Susan London

Major Finding: Each standard deviation rise in level of depressive symptoms at baseline, as assessed from CES-D score, was associated with a 26% increase in the risk of a first CHD event after inflammatory markers were taken into account.

Data Source: A population-based study of 1,794 randomly selected Nova Scotians with prospective 10-year follow-up.

Disclosures: Dr. Davidson reported that she had no relevant conflicts of interest.

VANCOUVER, B.C. — Depression increases the risk of coronary heart disease, but it does not do so through proinflammatory mechanisms, a study has shown.

In 1,794 healthy Nova Scotians randomly selected from the general population, those with higher levels of depressive symptoms did indeed have higher levels of inflammatory biomarkers.

But after these markers and traditional risk factors were taken into account, depressive symptoms still predicted first coronary heart disease (CHD) events over the next decade, with the risk rising by 26% with each standard deviation increase in Center for Epidemiologic Studies–Depression (CES-D) score.

“Inflammatory biomarkers neither fully nor partially explained the association between depression and CHD,” lead investigator Karina W. Davidson, Ph.D., said at the congress.

“We have also looked into recurrent CHD, and we did not find it,” she added. “So we are concluding that we need to explore other biological mechanisms, such as platelet aggregation or endothelial dysfunction.”

Several scenarios have been proposed to explain associations among depressive symptoms, inflammation, and CHD (Am. J. Cardiol. 2005;96:1016–21), according to Dr. Davidson of the department of medicine at Columbia University Medical Center in New York. In some of these scenarios, depression has effects on physiology or behavior that promote the development of CHD.

These effects could be indirect (for example, nonadherence to cardiac prevention regimens, increased unfavorable lifestyle behaviors, or cardiotoxicity of antidepressants) or direct (for example, increased inflammation, endothelial dysfunction, enhanced platelet aggregation, or autonomic and neuroendocrine perturbations).

“Depressive symptoms and inflammatory markers have been highly comorbid in a number of studies, so we felt this was a promising pattern to look at, whether inflammatory markers serve perhaps as the mechanism,” she explained.

Dr. Davidson and her colleagues studied apparently healthy adults recruited to the population-based Canadian Nova Scotia Health Survey in 1995.

At baseline, study participants underwent assessment of traditional risk factors such as lipids and smoking, completed the CES-D, and gave a blood sample for measurement of three inflammatory markers: high-sensitivity C-reactive protein (hs-CRP), soluble intercellular adhesion molecule (sICAM), and interleukin-6.

The study population was 46 years old on average and equally divided by sex. Participants' mean body mass index was about 27 kg/m

Participants had a mean score on the CES-D scale of 7.2 points out of a possible 60, with a standard deviation of 7.7 points. Only 3% of participants were taking antidepressants.

During a prospective 10-year follow-up, 8.5% of participants had a first CHD event, as ascertained from diagnostic codes on hospital admissions or death certificates, Dr. Davidson reported.

Each standard deviation increase in depressive symptoms was associated with an 8.3% higher hs-CRP level and a 4.7% higher interleukin-6 level.

In analyses adjusted for traditional risk factors, the higher participants' level of depressive symptoms, the greater their risk of CHD events, with a significant 28% greater risk for each standard deviation increase in CES-D score, she said.

The risk of events also rose with hs-CRP level and sICAM level (but not interleukin-6 level) at baseline, independent of traditional risk factors.

When all of the variables were included in a model, a higher level of depressive symptoms still significantly predicted a greater likelihood of CHD events, with a significant 26% greater risk for each standard deviation increase in CES-D score.

Furthermore, this association was consistent across a variety of subgroups: men, current smokers, participants aged 65 years or older, obese participants, those not prescribed any cardiac medications (although statins were not assessed), and those not prescribed antidepressants.

“Depression and inflammation are moderately correlated,” commented Dr. Davidson, but “depressive symptoms were independent for CHD, and inflammation was independent for CHD — there did not seem to be any kind of mechanistic association amongst them,” she added.

“We still believe in our hearts that there are some patients who have cytokine-induced depression,” Dr. Davidson concluded. “We are now trying the hypothesis that there may be a small, intermediary phenotype of depressed persons who actually have very high cytokines, and they may go on to have a higher rate of incident CHD, and we are testing them.”

Major Finding: Each standard deviation rise in level of depressive symptoms at baseline, as assessed from CES-D score, was associated with a 26% increase in the risk of a first CHD event after inflammatory markers were taken into account.

Data Source: A population-based study of 1,794 randomly selected Nova Scotians with prospective 10-year follow-up.

Disclosures: Dr. Davidson reported that she had no relevant conflicts of interest.

VANCOUVER, B.C. — Depression increases the risk of coronary heart disease, but it does not do so through proinflammatory mechanisms, a study has shown.

In 1,794 healthy Nova Scotians randomly selected from the general population, those with higher levels of depressive symptoms did indeed have higher levels of inflammatory biomarkers.

But after these markers and traditional risk factors were taken into account, depressive symptoms still predicted first coronary heart disease (CHD) events over the next decade, with the risk rising by 26% with each standard deviation increase in Center for Epidemiologic Studies–Depression (CES-D) score.

“Inflammatory biomarkers neither fully nor partially explained the association between depression and CHD,” lead investigator Karina W. Davidson, Ph.D., said at the congress.

“We have also looked into recurrent CHD, and we did not find it,” she added. “So we are concluding that we need to explore other biological mechanisms, such as platelet aggregation or endothelial dysfunction.”

Several scenarios have been proposed to explain associations among depressive symptoms, inflammation, and CHD (Am. J. Cardiol. 2005;96:1016–21), according to Dr. Davidson of the department of medicine at Columbia University Medical Center in New York. In some of these scenarios, depression has effects on physiology or behavior that promote the development of CHD.

These effects could be indirect (for example, nonadherence to cardiac prevention regimens, increased unfavorable lifestyle behaviors, or cardiotoxicity of antidepressants) or direct (for example, increased inflammation, endothelial dysfunction, enhanced platelet aggregation, or autonomic and neuroendocrine perturbations).

“Depressive symptoms and inflammatory markers have been highly comorbid in a number of studies, so we felt this was a promising pattern to look at, whether inflammatory markers serve perhaps as the mechanism,” she explained.

Dr. Davidson and her colleagues studied apparently healthy adults recruited to the population-based Canadian Nova Scotia Health Survey in 1995.

At baseline, study participants underwent assessment of traditional risk factors such as lipids and smoking, completed the CES-D, and gave a blood sample for measurement of three inflammatory markers: high-sensitivity C-reactive protein (hs-CRP), soluble intercellular adhesion molecule (sICAM), and interleukin-6.

The study population was 46 years old on average and equally divided by sex. Participants' mean body mass index was about 27 kg/m

Participants had a mean score on the CES-D scale of 7.2 points out of a possible 60, with a standard deviation of 7.7 points. Only 3% of participants were taking antidepressants.

During a prospective 10-year follow-up, 8.5% of participants had a first CHD event, as ascertained from diagnostic codes on hospital admissions or death certificates, Dr. Davidson reported.

Each standard deviation increase in depressive symptoms was associated with an 8.3% higher hs-CRP level and a 4.7% higher interleukin-6 level.

In analyses adjusted for traditional risk factors, the higher participants' level of depressive symptoms, the greater their risk of CHD events, with a significant 28% greater risk for each standard deviation increase in CES-D score, she said.

The risk of events also rose with hs-CRP level and sICAM level (but not interleukin-6 level) at baseline, independent of traditional risk factors.

When all of the variables were included in a model, a higher level of depressive symptoms still significantly predicted a greater likelihood of CHD events, with a significant 26% greater risk for each standard deviation increase in CES-D score.

Furthermore, this association was consistent across a variety of subgroups: men, current smokers, participants aged 65 years or older, obese participants, those not prescribed any cardiac medications (although statins were not assessed), and those not prescribed antidepressants.

“Depression and inflammation are moderately correlated,” commented Dr. Davidson, but “depressive symptoms were independent for CHD, and inflammation was independent for CHD — there did not seem to be any kind of mechanistic association amongst them,” she added.

“We still believe in our hearts that there are some patients who have cytokine-induced depression,” Dr. Davidson concluded. “We are now trying the hypothesis that there may be a small, intermediary phenotype of depressed persons who actually have very high cytokines, and they may go on to have a higher rate of incident CHD, and we are testing them.”

Major Finding: Each standard deviation rise in level of depressive symptoms at baseline, as assessed from CES-D score, was associated with a 26% increase in the risk of a first CHD event after inflammatory markers were taken into account.

Data Source: A population-based study of 1,794 randomly selected Nova Scotians with prospective 10-year follow-up.

Disclosures: Dr. Davidson reported that she had no relevant conflicts of interest.

VANCOUVER, B.C. — Depression increases the risk of coronary heart disease, but it does not do so through proinflammatory mechanisms, a study has shown.

In 1,794 healthy Nova Scotians randomly selected from the general population, those with higher levels of depressive symptoms did indeed have higher levels of inflammatory biomarkers.

But after these markers and traditional risk factors were taken into account, depressive symptoms still predicted first coronary heart disease (CHD) events over the next decade, with the risk rising by 26% with each standard deviation increase in Center for Epidemiologic Studies–Depression (CES-D) score.

“Inflammatory biomarkers neither fully nor partially explained the association between depression and CHD,” lead investigator Karina W. Davidson, Ph.D., said at the congress.

“We have also looked into recurrent CHD, and we did not find it,” she added. “So we are concluding that we need to explore other biological mechanisms, such as platelet aggregation or endothelial dysfunction.”

Several scenarios have been proposed to explain associations among depressive symptoms, inflammation, and CHD (Am. J. Cardiol. 2005;96:1016–21), according to Dr. Davidson of the department of medicine at Columbia University Medical Center in New York. In some of these scenarios, depression has effects on physiology or behavior that promote the development of CHD.

These effects could be indirect (for example, nonadherence to cardiac prevention regimens, increased unfavorable lifestyle behaviors, or cardiotoxicity of antidepressants) or direct (for example, increased inflammation, endothelial dysfunction, enhanced platelet aggregation, or autonomic and neuroendocrine perturbations).

“Depressive symptoms and inflammatory markers have been highly comorbid in a number of studies, so we felt this was a promising pattern to look at, whether inflammatory markers serve perhaps as the mechanism,” she explained.

Dr. Davidson and her colleagues studied apparently healthy adults recruited to the population-based Canadian Nova Scotia Health Survey in 1995.

At baseline, study participants underwent assessment of traditional risk factors such as lipids and smoking, completed the CES-D, and gave a blood sample for measurement of three inflammatory markers: high-sensitivity C-reactive protein (hs-CRP), soluble intercellular adhesion molecule (sICAM), and interleukin-6.

The study population was 46 years old on average and equally divided by sex. Participants' mean body mass index was about 27 kg/m

Participants had a mean score on the CES-D scale of 7.2 points out of a possible 60, with a standard deviation of 7.7 points. Only 3% of participants were taking antidepressants.

During a prospective 10-year follow-up, 8.5% of participants had a first CHD event, as ascertained from diagnostic codes on hospital admissions or death certificates, Dr. Davidson reported.

Each standard deviation increase in depressive symptoms was associated with an 8.3% higher hs-CRP level and a 4.7% higher interleukin-6 level.

In analyses adjusted for traditional risk factors, the higher participants' level of depressive symptoms, the greater their risk of CHD events, with a significant 28% greater risk for each standard deviation increase in CES-D score, she said.

The risk of events also rose with hs-CRP level and sICAM level (but not interleukin-6 level) at baseline, independent of traditional risk factors.

When all of the variables were included in a model, a higher level of depressive symptoms still significantly predicted a greater likelihood of CHD events, with a significant 26% greater risk for each standard deviation increase in CES-D score.

Furthermore, this association was consistent across a variety of subgroups: men, current smokers, participants aged 65 years or older, obese participants, those not prescribed any cardiac medications (although statins were not assessed), and those not prescribed antidepressants.

“Depression and inflammation are moderately correlated,” commented Dr. Davidson, but “depressive symptoms were independent for CHD, and inflammation was independent for CHD — there did not seem to be any kind of mechanistic association amongst them,” she added.

“We still believe in our hearts that there are some patients who have cytokine-induced depression,” Dr. Davidson concluded. “We are now trying the hypothesis that there may be a small, intermediary phenotype of depressed persons who actually have very high cytokines, and they may go on to have a higher rate of incident CHD, and we are testing them.”

Major Finding: Perceived burdensomeness was independently associated with suicidal ideation among patients with chronic pain, and a model including this measure correctly classified 95% of patients as to the presence or absence of suicidal ideation.

Data Source: A retrospective study of 109 patients in a military population with chronic pain referred to a psychology clinic.

Disclosures: Dr. Kanzler reported that she had no conflicts of interest related to the study.

SEATTLE – Asking patients with chronic pain a single question – “Do you believe it would be better for everyone involved if you were to die?” – can determine whether he or she is having suicidal thoughts or wishes, findings from a retrospective study suggest.

Among 109 patients with chronic pain, patients' perceptions that they were a burden to others as assessed with this question was the sole independent predictor of suicidal ideation even after depression and hopelessness were taken into account.

A model including perceived burdensomeness, in addition to conventional risk factors, correctly classified 95% of the patients regarding the presence or absence of suicidal ideation.

“It's important to consider perceived burdensomeness in the patients that you see,” advised lead investigator Kathryn E. Kanzler, Psy.D., who is a captain in the U.S. Air Force and a psychologist at Lackland Air Force Base in San Antonio. “Just one question – that's all it takes to get kind of a quick snapshot of what's going on.”

Patients with chronic conditions may be uniquely attuned to the impact of their health on their caregivers, Dr. Kanzler told attendees of the annual meeting of the Society of Behavioral Medicine. “Research has found that self-perceived burden … can have a direct impact on significant medical decision making,” she said, such as choosing to reduce or entirely skip dialysis.

In the study, she and her colleagues retrospectively reviewed the medical records of 109 outpatients with chronic pain who were referred to a psychology clinic for evaluation and possible behavioral and psychosocial interventions. All were active or retired military personnel, or their dependents or family members.

The patients were age 42 years on average. The majority were married (72%), female (65%), and white (66%).

The leading primary cause of pain was headache/migraine (seen in 28% of patients), followed by chronic low back pain (16%), fibromyalgia (13%), temporomandibular or myofascial pain (9%), arthritis (3%), and complex regional pain syndrome (1%). The remaining patients (30%) had pain due to other conditions, such as cancer or orthopedic injuries.

The investigators used responses on the Beck Depression Inventory–Second Edition (BDI-II) to assess patients' hopelessness, suicidal ideation, and depression.

Perceived burdensomeness was assessed from responses to a single statement, “It would be better for everyone involved if I were to die,” with possible response options ranging from 1 (never or none of the time) to 5 (always or a great many times).

Overall, 7% of patients were found to have suicidal ideation, Dr. Kanzler reported. A logistic regression model including age, sex, race, marital status, depression, and hopelessness improved the ability to predict suicidal ideation above a null model.

Adding patients' perceived burdensomeness to this model further improved the ability to predict suicidal ideation and also improved model fit.

When controlling for depression and hopelessness, perceived burdensomeness was the sole independent predictor of suicidal ideation.

There was no difference in the findings between patients who did and did not have an identified caregiver, a finding that corroborated those from other studies suggesting that perceived burdensomeness may apply to the people who are important in one's life generally.

Perceived burdensomeness performed better at correctly classifying patients without suicidal ideation (98%) than at correctly classifying those with suicidal ideation (63%).

“We hope this study adds to the understanding of the really complex relationship between chronic pain and suicide ideation,” Dr. Kanzler said. “Perceived burdensomeness as a risk factor might help explain high rates of suicide ideation beyond the types of things that definitely, immediately come to mind.”

Importantly, she noted, perceived burdensomeness is modifiable, in contrast to many of the other risk factors for suicidal ideation, such as age and sex. “Some kind of a cognitive intervention might be useful,” she proposed, such as intervening to change the meaning of the cognition of perceived burdensomeness or to challenge the cognition itself.

Encouraging increased communication with the key people in a patient's life may also be beneficial, according to Dr. Kanzler. “Sometimes, especially in our population, there is not necessarily an identified caregiver, but this perceived burdensomeness kind of affects the whole group that surrounds that person,” she explained. “So that type of intervention might also be useful, going beyond the individual patient.”

Major Finding: Perceived burdensomeness was independently associated with suicidal ideation among patients with chronic pain, and a model including this measure correctly classified 95% of patients as to the presence or absence of suicidal ideation.

Data Source: A retrospective study of 109 patients in a military population with chronic pain referred to a psychology clinic.

Disclosures: Dr. Kanzler reported that she had no conflicts of interest related to the study.

SEATTLE – Asking patients with chronic pain a single question – “Do you believe it would be better for everyone involved if you were to die?” – can determine whether he or she is having suicidal thoughts or wishes, findings from a retrospective study suggest.

Among 109 patients with chronic pain, patients' perceptions that they were a burden to others as assessed with this question was the sole independent predictor of suicidal ideation even after depression and hopelessness were taken into account.

A model including perceived burdensomeness, in addition to conventional risk factors, correctly classified 95% of the patients regarding the presence or absence of suicidal ideation.

“It's important to consider perceived burdensomeness in the patients that you see,” advised lead investigator Kathryn E. Kanzler, Psy.D., who is a captain in the U.S. Air Force and a psychologist at Lackland Air Force Base in San Antonio. “Just one question – that's all it takes to get kind of a quick snapshot of what's going on.”

Patients with chronic conditions may be uniquely attuned to the impact of their health on their caregivers, Dr. Kanzler told attendees of the annual meeting of the Society of Behavioral Medicine. “Research has found that self-perceived burden … can have a direct impact on significant medical decision making,” she said, such as choosing to reduce or entirely skip dialysis.

In the study, she and her colleagues retrospectively reviewed the medical records of 109 outpatients with chronic pain who were referred to a psychology clinic for evaluation and possible behavioral and psychosocial interventions. All were active or retired military personnel, or their dependents or family members.

The patients were age 42 years on average. The majority were married (72%), female (65%), and white (66%).

The leading primary cause of pain was headache/migraine (seen in 28% of patients), followed by chronic low back pain (16%), fibromyalgia (13%), temporomandibular or myofascial pain (9%), arthritis (3%), and complex regional pain syndrome (1%). The remaining patients (30%) had pain due to other conditions, such as cancer or orthopedic injuries.

The investigators used responses on the Beck Depression Inventory–Second Edition (BDI-II) to assess patients' hopelessness, suicidal ideation, and depression.

Perceived burdensomeness was assessed from responses to a single statement, “It would be better for everyone involved if I were to die,” with possible response options ranging from 1 (never or none of the time) to 5 (always or a great many times).

Overall, 7% of patients were found to have suicidal ideation, Dr. Kanzler reported. A logistic regression model including age, sex, race, marital status, depression, and hopelessness improved the ability to predict suicidal ideation above a null model.

Adding patients' perceived burdensomeness to this model further improved the ability to predict suicidal ideation and also improved model fit.

When controlling for depression and hopelessness, perceived burdensomeness was the sole independent predictor of suicidal ideation.

There was no difference in the findings between patients who did and did not have an identified caregiver, a finding that corroborated those from other studies suggesting that perceived burdensomeness may apply to the people who are important in one's life generally.

Perceived burdensomeness performed better at correctly classifying patients without suicidal ideation (98%) than at correctly classifying those with suicidal ideation (63%).

“We hope this study adds to the understanding of the really complex relationship between chronic pain and suicide ideation,” Dr. Kanzler said. “Perceived burdensomeness as a risk factor might help explain high rates of suicide ideation beyond the types of things that definitely, immediately come to mind.”

Importantly, she noted, perceived burdensomeness is modifiable, in contrast to many of the other risk factors for suicidal ideation, such as age and sex. “Some kind of a cognitive intervention might be useful,” she proposed, such as intervening to change the meaning of the cognition of perceived burdensomeness or to challenge the cognition itself.

Encouraging increased communication with the key people in a patient's life may also be beneficial, according to Dr. Kanzler. “Sometimes, especially in our population, there is not necessarily an identified caregiver, but this perceived burdensomeness kind of affects the whole group that surrounds that person,” she explained. “So that type of intervention might also be useful, going beyond the individual patient.”

Major Finding: Perceived burdensomeness was independently associated with suicidal ideation among patients with chronic pain, and a model including this measure correctly classified 95% of patients as to the presence or absence of suicidal ideation.

Data Source: A retrospective study of 109 patients in a military population with chronic pain referred to a psychology clinic.

Disclosures: Dr. Kanzler reported that she had no conflicts of interest related to the study.

SEATTLE – Asking patients with chronic pain a single question – “Do you believe it would be better for everyone involved if you were to die?” – can determine whether he or she is having suicidal thoughts or wishes, findings from a retrospective study suggest.

Among 109 patients with chronic pain, patients' perceptions that they were a burden to others as assessed with this question was the sole independent predictor of suicidal ideation even after depression and hopelessness were taken into account.

A model including perceived burdensomeness, in addition to conventional risk factors, correctly classified 95% of the patients regarding the presence or absence of suicidal ideation.

“It's important to consider perceived burdensomeness in the patients that you see,” advised lead investigator Kathryn E. Kanzler, Psy.D., who is a captain in the U.S. Air Force and a psychologist at Lackland Air Force Base in San Antonio. “Just one question – that's all it takes to get kind of a quick snapshot of what's going on.”

Patients with chronic conditions may be uniquely attuned to the impact of their health on their caregivers, Dr. Kanzler told attendees of the annual meeting of the Society of Behavioral Medicine. “Research has found that self-perceived burden … can have a direct impact on significant medical decision making,” she said, such as choosing to reduce or entirely skip dialysis.

In the study, she and her colleagues retrospectively reviewed the medical records of 109 outpatients with chronic pain who were referred to a psychology clinic for evaluation and possible behavioral and psychosocial interventions. All were active or retired military personnel, or their dependents or family members.

The patients were age 42 years on average. The majority were married (72%), female (65%), and white (66%).

The leading primary cause of pain was headache/migraine (seen in 28% of patients), followed by chronic low back pain (16%), fibromyalgia (13%), temporomandibular or myofascial pain (9%), arthritis (3%), and complex regional pain syndrome (1%). The remaining patients (30%) had pain due to other conditions, such as cancer or orthopedic injuries.

The investigators used responses on the Beck Depression Inventory–Second Edition (BDI-II) to assess patients' hopelessness, suicidal ideation, and depression.

Perceived burdensomeness was assessed from responses to a single statement, “It would be better for everyone involved if I were to die,” with possible response options ranging from 1 (never or none of the time) to 5 (always or a great many times).

Overall, 7% of patients were found to have suicidal ideation, Dr. Kanzler reported. A logistic regression model including age, sex, race, marital status, depression, and hopelessness improved the ability to predict suicidal ideation above a null model.

Adding patients' perceived burdensomeness to this model further improved the ability to predict suicidal ideation and also improved model fit.

When controlling for depression and hopelessness, perceived burdensomeness was the sole independent predictor of suicidal ideation.

There was no difference in the findings between patients who did and did not have an identified caregiver, a finding that corroborated those from other studies suggesting that perceived burdensomeness may apply to the people who are important in one's life generally.

Perceived burdensomeness performed better at correctly classifying patients without suicidal ideation (98%) than at correctly classifying those with suicidal ideation (63%).

“We hope this study adds to the understanding of the really complex relationship between chronic pain and suicide ideation,” Dr. Kanzler said. “Perceived burdensomeness as a risk factor might help explain high rates of suicide ideation beyond the types of things that definitely, immediately come to mind.”

Importantly, she noted, perceived burdensomeness is modifiable, in contrast to many of the other risk factors for suicidal ideation, such as age and sex. “Some kind of a cognitive intervention might be useful,” she proposed, such as intervening to change the meaning of the cognition of perceived burdensomeness or to challenge the cognition itself.

Encouraging increased communication with the key people in a patient's life may also be beneficial, according to Dr. Kanzler. “Sometimes, especially in our population, there is not necessarily an identified caregiver, but this perceived burdensomeness kind of affects the whole group that surrounds that person,” she explained. “So that type of intervention might also be useful, going beyond the individual patient.”

Major Finding: Compared with women who had not had gestational diabetes, women who had the condition were 41% more likely to develop hypertension during a period of up to 14 years.

Data Source: A nested prospective cohort study of 26,384 women initially aged 25-44 years who had at least one singleton pregnancy.

Disclosures: Ms. Tobias reported that she had no relevant conflicts of interest.

SEATTLE — Women who have had gestational diabetes may be at elevated risk for hypertension even after established risk factors are taken into account, a nested cohort study has shown.

Using data from the Nurses' Health Study II, researchers followed more than 26,000 women from an index pregnancy for up to 14 years. Those who had gestational diabetes during that pregnancy were 41% more likely to develop hypertension even after adjustment for potential confounders such as body mass index, diet, and a family history of hypertension.

“These women may represent a target group for intervention to prevent or delay the onset of hypertension, which is a public health concern in the United States,” lead investigator Deirdre K. Tobias said at the meeting.

“The mechanism by which gestational diabetes mellitus could lead to an increased risk of hypertension and other metabolic complications is not yet established,” said Ms. Tobias, a doctoral student at Harvard School of Public Health in Boston.

One possibility is that these conditions have shared risk factors, she noted. Another is that gestational diabetes itself increases the risk of hypertension, for example, by causing vascular damage that manifests later in time.

The Nurses' Health Study II is a prospective cohort study of women 25-44 years old at baseline that began in 1989.

Women in the study complete extensive questionnaires at baseline and biennially, which include questions about physician-diagnosed gestational diabetes and hypertension.

Ms. Tobias and her colleagues included in their sample the 26,384 women who reported having at least one singleton pregnancy between 1991 (the first year in which dietary data were collected) and 2005, did not have type 2 diabetes or hypertension, had not had gestational diabetes in a previous pregnancy, and had not had previous cardiovascular disease.

Study results showed that 6% of the women had gestational diabetes during their index pregnancy. Both the women with and without gestational diabetes had a mean age of 32 years at baseline, and 7% in each group were current smokers.

However, those with gestational diabetes had a higher body mass index (25 vs. 23 kg/m

Overall, 9% of the women developed hypertension during follow-up, and the cumulative incidence was higher among those who had had gestational diabetes.

After adjustment for a dozen potential confounders, women who had gestational diabetes still had a 41% higher risk of developing hypertension (hazard ratio, 1.41).

Moreover, compared with women who had neither gestational diabetes nor type 2 diabetes, those who had both had a near tripling of the risk of developing hypertension (HR, 2.95).

“It is possible that there was residual or unmeasured confounding,” acknowledged Ms. Tobias. “For example, in our cohort, we were unable to capture pregnancy-related characteristics, such as weight gain or severity of gestational diabetes.”

In addition, she noted, the generalizability of the study's findings may be limited, given the women's somewhat higher age at baseline and the fact that most were white.

Major Finding: Compared with women who had not had gestational diabetes, women who had the condition were 41% more likely to develop hypertension during a period of up to 14 years.

Data Source: A nested prospective cohort study of 26,384 women initially aged 25-44 years who had at least one singleton pregnancy.

Disclosures: Ms. Tobias reported that she had no relevant conflicts of interest.

SEATTLE — Women who have had gestational diabetes may be at elevated risk for hypertension even after established risk factors are taken into account, a nested cohort study has shown.

Using data from the Nurses' Health Study II, researchers followed more than 26,000 women from an index pregnancy for up to 14 years. Those who had gestational diabetes during that pregnancy were 41% more likely to develop hypertension even after adjustment for potential confounders such as body mass index, diet, and a family history of hypertension.

“These women may represent a target group for intervention to prevent or delay the onset of hypertension, which is a public health concern in the United States,” lead investigator Deirdre K. Tobias said at the meeting.

“The mechanism by which gestational diabetes mellitus could lead to an increased risk of hypertension and other metabolic complications is not yet established,” said Ms. Tobias, a doctoral student at Harvard School of Public Health in Boston.

One possibility is that these conditions have shared risk factors, she noted. Another is that gestational diabetes itself increases the risk of hypertension, for example, by causing vascular damage that manifests later in time.

The Nurses' Health Study II is a prospective cohort study of women 25-44 years old at baseline that began in 1989.

Women in the study complete extensive questionnaires at baseline and biennially, which include questions about physician-diagnosed gestational diabetes and hypertension.

Ms. Tobias and her colleagues included in their sample the 26,384 women who reported having at least one singleton pregnancy between 1991 (the first year in which dietary data were collected) and 2005, did not have type 2 diabetes or hypertension, had not had gestational diabetes in a previous pregnancy, and had not had previous cardiovascular disease.

Study results showed that 6% of the women had gestational diabetes during their index pregnancy. Both the women with and without gestational diabetes had a mean age of 32 years at baseline, and 7% in each group were current smokers.

However, those with gestational diabetes had a higher body mass index (25 vs. 23 kg/m

Overall, 9% of the women developed hypertension during follow-up, and the cumulative incidence was higher among those who had had gestational diabetes.

After adjustment for a dozen potential confounders, women who had gestational diabetes still had a 41% higher risk of developing hypertension (hazard ratio, 1.41).

Moreover, compared with women who had neither gestational diabetes nor type 2 diabetes, those who had both had a near tripling of the risk of developing hypertension (HR, 2.95).

“It is possible that there was residual or unmeasured confounding,” acknowledged Ms. Tobias. “For example, in our cohort, we were unable to capture pregnancy-related characteristics, such as weight gain or severity of gestational diabetes.”

In addition, she noted, the generalizability of the study's findings may be limited, given the women's somewhat higher age at baseline and the fact that most were white.

Major Finding: Compared with women who had not had gestational diabetes, women who had the condition were 41% more likely to develop hypertension during a period of up to 14 years.

Data Source: A nested prospective cohort study of 26,384 women initially aged 25-44 years who had at least one singleton pregnancy.

Disclosures: Ms. Tobias reported that she had no relevant conflicts of interest.

SEATTLE — Women who have had gestational diabetes may be at elevated risk for hypertension even after established risk factors are taken into account, a nested cohort study has shown.

Using data from the Nurses' Health Study II, researchers followed more than 26,000 women from an index pregnancy for up to 14 years. Those who had gestational diabetes during that pregnancy were 41% more likely to develop hypertension even after adjustment for potential confounders such as body mass index, diet, and a family history of hypertension.

“These women may represent a target group for intervention to prevent or delay the onset of hypertension, which is a public health concern in the United States,” lead investigator Deirdre K. Tobias said at the meeting.

“The mechanism by which gestational diabetes mellitus could lead to an increased risk of hypertension and other metabolic complications is not yet established,” said Ms. Tobias, a doctoral student at Harvard School of Public Health in Boston.

One possibility is that these conditions have shared risk factors, she noted. Another is that gestational diabetes itself increases the risk of hypertension, for example, by causing vascular damage that manifests later in time.

The Nurses' Health Study II is a prospective cohort study of women 25-44 years old at baseline that began in 1989.

Women in the study complete extensive questionnaires at baseline and biennially, which include questions about physician-diagnosed gestational diabetes and hypertension.

Ms. Tobias and her colleagues included in their sample the 26,384 women who reported having at least one singleton pregnancy between 1991 (the first year in which dietary data were collected) and 2005, did not have type 2 diabetes or hypertension, had not had gestational diabetes in a previous pregnancy, and had not had previous cardiovascular disease.

Study results showed that 6% of the women had gestational diabetes during their index pregnancy. Both the women with and without gestational diabetes had a mean age of 32 years at baseline, and 7% in each group were current smokers.

However, those with gestational diabetes had a higher body mass index (25 vs. 23 kg/m

Overall, 9% of the women developed hypertension during follow-up, and the cumulative incidence was higher among those who had had gestational diabetes.

After adjustment for a dozen potential confounders, women who had gestational diabetes still had a 41% higher risk of developing hypertension (hazard ratio, 1.41).

Moreover, compared with women who had neither gestational diabetes nor type 2 diabetes, those who had both had a near tripling of the risk of developing hypertension (HR, 2.95).

“It is possible that there was residual or unmeasured confounding,” acknowledged Ms. Tobias. “For example, in our cohort, we were unable to capture pregnancy-related characteristics, such as weight gain or severity of gestational diabetes.”

In addition, she noted, the generalizability of the study's findings may be limited, given the women's somewhat higher age at baseline and the fact that most were white.

Major Finding: Normal-weight women who took multivitamins in the periconceptional period were 16% less likely to give birth preterm than were women who did not take them; the benefit was due to a reduced risk of preterm birth after spontaneous preterm labor.

Data Source: An observational study of 27,259 women with singleton pregnancies enrolled in the Danish National Birth Cohort.

Disclosures: Dr. Catov reported that she had no relevant conflicts of interest.

SEATTLE — Use of multivitamins around the time of conception may protect against preterm birth, but the benefit depends on a woman's weight and the type of preterm birth, according to findings of a cohort study among more than 27,000 Danish women.

Normal-weight women were 16% less likely to give birth preterm if they took multivitamins in the periconceptional period. This reduction was due to a lower risk of preterm birth after spontaneous (idiopathic) preterm labor. In contrast, overweight women were not less likely to have a preterm birth if they took multivitamins periconceptionally. Also, use did not reduce the risk of preterm births that were medically indicated or that occurred after premature rupture of membranes (PROM).

“Our data suggest that multivitamin use around the time of conception and implantation may reduce risk of idiopathic preterm labor among normal-weight women,” lead investigator Janet M. Catov, Ph.D., said at the meeting. The dose of vitamins may have been insufficient in overweight women or perhaps they had higher levels of systemic inflammation, she speculated. “As we better understand the complexity of overweight and obesity in pregnancy, I hope that we might better understand that.”

“The fact that we did not see a relationship to membrane rupture suggests that what we might be describing are more sort of early and important characteristics of pregnancy that may… set the stage, if you will, for successful pregnancy or less successful pregnancy,” she commented.

Explaining the rationale for the study, Dr. Catov, assistant professor of epidemiology at the University of Pittsburgh, noted that women's nutritional status is likely important for placentation, and abnormal placentation has been associated with both spontaneous and medically indicated preterm births. “Previous work by our group and others has suggested that periconceptional multivitamin use is related to reduced risk for preeclampsia, early preterm birth, and growth restriction.”

The investigators analyzed data from the Danish National Birth Cohort, which enrolled pregnant women in Denmark between 1997 and 2003.

Analyses were based on women who were recruited at more than 5 weeks' but less than 24 weeks' gestation of a singleton pregnancy, provided detailed information on vitamin use in the periconceptional period (extending from 4 weeks before the last menstrual period to 8 weeks after), and had a live birth.

To look more closely at the issue of timing of multivitamin use, they subdivided the periconceptional period into a preconceptional period (4 weeks before last menstrual period to 2 weeks after) and a postconceptional period (2 weeks after last menstrual period to 8 weeks after). To look at the issue of frequency of multivitamin use, within each 6-week period, they classified use as partial (3 weeks or less) or regular (4 weeks or more).

A total of 19,677 women reported at least some use of multivitamins periconceptionally, and 7,582 did not report any use. Users were somewhat less likely to be younger than age 25 than were nonusers (13% vs. 19%, respectively), to be overweight, defined as having a prepregnancy body mass index of 25 kg/m

Overall, in adjusted analyses, multivitamin users were significantly less likely than nonusers to give birth preterm, meaning before 37 weeks' gestation (hazard ratio, 0.88). But after stratification by weight, this benefit was seen only among women who were of normal weight, defined as having a prepregnancy BMI of less than 25 kg/m

Timing and frequency of multivitamin use were important, according to Dr. Catov. Normal-weight women were significantly less likely to have a preterm birth if they partially used multivitamins preconception and regularly used them post conception (HR, 0.77) or regularly used them in both periods (HR, 0.82). However, they did not have a significant reduction in risk if they only used them regularly post conception or used them partially in both periods.

“Future studies are needed to determine the actual nutrients that might be involved … and to really understand what mechanisms might be involved,” she concluded. “We also need to better understand the relationship between periconceptional multivitamin use and adverse events, and that work is actually under way with our colleagues in Denmark.”

The effect was limited to idiopathic preterm labor in normal-weight women.

Source ©dragon fang/Fotolia.com

Major Finding: Normal-weight women who took multivitamins in the periconceptional period were 16% less likely to give birth preterm than were women who did not take them; the benefit was due to a reduced risk of preterm birth after spontaneous preterm labor.

Data Source: An observational study of 27,259 women with singleton pregnancies enrolled in the Danish National Birth Cohort.

Disclosures: Dr. Catov reported that she had no relevant conflicts of interest.

SEATTLE — Use of multivitamins around the time of conception may protect against preterm birth, but the benefit depends on a woman's weight and the type of preterm birth, according to findings of a cohort study among more than 27,000 Danish women.

Normal-weight women were 16% less likely to give birth preterm if they took multivitamins in the periconceptional period. This reduction was due to a lower risk of preterm birth after spontaneous (idiopathic) preterm labor. In contrast, overweight women were not less likely to have a preterm birth if they took multivitamins periconceptionally. Also, use did not reduce the risk of preterm births that were medically indicated or that occurred after premature rupture of membranes (PROM).

“Our data suggest that multivitamin use around the time of conception and implantation may reduce risk of idiopathic preterm labor among normal-weight women,” lead investigator Janet M. Catov, Ph.D., said at the meeting. The dose of vitamins may have been insufficient in overweight women or perhaps they had higher levels of systemic inflammation, she speculated. “As we better understand the complexity of overweight and obesity in pregnancy, I hope that we might better understand that.”

“The fact that we did not see a relationship to membrane rupture suggests that what we might be describing are more sort of early and important characteristics of pregnancy that may… set the stage, if you will, for successful pregnancy or less successful pregnancy,” she commented.

Explaining the rationale for the study, Dr. Catov, assistant professor of epidemiology at the University of Pittsburgh, noted that women's nutritional status is likely important for placentation, and abnormal placentation has been associated with both spontaneous and medically indicated preterm births. “Previous work by our group and others has suggested that periconceptional multivitamin use is related to reduced risk for preeclampsia, early preterm birth, and growth restriction.”

The investigators analyzed data from the Danish National Birth Cohort, which enrolled pregnant women in Denmark between 1997 and 2003.

Analyses were based on women who were recruited at more than 5 weeks' but less than 24 weeks' gestation of a singleton pregnancy, provided detailed information on vitamin use in the periconceptional period (extending from 4 weeks before the last menstrual period to 8 weeks after), and had a live birth.

To look more closely at the issue of timing of multivitamin use, they subdivided the periconceptional period into a preconceptional period (4 weeks before last menstrual period to 2 weeks after) and a postconceptional period (2 weeks after last menstrual period to 8 weeks after). To look at the issue of frequency of multivitamin use, within each 6-week period, they classified use as partial (3 weeks or less) or regular (4 weeks or more).

A total of 19,677 women reported at least some use of multivitamins periconceptionally, and 7,582 did not report any use. Users were somewhat less likely to be younger than age 25 than were nonusers (13% vs. 19%, respectively), to be overweight, defined as having a prepregnancy body mass index of 25 kg/m

Overall, in adjusted analyses, multivitamin users were significantly less likely than nonusers to give birth preterm, meaning before 37 weeks' gestation (hazard ratio, 0.88). But after stratification by weight, this benefit was seen only among women who were of normal weight, defined as having a prepregnancy BMI of less than 25 kg/m

Timing and frequency of multivitamin use were important, according to Dr. Catov. Normal-weight women were significantly less likely to have a preterm birth if they partially used multivitamins preconception and regularly used them post conception (HR, 0.77) or regularly used them in both periods (HR, 0.82). However, they did not have a significant reduction in risk if they only used them regularly post conception or used them partially in both periods.

“Future studies are needed to determine the actual nutrients that might be involved … and to really understand what mechanisms might be involved,” she concluded. “We also need to better understand the relationship between periconceptional multivitamin use and adverse events, and that work is actually under way with our colleagues in Denmark.”

The effect was limited to idiopathic preterm labor in normal-weight women.

Source ©dragon fang/Fotolia.com

Major Finding: Normal-weight women who took multivitamins in the periconceptional period were 16% less likely to give birth preterm than were women who did not take them; the benefit was due to a reduced risk of preterm birth after spontaneous preterm labor.

Data Source: An observational study of 27,259 women with singleton pregnancies enrolled in the Danish National Birth Cohort.

Disclosures: Dr. Catov reported that she had no relevant conflicts of interest.

SEATTLE — Use of multivitamins around the time of conception may protect against preterm birth, but the benefit depends on a woman's weight and the type of preterm birth, according to findings of a cohort study among more than 27,000 Danish women.

Normal-weight women were 16% less likely to give birth preterm if they took multivitamins in the periconceptional period. This reduction was due to a lower risk of preterm birth after spontaneous (idiopathic) preterm labor. In contrast, overweight women were not less likely to have a preterm birth if they took multivitamins periconceptionally. Also, use did not reduce the risk of preterm births that were medically indicated or that occurred after premature rupture of membranes (PROM).

“Our data suggest that multivitamin use around the time of conception and implantation may reduce risk of idiopathic preterm labor among normal-weight women,” lead investigator Janet M. Catov, Ph.D., said at the meeting. The dose of vitamins may have been insufficient in overweight women or perhaps they had higher levels of systemic inflammation, she speculated. “As we better understand the complexity of overweight and obesity in pregnancy, I hope that we might better understand that.”

“The fact that we did not see a relationship to membrane rupture suggests that what we might be describing are more sort of early and important characteristics of pregnancy that may… set the stage, if you will, for successful pregnancy or less successful pregnancy,” she commented.

Explaining the rationale for the study, Dr. Catov, assistant professor of epidemiology at the University of Pittsburgh, noted that women's nutritional status is likely important for placentation, and abnormal placentation has been associated with both spontaneous and medically indicated preterm births. “Previous work by our group and others has suggested that periconceptional multivitamin use is related to reduced risk for preeclampsia, early preterm birth, and growth restriction.”

The investigators analyzed data from the Danish National Birth Cohort, which enrolled pregnant women in Denmark between 1997 and 2003.

Analyses were based on women who were recruited at more than 5 weeks' but less than 24 weeks' gestation of a singleton pregnancy, provided detailed information on vitamin use in the periconceptional period (extending from 4 weeks before the last menstrual period to 8 weeks after), and had a live birth.

To look more closely at the issue of timing of multivitamin use, they subdivided the periconceptional period into a preconceptional period (4 weeks before last menstrual period to 2 weeks after) and a postconceptional period (2 weeks after last menstrual period to 8 weeks after). To look at the issue of frequency of multivitamin use, within each 6-week period, they classified use as partial (3 weeks or less) or regular (4 weeks or more).

A total of 19,677 women reported at least some use of multivitamins periconceptionally, and 7,582 did not report any use. Users were somewhat less likely to be younger than age 25 than were nonusers (13% vs. 19%, respectively), to be overweight, defined as having a prepregnancy body mass index of 25 kg/m

Overall, in adjusted analyses, multivitamin users were significantly less likely than nonusers to give birth preterm, meaning before 37 weeks' gestation (hazard ratio, 0.88). But after stratification by weight, this benefit was seen only among women who were of normal weight, defined as having a prepregnancy BMI of less than 25 kg/m

Timing and frequency of multivitamin use were important, according to Dr. Catov. Normal-weight women were significantly less likely to have a preterm birth if they partially used multivitamins preconception and regularly used them post conception (HR, 0.77) or regularly used them in both periods (HR, 0.82). However, they did not have a significant reduction in risk if they only used them regularly post conception or used them partially in both periods.

“Future studies are needed to determine the actual nutrients that might be involved … and to really understand what mechanisms might be involved,” she concluded. “We also need to better understand the relationship between periconceptional multivitamin use and adverse events, and that work is actually under way with our colleagues in Denmark.”

The effect was limited to idiopathic preterm labor in normal-weight women.

Source ©dragon fang/Fotolia.com

Major Finding: Some 11% of women reported abuse in the period surrounding pregnancy. Abused women who were younger than age 20 or were aged 35 or older, or who were depressed before pregnancy, were more than twice as likely to have postpartum depression.

Data Source: The Maternity Experiences Survey, a population-based survey of a weighted sample of 76,508 women who were an average of 7 months post partum.

Disclosures: Dr. O'Campo and Dr. Janssen reported having no relevant conflicts of interest.

SEATTLE — Abuse is prevalent in the period surrounding pregnancy and is associated with a higher rate of postpartum depression, new data show. But certain factors might help identify abused women who are at highest risk and might benefit from targeted intervention.

In a national survey of Canadian women who had had a recent birth, 11% reported abuse in the past 2 years, according to study results presented at the meeting. Abused women were roughly three times more likely than their nonabused counterparts to have postpartum depression. Within the abused group, this risk was more than doubled for teenagers and women aged at least 35 years, as well as for women who were depressed before pregnancy.

Prevalence of Abuse

Studies in recent decades have explored the issue of violence around the time of pregnancy, noted coinvestigator Patricia J. O'Campo, Ph.D. “Yet, despite that, we actually don't have good estimates of prevalence,” she said. Reasons include variation in the types of abuse captured, the time period and perpetrators assessed, and the women studied.

In the national, population-based Maternity Experiences Survey, Dr. O'Campo and her colleagues used census data to identify Canadian women with a singleton infant aged 5-14 months. A random sample was selected for a computer-assisted telephone interview that asked whether they had experienced any of 10 types of abuse (actual or threatened) in the past 2 years, as well as questions from the EPDS (Edinburgh Postnatal Depression Scale).

Interviews were conducted with 6,421 women, who represented a weighted sample of 76,508 women, according to Dr. O'Campo, an epidemiologist at the University of Toronto. Most were 5-9 months post partum (average, 7.3 months).

Fully 11% of the women reported experiencing abuse in the past 2 years. In stratified analyses, the prevalence was highest among teenagers (40%); women with an annual income less than $20,000 (28%); aboriginal women (30%); and nonmarried, noncohabiting women (35%). By far, the leading perpetrators were partners (reported by 6% overall), followed by family members (2%), strangers/others (2%), and friends (1%).

Abused women most often said they had been pushed, grabbed, or shoved; were threatened with being hit; or had something thrown at them, noted Dr. O'Campo. And they most commonly reported that just one incident of abuse had occurred, and that the abuse had taken place only before pregnancy. The patterns of abuse type, timing, and perpetrator were generally the same among low-income and nonmarried subgroups. But those two subgroups differed with respect to the frequency of abuse, more commonly reporting two to five incidents or six or more.

“The 11%, I think, is significant,” she said, adding that previous population-based studies have found prevalences of less than 5% for 1-year periods. “But I think it could be higher, actually, if we had had a full spectrum of abuse items that were asked about,” such as control and restriction, she added. The diverse nature of perpetrators suggests that future research should not focus solely on partners, according to Dr. O'Campo.

“Contrary to common perception,” abuse is not necessarily a high risk around the time of pregnancy; “in fact, it's lower than, say, the abuse that was experienced before pregnancy,” she observed. “It would be important for future studies to be able to ask questions so that we know more about why abuse might decline during pregnancy,” she said. For example, “is it because the pregnancy is disclosed and known, and that tends to be protective, or that partners change?”

Predictors of Postpartum Depression

“Exposure to family violence is increasingly understood to be an important risk factor for adverse pregnancy outcomes,” said coinvestigator Patricia A. Janssen, Ph.D. “But its role in the development of postpartum depression has not been well studied yet.”

Certain psychosocial factors have been linked to the risk of family violence and postpartum depression, questioning the relationship between the two, she added.

Overall, 8% of the women in the survey screened positive for postpartum depression (defined as an EPDS score of at least 13 of 30), reported Dr. Janssen, an epidemiologist at the University of British Columbia in Vancouver. The rate was nearly threefold higher among abused women (17%) than among their nonabused counterparts (6%).

Within the abused group, the risk of postpartum depression was more than doubled for women who were younger than age 20 (odds ratio, 2.29) or aged 35 years or older (OR, 2.33) relative to their peers aged 20-34 years. In addition, the risk was more than twice as high for abused women who had depression before pregnancy (OR, 2.23).

Marital status, education, employment, ethnicity and immigration status, income, and whether pregnancy was planned did not significantly influence this outcome among abused women.

In adjusted analyses that focused on the timing of abuse, women who were abused only before pregnancy (OR, 3.28), starting after pregnancy (OR, 4.76), or before, during, and after pregnancy (OR, 6.62) were all significantly more likely to have postpartum depression than were their nonabused peers. The results were generally the same when women with preexisting depression were excluded.

When analyses were restricted to just the threatened types of abuse, risk was significantly elevated among women whose abuse occurred only before pregnancy (OR, 2.29), began during pregnancy and continued afterward (OR, 2.62), started after pregnancy (OR, 5.28), or occurred before, during, and after pregnancy (OR, 10.52). Again, the results were generally the same when women with preexisting depression were excluded.

“Women who are abused women have higher rates of postpartum depression,” Dr. Janssen said. “I hope this will encourage people to screen more for abuse, if they recognize that it's a risk factor for depression.”

Those at the extremes of childbearing age or with preexisting depression appear to be especially vulnerable. Thus, “if we do know that women are abused, we should be paying particular attention to the risk for postpartum depression among those women.”

Major Finding: Some 11% of women reported abuse in the period surrounding pregnancy. Abused women who were younger than age 20 or were aged 35 or older, or who were depressed before pregnancy, were more than twice as likely to have postpartum depression.

Data Source: The Maternity Experiences Survey, a population-based survey of a weighted sample of 76,508 women who were an average of 7 months post partum.

Disclosures: Dr. O'Campo and Dr. Janssen reported having no relevant conflicts of interest.

SEATTLE — Abuse is prevalent in the period surrounding pregnancy and is associated with a higher rate of postpartum depression, new data show. But certain factors might help identify abused women who are at highest risk and might benefit from targeted intervention.

In a national survey of Canadian women who had had a recent birth, 11% reported abuse in the past 2 years, according to study results presented at the meeting. Abused women were roughly three times more likely than their nonabused counterparts to have postpartum depression. Within the abused group, this risk was more than doubled for teenagers and women aged at least 35 years, as well as for women who were depressed before pregnancy.

Prevalence of Abuse

Studies in recent decades have explored the issue of violence around the time of pregnancy, noted coinvestigator Patricia J. O'Campo, Ph.D. “Yet, despite that, we actually don't have good estimates of prevalence,” she said. Reasons include variation in the types of abuse captured, the time period and perpetrators assessed, and the women studied.

In the national, population-based Maternity Experiences Survey, Dr. O'Campo and her colleagues used census data to identify Canadian women with a singleton infant aged 5-14 months. A random sample was selected for a computer-assisted telephone interview that asked whether they had experienced any of 10 types of abuse (actual or threatened) in the past 2 years, as well as questions from the EPDS (Edinburgh Postnatal Depression Scale).

Interviews were conducted with 6,421 women, who represented a weighted sample of 76,508 women, according to Dr. O'Campo, an epidemiologist at the University of Toronto. Most were 5-9 months post partum (average, 7.3 months).

Fully 11% of the women reported experiencing abuse in the past 2 years. In stratified analyses, the prevalence was highest among teenagers (40%); women with an annual income less than $20,000 (28%); aboriginal women (30%); and nonmarried, noncohabiting women (35%). By far, the leading perpetrators were partners (reported by 6% overall), followed by family members (2%), strangers/others (2%), and friends (1%).

Abused women most often said they had been pushed, grabbed, or shoved; were threatened with being hit; or had something thrown at them, noted Dr. O'Campo. And they most commonly reported that just one incident of abuse had occurred, and that the abuse had taken place only before pregnancy. The patterns of abuse type, timing, and perpetrator were generally the same among low-income and nonmarried subgroups. But those two subgroups differed with respect to the frequency of abuse, more commonly reporting two to five incidents or six or more.

“The 11%, I think, is significant,” she said, adding that previous population-based studies have found prevalences of less than 5% for 1-year periods. “But I think it could be higher, actually, if we had had a full spectrum of abuse items that were asked about,” such as control and restriction, she added. The diverse nature of perpetrators suggests that future research should not focus solely on partners, according to Dr. O'Campo.

“Contrary to common perception,” abuse is not necessarily a high risk around the time of pregnancy; “in fact, it's lower than, say, the abuse that was experienced before pregnancy,” she observed. “It would be important for future studies to be able to ask questions so that we know more about why abuse might decline during pregnancy,” she said. For example, “is it because the pregnancy is disclosed and known, and that tends to be protective, or that partners change?”

Predictors of Postpartum Depression

“Exposure to family violence is increasingly understood to be an important risk factor for adverse pregnancy outcomes,” said coinvestigator Patricia A. Janssen, Ph.D. “But its role in the development of postpartum depression has not been well studied yet.”

Certain psychosocial factors have been linked to the risk of family violence and postpartum depression, questioning the relationship between the two, she added.

Overall, 8% of the women in the survey screened positive for postpartum depression (defined as an EPDS score of at least 13 of 30), reported Dr. Janssen, an epidemiologist at the University of British Columbia in Vancouver. The rate was nearly threefold higher among abused women (17%) than among their nonabused counterparts (6%).

Within the abused group, the risk of postpartum depression was more than doubled for women who were younger than age 20 (odds ratio, 2.29) or aged 35 years or older (OR, 2.33) relative to their peers aged 20-34 years. In addition, the risk was more than twice as high for abused women who had depression before pregnancy (OR, 2.23).

Marital status, education, employment, ethnicity and immigration status, income, and whether pregnancy was planned did not significantly influence this outcome among abused women.

In adjusted analyses that focused on the timing of abuse, women who were abused only before pregnancy (OR, 3.28), starting after pregnancy (OR, 4.76), or before, during, and after pregnancy (OR, 6.62) were all significantly more likely to have postpartum depression than were their nonabused peers. The results were generally the same when women with preexisting depression were excluded.

When analyses were restricted to just the threatened types of abuse, risk was significantly elevated among women whose abuse occurred only before pregnancy (OR, 2.29), began during pregnancy and continued afterward (OR, 2.62), started after pregnancy (OR, 5.28), or occurred before, during, and after pregnancy (OR, 10.52). Again, the results were generally the same when women with preexisting depression were excluded.

“Women who are abused women have higher rates of postpartum depression,” Dr. Janssen said. “I hope this will encourage people to screen more for abuse, if they recognize that it's a risk factor for depression.”

Those at the extremes of childbearing age or with preexisting depression appear to be especially vulnerable. Thus, “if we do know that women are abused, we should be paying particular attention to the risk for postpartum depression among those women.”

Major Finding: Some 11% of women reported abuse in the period surrounding pregnancy. Abused women who were younger than age 20 or were aged 35 or older, or who were depressed before pregnancy, were more than twice as likely to have postpartum depression.

Data Source: The Maternity Experiences Survey, a population-based survey of a weighted sample of 76,508 women who were an average of 7 months post partum.

Disclosures: Dr. O'Campo and Dr. Janssen reported having no relevant conflicts of interest.

SEATTLE — Abuse is prevalent in the period surrounding pregnancy and is associated with a higher rate of postpartum depression, new data show. But certain factors might help identify abused women who are at highest risk and might benefit from targeted intervention.

In a national survey of Canadian women who had had a recent birth, 11% reported abuse in the past 2 years, according to study results presented at the meeting. Abused women were roughly three times more likely than their nonabused counterparts to have postpartum depression. Within the abused group, this risk was more than doubled for teenagers and women aged at least 35 years, as well as for women who were depressed before pregnancy.

Prevalence of Abuse

Studies in recent decades have explored the issue of violence around the time of pregnancy, noted coinvestigator Patricia J. O'Campo, Ph.D. “Yet, despite that, we actually don't have good estimates of prevalence,” she said. Reasons include variation in the types of abuse captured, the time period and perpetrators assessed, and the women studied.

In the national, population-based Maternity Experiences Survey, Dr. O'Campo and her colleagues used census data to identify Canadian women with a singleton infant aged 5-14 months. A random sample was selected for a computer-assisted telephone interview that asked whether they had experienced any of 10 types of abuse (actual or threatened) in the past 2 years, as well as questions from the EPDS (Edinburgh Postnatal Depression Scale).

Interviews were conducted with 6,421 women, who represented a weighted sample of 76,508 women, according to Dr. O'Campo, an epidemiologist at the University of Toronto. Most were 5-9 months post partum (average, 7.3 months).

Fully 11% of the women reported experiencing abuse in the past 2 years. In stratified analyses, the prevalence was highest among teenagers (40%); women with an annual income less than $20,000 (28%); aboriginal women (30%); and nonmarried, noncohabiting women (35%). By far, the leading perpetrators were partners (reported by 6% overall), followed by family members (2%), strangers/others (2%), and friends (1%).

Abused women most often said they had been pushed, grabbed, or shoved; were threatened with being hit; or had something thrown at them, noted Dr. O'Campo. And they most commonly reported that just one incident of abuse had occurred, and that the abuse had taken place only before pregnancy. The patterns of abuse type, timing, and perpetrator were generally the same among low-income and nonmarried subgroups. But those two subgroups differed with respect to the frequency of abuse, more commonly reporting two to five incidents or six or more.

“The 11%, I think, is significant,” she said, adding that previous population-based studies have found prevalences of less than 5% for 1-year periods. “But I think it could be higher, actually, if we had had a full spectrum of abuse items that were asked about,” such as control and restriction, she added. The diverse nature of perpetrators suggests that future research should not focus solely on partners, according to Dr. O'Campo.

“Contrary to common perception,” abuse is not necessarily a high risk around the time of pregnancy; “in fact, it's lower than, say, the abuse that was experienced before pregnancy,” she observed. “It would be important for future studies to be able to ask questions so that we know more about why abuse might decline during pregnancy,” she said. For example, “is it because the pregnancy is disclosed and known, and that tends to be protective, or that partners change?”

Predictors of Postpartum Depression

“Exposure to family violence is increasingly understood to be an important risk factor for adverse pregnancy outcomes,” said coinvestigator Patricia A. Janssen, Ph.D. “But its role in the development of postpartum depression has not been well studied yet.”

Certain psychosocial factors have been linked to the risk of family violence and postpartum depression, questioning the relationship between the two, she added.

Overall, 8% of the women in the survey screened positive for postpartum depression (defined as an EPDS score of at least 13 of 30), reported Dr. Janssen, an epidemiologist at the University of British Columbia in Vancouver. The rate was nearly threefold higher among abused women (17%) than among their nonabused counterparts (6%).

Within the abused group, the risk of postpartum depression was more than doubled for women who were younger than age 20 (odds ratio, 2.29) or aged 35 years or older (OR, 2.33) relative to their peers aged 20-34 years. In addition, the risk was more than twice as high for abused women who had depression before pregnancy (OR, 2.23).

Marital status, education, employment, ethnicity and immigration status, income, and whether pregnancy was planned did not significantly influence this outcome among abused women.

In adjusted analyses that focused on the timing of abuse, women who were abused only before pregnancy (OR, 3.28), starting after pregnancy (OR, 4.76), or before, during, and after pregnancy (OR, 6.62) were all significantly more likely to have postpartum depression than were their nonabused peers. The results were generally the same when women with preexisting depression were excluded.

When analyses were restricted to just the threatened types of abuse, risk was significantly elevated among women whose abuse occurred only before pregnancy (OR, 2.29), began during pregnancy and continued afterward (OR, 2.62), started after pregnancy (OR, 5.28), or occurred before, during, and after pregnancy (OR, 10.52). Again, the results were generally the same when women with preexisting depression were excluded.

“Women who are abused women have higher rates of postpartum depression,” Dr. Janssen said. “I hope this will encourage people to screen more for abuse, if they recognize that it's a risk factor for depression.”

Those at the extremes of childbearing age or with preexisting depression appear to be especially vulnerable. Thus, “if we do know that women are abused, we should be paying particular attention to the risk for postpartum depression among those women.”

Major Finding: Adolescent girls whose physicians had recommended they get the HPV vaccine were more likely to do so than were girls whose physicians had not made such a recommendation (49% vs. 4%).

Data Source: A cross-sectional study among 36,284 adolescents aged 12–17 years from the National Survey of Children's Health.

Disclosures: None was reported.

VANCOUVER, B.C. — Your recommendation may be the deciding factor in whether adolescents get vaccinated, and this appears to be especially true in adolescent girls, according to Dr. Paul M. Darden.

“The girls who got recommended HPV [human papillomavirus vaccine] had much higher rates of every vaccine than those who did not and in fact did much better than boys just in general,” Dr. Darden said at the meeting.

This was the finding in a nationwide, cross-sectional study that assessed predictors of immunization among 36,284 adolescents 12–17 years old. The 2007 National Survey of Children's Health used random-digit dialing to contact and poll parents by telephone about their children's immunization status, especially the tetanus-diphtheria (Td) vaccine, the tetanus-diphtheria-pertussis (Tdap) vaccine, the meningococcal conjugate vaccine (MCV4), and HPV vaccine.

Specifically, relative to girls whose physicians had not recommended the HPV vaccine, those whose physicians had were more likely to have received that vaccine (49% vs 4%, P less than .01), but also more likely to have received the Td or Tdap vaccine (91% vs 77%, P less than .01) and the MCV4 vaccine (52% vs 30%, P less than .01).

“I think this just points out that a physician's or health care provider's recommendation has a huge influence on who gets a vaccine,” he commented. “They usually come to you because they trust you, and if you recommend something, they are much more likely to do it.”

An analysis of the independent predictors of up-to-date status showed the predictors “were actually the opposite of what we typically see for childhood immunizations,” said Dr. Darden, professor of pediatrics at the University of Oklahoma in Oklahoma City.

The proportion of adolescents current on both Td/Tdap and MCV4 was highest among those with Medicaid insurance, at 38%, vs. 33% among those with no insurance and 31% among those with private insurance (P less than .01).

In addition, the proportion up to date on both vaccines was highest among adolescents whose mothers had not completed high school, at 39%, vs. 33% among those whose mothers had schooling beyond this level and 31% among those whose mothers had stopped at high school (P = .02).

Speculating on the reasons for these findings, Dr. Darden noted that the costs of vaccines are covered for adolescents with Medicaid, whereas adolescents with other types of insurance or none may face substantial out-of-pocket costs.

As for the influence of maternal education, higher education has been linked with greater reluctance to vaccinate. Dr. Darden proposed that there are likely additional factors at play. “A 6–percentage point difference seems too big for just vaccine reluctance.”

Major Finding: Adolescent girls whose physicians had recommended they get the HPV vaccine were more likely to do so than were girls whose physicians had not made such a recommendation (49% vs. 4%).

Data Source: A cross-sectional study among 36,284 adolescents aged 12–17 years from the National Survey of Children's Health.

Disclosures: None was reported.

VANCOUVER, B.C. — Your recommendation may be the deciding factor in whether adolescents get vaccinated, and this appears to be especially true in adolescent girls, according to Dr. Paul M. Darden.

“The girls who got recommended HPV [human papillomavirus vaccine] had much higher rates of every vaccine than those who did not and in fact did much better than boys just in general,” Dr. Darden said at the meeting.

This was the finding in a nationwide, cross-sectional study that assessed predictors of immunization among 36,284 adolescents 12–17 years old. The 2007 National Survey of Children's Health used random-digit dialing to contact and poll parents by telephone about their children's immunization status, especially the tetanus-diphtheria (Td) vaccine, the tetanus-diphtheria-pertussis (Tdap) vaccine, the meningococcal conjugate vaccine (MCV4), and HPV vaccine.

Specifically, relative to girls whose physicians had not recommended the HPV vaccine, those whose physicians had were more likely to have received that vaccine (49% vs 4%, P less than .01), but also more likely to have received the Td or Tdap vaccine (91% vs 77%, P less than .01) and the MCV4 vaccine (52% vs 30%, P less than .01).

“I think this just points out that a physician's or health care provider's recommendation has a huge influence on who gets a vaccine,” he commented. “They usually come to you because they trust you, and if you recommend something, they are much more likely to do it.”

An analysis of the independent predictors of up-to-date status showed the predictors “were actually the opposite of what we typically see for childhood immunizations,” said Dr. Darden, professor of pediatrics at the University of Oklahoma in Oklahoma City.

The proportion of adolescents current on both Td/Tdap and MCV4 was highest among those with Medicaid insurance, at 38%, vs. 33% among those with no insurance and 31% among those with private insurance (P less than .01).

In addition, the proportion up to date on both vaccines was highest among adolescents whose mothers had not completed high school, at 39%, vs. 33% among those whose mothers had schooling beyond this level and 31% among those whose mothers had stopped at high school (P = .02).

Speculating on the reasons for these findings, Dr. Darden noted that the costs of vaccines are covered for adolescents with Medicaid, whereas adolescents with other types of insurance or none may face substantial out-of-pocket costs.

As for the influence of maternal education, higher education has been linked with greater reluctance to vaccinate. Dr. Darden proposed that there are likely additional factors at play. “A 6–percentage point difference seems too big for just vaccine reluctance.”

Major Finding: Adolescent girls whose physicians had recommended they get the HPV vaccine were more likely to do so than were girls whose physicians had not made such a recommendation (49% vs. 4%).

Data Source: A cross-sectional study among 36,284 adolescents aged 12–17 years from the National Survey of Children's Health.

Disclosures: None was reported.

VANCOUVER, B.C. — Your recommendation may be the deciding factor in whether adolescents get vaccinated, and this appears to be especially true in adolescent girls, according to Dr. Paul M. Darden.

“The girls who got recommended HPV [human papillomavirus vaccine] had much higher rates of every vaccine than those who did not and in fact did much better than boys just in general,” Dr. Darden said at the meeting.

This was the finding in a nationwide, cross-sectional study that assessed predictors of immunization among 36,284 adolescents 12–17 years old. The 2007 National Survey of Children's Health used random-digit dialing to contact and poll parents by telephone about their children's immunization status, especially the tetanus-diphtheria (Td) vaccine, the tetanus-diphtheria-pertussis (Tdap) vaccine, the meningococcal conjugate vaccine (MCV4), and HPV vaccine.

Specifically, relative to girls whose physicians had not recommended the HPV vaccine, those whose physicians had were more likely to have received that vaccine (49% vs 4%, P less than .01), but also more likely to have received the Td or Tdap vaccine (91% vs 77%, P less than .01) and the MCV4 vaccine (52% vs 30%, P less than .01).

“I think this just points out that a physician's or health care provider's recommendation has a huge influence on who gets a vaccine,” he commented. “They usually come to you because they trust you, and if you recommend something, they are much more likely to do it.”

An analysis of the independent predictors of up-to-date status showed the predictors “were actually the opposite of what we typically see for childhood immunizations,” said Dr. Darden, professor of pediatrics at the University of Oklahoma in Oklahoma City.

The proportion of adolescents current on both Td/Tdap and MCV4 was highest among those with Medicaid insurance, at 38%, vs. 33% among those with no insurance and 31% among those with private insurance (P less than .01).

In addition, the proportion up to date on both vaccines was highest among adolescents whose mothers had not completed high school, at 39%, vs. 33% among those whose mothers had schooling beyond this level and 31% among those whose mothers had stopped at high school (P = .02).

Speculating on the reasons for these findings, Dr. Darden noted that the costs of vaccines are covered for adolescents with Medicaid, whereas adolescents with other types of insurance or none may face substantial out-of-pocket costs.

As for the influence of maternal education, higher education has been linked with greater reluctance to vaccinate. Dr. Darden proposed that there are likely additional factors at play. “A 6–percentage point difference seems too big for just vaccine reluctance.”

Major Finding: Adolescents whose parents were sent personalized text messages reminding them that their children needed vaccines were 2.5 times more likely to receive the MCV4 vaccine, the Tdap vaccine, or both.

Data Source: A randomized trial among 361 low-income urban adolescents and their parents.

Disclosures: Dr. Stockwell reported that she had no conflicts, but that one of the study's investigators is on an advisory board for, and has received research funding from, Merck.

VANCOUVER, B.C. — Text messages reminding parents that their adolescents need vaccines are an effective and efficient means of improving immunization rates in this population, according to findings of a randomized trial conducted in New York City.

Adolescents whose parents were sent automated, personalized text message reminders that included walk-in times for vaccination were 2.5 times more likely to receive the meningococcal conjugate (MCV4) vaccine, the tetanus-diphtheria-pertussis (Tdap) vaccine, or both.

They were also more likely to receive other vaccines for which they were not up to date.

The observed gains in immunization rates are greater than are those seen previously with reminder and recall strategies, lead investigator Dr. Melissa S. Stockwell said.

In fact, this strategy works so well that busy clinics with many competing priorities will need to be prepared to meet the increased demand for vaccines.

“The actual texting is instantaneous and fast, but you have the repercussions of bringing all those kids in,” she explained. One option is to stagger the messages to control the numbers of adolescents who might show up at one time.

The trial was conducted in six networked primary care practices in New York City among parents who had a child aged 11–18 years needing the MCV4 vaccine, the Tdap vaccine, or both, and had a cell phone number in the system. The practices were randomized to a control (usual-care) group or an intervention (Text4Health) group.

In intervention practices, parents were sent text message reminders that were automatically generated using data from a linked immunization registry. The reminders were sent up to five times over a 7-week period until the adolescent had received the two target vaccines.

“Based on our focus group [of parents], they really wanted the messages to be personal,” commented Dr. Stockwell, a pediatrician at Columbia University in New York. So the messages included their child's name, their clinic's name, and walk-in times for vaccination, plus offered the option of switching between English and Spanish.

Study results reported in a poster were based on 195 adolescents in the intervention group and 166 in the control group. The adolescents were 16 years old on average, 55% were Hispanic, and 80% had public health insurance.

In the intervention group, 821 text messages were sent, only 7 of which bounced. Merely 6% of parents in this group were found to have wrong numbers, and just 3% opted out of receiving the messages.

At 24 weeks, adolescents in the intervention group had a higher rate of receipt of the MCV4 vaccine, Tdap vaccine, or both, compared with their peers in the control group (35% vs. 17%)—for a difference between groups of 18% (P less than .001). After adjustment for potential confounders, the intervention was associated with a 2.5-fold higher odds of receiving these vaccines.

Adolescents in the intervention group also had a higher rate of receipt of any other needed vaccines, such as the human papillomavirus vaccine and the hepatitis A vaccine (42% vs. 29.5%), with a difference between groups of 12.5% (P less than .05).

The estimated cost of the intervention, assuming that it were sustained for 2 years in a hypothetical cohort of 100,000 adolescents, was $1.71 per adolescent immunized and $0.42 per additional vaccine delivered.

There are several possible reasons why a text message might work when a reminder phone call does not, according to Dr. Stockwell.

“One, it always reaches the right recipient, the intended person,” she said, whereas a phone call might be answered by someone else or go to voicemail that is picked up by someone else. “Also, it stays on the phone, so it might sort of be a constant reminder, or the information—especially the walk-in times—is right there so they can send their child at that time.”

Finally, focus groups suggest that parents in their study population perceive information sent by text messages to be more important than that conveyed by more commonplace phone calls, although she said that might change as the novelty of texting wears off.

Major Finding: Adolescents whose parents were sent personalized text messages reminding them that their children needed vaccines were 2.5 times more likely to receive the MCV4 vaccine, the Tdap vaccine, or both.

Data Source: A randomized trial among 361 low-income urban adolescents and their parents.

Disclosures: Dr. Stockwell reported that she had no conflicts, but that one of the study's investigators is on an advisory board for, and has received research funding from, Merck.

VANCOUVER, B.C. — Text messages reminding parents that their adolescents need vaccines are an effective and efficient means of improving immunization rates in this population, according to findings of a randomized trial conducted in New York City.

Adolescents whose parents were sent automated, personalized text message reminders that included walk-in times for vaccination were 2.5 times more likely to receive the meningococcal conjugate (MCV4) vaccine, the tetanus-diphtheria-pertussis (Tdap) vaccine, or both.

They were also more likely to receive other vaccines for which they were not up to date.

The observed gains in immunization rates are greater than are those seen previously with reminder and recall strategies, lead investigator Dr. Melissa S. Stockwell said.

In fact, this strategy works so well that busy clinics with many competing priorities will need to be prepared to meet the increased demand for vaccines.

“The actual texting is instantaneous and fast, but you have the repercussions of bringing all those kids in,” she explained. One option is to stagger the messages to control the numbers of adolescents who might show up at one time.

The trial was conducted in six networked primary care practices in New York City among parents who had a child aged 11–18 years needing the MCV4 vaccine, the Tdap vaccine, or both, and had a cell phone number in the system. The practices were randomized to a control (usual-care) group or an intervention (Text4Health) group.

In intervention practices, parents were sent text message reminders that were automatically generated using data from a linked immunization registry. The reminders were sent up to five times over a 7-week period until the adolescent had received the two target vaccines.

“Based on our focus group [of parents], they really wanted the messages to be personal,” commented Dr. Stockwell, a pediatrician at Columbia University in New York. So the messages included their child's name, their clinic's name, and walk-in times for vaccination, plus offered the option of switching between English and Spanish.

Study results reported in a poster were based on 195 adolescents in the intervention group and 166 in the control group. The adolescents were 16 years old on average, 55% were Hispanic, and 80% had public health insurance.

In the intervention group, 821 text messages were sent, only 7 of which bounced. Merely 6% of parents in this group were found to have wrong numbers, and just 3% opted out of receiving the messages.

At 24 weeks, adolescents in the intervention group had a higher rate of receipt of the MCV4 vaccine, Tdap vaccine, or both, compared with their peers in the control group (35% vs. 17%)—for a difference between groups of 18% (P less than .001). After adjustment for potential confounders, the intervention was associated with a 2.5-fold higher odds of receiving these vaccines.

Adolescents in the intervention group also had a higher rate of receipt of any other needed vaccines, such as the human papillomavirus vaccine and the hepatitis A vaccine (42% vs. 29.5%), with a difference between groups of 12.5% (P less than .05).

The estimated cost of the intervention, assuming that it were sustained for 2 years in a hypothetical cohort of 100,000 adolescents, was $1.71 per adolescent immunized and $0.42 per additional vaccine delivered.

There are several possible reasons why a text message might work when a reminder phone call does not, according to Dr. Stockwell.

“One, it always reaches the right recipient, the intended person,” she said, whereas a phone call might be answered by someone else or go to voicemail that is picked up by someone else. “Also, it stays on the phone, so it might sort of be a constant reminder, or the information—especially the walk-in times—is right there so they can send their child at that time.”

Finally, focus groups suggest that parents in their study population perceive information sent by text messages to be more important than that conveyed by more commonplace phone calls, although she said that might change as the novelty of texting wears off.

Major Finding: Adolescents whose parents were sent personalized text messages reminding them that their children needed vaccines were 2.5 times more likely to receive the MCV4 vaccine, the Tdap vaccine, or both.

Data Source: A randomized trial among 361 low-income urban adolescents and their parents.

Disclosures: Dr. Stockwell reported that she had no conflicts, but that one of the study's investigators is on an advisory board for, and has received research funding from, Merck.

VANCOUVER, B.C. — Text messages reminding parents that their adolescents need vaccines are an effective and efficient means of improving immunization rates in this population, according to findings of a randomized trial conducted in New York City.

Adolescents whose parents were sent automated, personalized text message reminders that included walk-in times for vaccination were 2.5 times more likely to receive the meningococcal conjugate (MCV4) vaccine, the tetanus-diphtheria-pertussis (Tdap) vaccine, or both.

They were also more likely to receive other vaccines for which they were not up to date.

The observed gains in immunization rates are greater than are those seen previously with reminder and recall strategies, lead investigator Dr. Melissa S. Stockwell said.

In fact, this strategy works so well that busy clinics with many competing priorities will need to be prepared to meet the increased demand for vaccines.

“The actual texting is instantaneous and fast, but you have the repercussions of bringing all those kids in,” she explained. One option is to stagger the messages to control the numbers of adolescents who might show up at one time.

The trial was conducted in six networked primary care practices in New York City among parents who had a child aged 11–18 years needing the MCV4 vaccine, the Tdap vaccine, or both, and had a cell phone number in the system. The practices were randomized to a control (usual-care) group or an intervention (Text4Health) group.

In intervention practices, parents were sent text message reminders that were automatically generated using data from a linked immunization registry. The reminders were sent up to five times over a 7-week period until the adolescent had received the two target vaccines.

“Based on our focus group [of parents], they really wanted the messages to be personal,” commented Dr. Stockwell, a pediatrician at Columbia University in New York. So the messages included their child's name, their clinic's name, and walk-in times for vaccination, plus offered the option of switching between English and Spanish.

Study results reported in a poster were based on 195 adolescents in the intervention group and 166 in the control group. The adolescents were 16 years old on average, 55% were Hispanic, and 80% had public health insurance.

In the intervention group, 821 text messages were sent, only 7 of which bounced. Merely 6% of parents in this group were found to have wrong numbers, and just 3% opted out of receiving the messages.

At 24 weeks, adolescents in the intervention group had a higher rate of receipt of the MCV4 vaccine, Tdap vaccine, or both, compared with their peers in the control group (35% vs. 17%)—for a difference between groups of 18% (P less than .001). After adjustment for potential confounders, the intervention was associated with a 2.5-fold higher odds of receiving these vaccines.

Adolescents in the intervention group also had a higher rate of receipt of any other needed vaccines, such as the human papillomavirus vaccine and the hepatitis A vaccine (42% vs. 29.5%), with a difference between groups of 12.5% (P less than .05).

The estimated cost of the intervention, assuming that it were sustained for 2 years in a hypothetical cohort of 100,000 adolescents, was $1.71 per adolescent immunized and $0.42 per additional vaccine delivered.

There are several possible reasons why a text message might work when a reminder phone call does not, according to Dr. Stockwell.

“One, it always reaches the right recipient, the intended person,” she said, whereas a phone call might be answered by someone else or go to voicemail that is picked up by someone else. “Also, it stays on the phone, so it might sort of be a constant reminder, or the information—especially the walk-in times—is right there so they can send their child at that time.”

Finally, focus groups suggest that parents in their study population perceive information sent by text messages to be more important than that conveyed by more commonplace phone calls, although she said that might change as the novelty of texting wears off.

VANCOUVER, B.C. – Information about a patient’s risk of developing statin-induced myotoxicity may come from a surprising source: the composition of his or her high-density lipoprotein particles, a study has shown.

In the observational study of 670 patients, those having higher levels of HDL particles rich in triglycerides before starting statins were 3-21 times more likely to develop various types of myotoxicity while on these medications.

“These results suggest an association between HDL-triglyceride concentration at baseline and markers of statin intolerance,” lead investigator Dr. Tarek Bouhali said at the annual World Congress on Heart Disease sponsored by the International Academy of Cardiology.

“Right now, we can’t tell what the mechanism is,” he commented. “But ... HDL is a heterogeneous group of lipoproteins and is not only cholesterol,” and in this setting, it appears that heterogeneity may be useful for risk stratification.

Skeletal muscle adverse events are the leading complication of statin therapy, according to Dr. Bouhali. “Muscular side effects are usually mild and reversible; however, knowledge of the underlying mechanism and risk factors is required for prompt identification and proper management for these side effects.”

Furthermore, he noted, understanding and preventing statin adverse effects are important because side effects can lead to both reduced compliance and increased health care costs.

The investigators collected fasting plasma samples from adult French Canadian patients who were about to start lipid-lowering therapy and were receiving care according to American Heart Association guidelines. The samples were ultracentrifuged to determine levels of cholesterol and triglyceride associated with HDL particles.

The study patients were 50 years old on average, and 63% were male. About 12% were taking statins in combination with other lipid-lowering medications (fibrates, bile acid sequestrants, nicotinic acid, or inhibitors of cholesterol absorption).

Overall, 22% of the patients had at least one adverse effect of any type (muscular or nonmuscular), reported Dr. Bouhali of the University of Montreal and the lead investigator with the ECOGENE-21 project.

In terms of the muscular adverse effects, 20% of patients developed myalgia, 7% had a creatine kinase level of 300 U/L or higher, 3% had trace proteinuria or myoglobinuria, and 2% had a creatine kinase level of 500 U/L or higher.

After adjustment for age, sex, statin dose, and concomitant medications, a higher plasma HDL-triglyceride level before starting treatment was associated with a greater risk of myotoxicity, and especially the more severe forms, noted Dr. Bouhali.

Specifically, the log10-transformed HDL-triglyceride level in millimoles per liter was significantly associated with the development of myalgia (odds ratio, 2.9); a creatine kinase level of 300 U/L or higher (OR, 6.0); proteinuria or myoglobinuria (OR, 12.6); and a creatine kinase level of 500 U/L or higher (OR, 20.9).

In addition, a higher HDL-triglyceride level at baseline was associated with a borderline increase in the risk of any (muscular or nonmuscular) adverse effect (OR, 2.5; P = .065).

In contrast, the level of HDL cholesterol was not significantly associated with the likelihood of statin adverse effects.

The study’s findings should be validated and replicated in other populations, said Dr. Bouhali.

In addition, “further mechanistic studies are needed to explore the implication of HDL particle composition in the development of statin intolerance,” he said in comments provided by e-mail.

Dr. Bouhali reported that he had no relevant conflicts of interest.

VANCOUVER, B.C. – Information about a patient’s risk of developing statin-induced myotoxicity may come from a surprising source: the composition of his or her high-density lipoprotein particles, a study has shown.

In the observational study of 670 patients, those having higher levels of HDL particles rich in triglycerides before starting statins were 3-21 times more likely to develop various types of myotoxicity while on these medications.

“These results suggest an association between HDL-triglyceride concentration at baseline and markers of statin intolerance,” lead investigator Dr. Tarek Bouhali said at the annual World Congress on Heart Disease sponsored by the International Academy of Cardiology.

“Right now, we can’t tell what the mechanism is,” he commented. “But ... HDL is a heterogeneous group of lipoproteins and is not only cholesterol,” and in this setting, it appears that heterogeneity may be useful for risk stratification.

Skeletal muscle adverse events are the leading complication of statin therapy, according to Dr. Bouhali. “Muscular side effects are usually mild and reversible; however, knowledge of the underlying mechanism and risk factors is required for prompt identification and proper management for these side effects.”

Furthermore, he noted, understanding and preventing statin adverse effects are important because side effects can lead to both reduced compliance and increased health care costs.

The investigators collected fasting plasma samples from adult French Canadian patients who were about to start lipid-lowering therapy and were receiving care according to American Heart Association guidelines. The samples were ultracentrifuged to determine levels of cholesterol and triglyceride associated with HDL particles.

The study patients were 50 years old on average, and 63% were male. About 12% were taking statins in combination with other lipid-lowering medications (fibrates, bile acid sequestrants, nicotinic acid, or inhibitors of cholesterol absorption).

Overall, 22% of the patients had at least one adverse effect of any type (muscular or nonmuscular), reported Dr. Bouhali of the University of Montreal and the lead investigator with the ECOGENE-21 project.

In terms of the muscular adverse effects, 20% of patients developed myalgia, 7% had a creatine kinase level of 300 U/L or higher, 3% had trace proteinuria or myoglobinuria, and 2% had a creatine kinase level of 500 U/L or higher.

After adjustment for age, sex, statin dose, and concomitant medications, a higher plasma HDL-triglyceride level before starting treatment was associated with a greater risk of myotoxicity, and especially the more severe forms, noted Dr. Bouhali.

Specifically, the log10-transformed HDL-triglyceride level in millimoles per liter was significantly associated with the development of myalgia (odds ratio, 2.9); a creatine kinase level of 300 U/L or higher (OR, 6.0); proteinuria or myoglobinuria (OR, 12.6); and a creatine kinase level of 500 U/L or higher (OR, 20.9).

In addition, a higher HDL-triglyceride level at baseline was associated with a borderline increase in the risk of any (muscular or nonmuscular) adverse effect (OR, 2.5; P = .065).

In contrast, the level of HDL cholesterol was not significantly associated with the likelihood of statin adverse effects.

The study’s findings should be validated and replicated in other populations, said Dr. Bouhali.

In addition, “further mechanistic studies are needed to explore the implication of HDL particle composition in the development of statin intolerance,” he said in comments provided by e-mail.

Dr. Bouhali reported that he had no relevant conflicts of interest.

VANCOUVER, B.C. – Information about a patient’s risk of developing statin-induced myotoxicity may come from a surprising source: the composition of his or her high-density lipoprotein particles, a study has shown.

In the observational study of 670 patients, those having higher levels of HDL particles rich in triglycerides before starting statins were 3-21 times more likely to develop various types of myotoxicity while on these medications.

“These results suggest an association between HDL-triglyceride concentration at baseline and markers of statin intolerance,” lead investigator Dr. Tarek Bouhali said at the annual World Congress on Heart Disease sponsored by the International Academy of Cardiology.

“Right now, we can’t tell what the mechanism is,” he commented. “But ... HDL is a heterogeneous group of lipoproteins and is not only cholesterol,” and in this setting, it appears that heterogeneity may be useful for risk stratification.

Skeletal muscle adverse events are the leading complication of statin therapy, according to Dr. Bouhali. “Muscular side effects are usually mild and reversible; however, knowledge of the underlying mechanism and risk factors is required for prompt identification and proper management for these side effects.”

Furthermore, he noted, understanding and preventing statin adverse effects are important because side effects can lead to both reduced compliance and increased health care costs.

The investigators collected fasting plasma samples from adult French Canadian patients who were about to start lipid-lowering therapy and were receiving care according to American Heart Association guidelines. The samples were ultracentrifuged to determine levels of cholesterol and triglyceride associated with HDL particles.

The study patients were 50 years old on average, and 63% were male. About 12% were taking statins in combination with other lipid-lowering medications (fibrates, bile acid sequestrants, nicotinic acid, or inhibitors of cholesterol absorption).

Overall, 22% of the patients had at least one adverse effect of any type (muscular or nonmuscular), reported Dr. Bouhali of the University of Montreal and the lead investigator with the ECOGENE-21 project.

In terms of the muscular adverse effects, 20% of patients developed myalgia, 7% had a creatine kinase level of 300 U/L or higher, 3% had trace proteinuria or myoglobinuria, and 2% had a creatine kinase level of 500 U/L or higher.

After adjustment for age, sex, statin dose, and concomitant medications, a higher plasma HDL-triglyceride level before starting treatment was associated with a greater risk of myotoxicity, and especially the more severe forms, noted Dr. Bouhali.

Specifically, the log10-transformed HDL-triglyceride level in millimoles per liter was significantly associated with the development of myalgia (odds ratio, 2.9); a creatine kinase level of 300 U/L or higher (OR, 6.0); proteinuria or myoglobinuria (OR, 12.6); and a creatine kinase level of 500 U/L or higher (OR, 20.9).

In addition, a higher HDL-triglyceride level at baseline was associated with a borderline increase in the risk of any (muscular or nonmuscular) adverse effect (OR, 2.5; P = .065).

In contrast, the level of HDL cholesterol was not significantly associated with the likelihood of statin adverse effects.

The study’s findings should be validated and replicated in other populations, said Dr. Bouhali.

In addition, “further mechanistic studies are needed to explore the implication of HDL particle composition in the development of statin intolerance,” he said in comments provided by e-mail.

Dr. Bouhali reported that he had no relevant conflicts of interest.

VANCOUVER, B.C. – Information about a patient’s risk of developing statin-induced myotoxicity may come from a surprising source: the composition of his or her high-density lipoprotein particles, a study has shown.

In the observational study of 670 patients, those having higher levels of HDL particles rich in triglycerides before starting statins were 3-21 times more likely to develop various types of myotoxicity while on these medications.

“These results suggest an association between HDL-triglyceride concentration at baseline and markers of statin intolerance,” lead investigator Dr. Tarek Bouhali said at the annual World Congress on Heart Disease sponsored by the International Academy of Cardiology.

“Right now, we can’t tell what the mechanism is,” he commented. “But ... HDL is a heterogeneous group of lipoproteins and is not only cholesterol,” and in this setting, it appears that heterogeneity may be useful for risk stratification.

Skeletal muscle adverse events are the leading complication of statin therapy, according to Dr. Bouhali. “Muscular side effects are usually mild and reversible; however, knowledge of the underlying mechanism and risk factors is required for prompt identification and proper management for these side effects.”

Furthermore, he noted, understanding and preventing statin adverse effects are important because side effects can lead to both reduced compliance and increased health care costs.

The investigators collected fasting plasma samples from adult French Canadian patients who were about to start lipid-lowering therapy and were receiving care according to American Heart Association guidelines. The samples were ultracentrifuged to determine levels of cholesterol and triglyceride associated with HDL particles.

The study patients were 50 years old on average, and 63% were male. About 12% were taking statins in combination with other lipid-lowering medications (fibrates, bile acid sequestrants, nicotinic acid, or inhibitors of cholesterol absorption).

Overall, 22% of the patients had at least one adverse effect of any type (muscular or nonmuscular), reported Dr. Bouhali of the University of Montreal and the lead investigator with the ECOGENE-21 project.

In terms of the muscular adverse effects, 20% of patients developed myalgia, 7% had a creatine kinase level of 300 U/L or higher, 3% had trace proteinuria or myoglobinuria, and 2% had a creatine kinase level of 500 U/L or higher.

After adjustment for age, sex, statin dose, and concomitant medications, a higher plasma HDL-triglyceride level before starting treatment was associated with a greater risk of myotoxicity, and especially the more severe forms, noted Dr. Bouhali.

Specifically, the log10-transformed HDL-triglyceride level in millimoles per liter was significantly associated with the development of myalgia (odds ratio, 2.9); a creatine kinase level of 300 U/L or higher (OR, 6.0); proteinuria or myoglobinuria (OR, 12.6); and a creatine kinase level of 500 U/L or higher (OR, 20.9).

In addition, a higher HDL-triglyceride level at baseline was associated with a borderline increase in the risk of any (muscular or nonmuscular) adverse effect (OR, 2.5; P = .065).

In contrast, the level of HDL cholesterol was not significantly associated with the likelihood of statin adverse effects.

The study’s findings should be validated and replicated in other populations, said Dr. Bouhali.

In addition, “further mechanistic studies are needed to explore the implication of HDL particle composition in the development of statin intolerance,” he said in comments provided by e-mail.

Dr. Bouhali reported that he had no relevant conflicts of interest.

VANCOUVER, B.C. – Information about a patient’s risk of developing statin-induced myotoxicity may come from a surprising source: the composition of his or her high-density lipoprotein particles, a study has shown.

In the observational study of 670 patients, those having higher levels of HDL particles rich in triglycerides before starting statins were 3-21 times more likely to develop various types of myotoxicity while on these medications.

“These results suggest an association between HDL-triglyceride concentration at baseline and markers of statin intolerance,” lead investigator Dr. Tarek Bouhali said at the annual World Congress on Heart Disease sponsored by the International Academy of Cardiology.

“Right now, we can’t tell what the mechanism is,” he commented. “But ... HDL is a heterogeneous group of lipoproteins and is not only cholesterol,” and in this setting, it appears that heterogeneity may be useful for risk stratification.

Skeletal muscle adverse events are the leading complication of statin therapy, according to Dr. Bouhali. “Muscular side effects are usually mild and reversible; however, knowledge of the underlying mechanism and risk factors is required for prompt identification and proper management for these side effects.”

Furthermore, he noted, understanding and preventing statin adverse effects are important because side effects can lead to both reduced compliance and increased health care costs.

The investigators collected fasting plasma samples from adult French Canadian patients who were about to start lipid-lowering therapy and were receiving care according to American Heart Association guidelines. The samples were ultracentrifuged to determine levels of cholesterol and triglyceride associated with HDL particles.

The study patients were 50 years old on average, and 63% were male. About 12% were taking statins in combination with other lipid-lowering medications (fibrates, bile acid sequestrants, nicotinic acid, or inhibitors of cholesterol absorption).

Overall, 22% of the patients had at least one adverse effect of any type (muscular or nonmuscular), reported Dr. Bouhali of the University of Montreal and the lead investigator with the ECOGENE-21 project.

In terms of the muscular adverse effects, 20% of patients developed myalgia, 7% had a creatine kinase level of 300 U/L or higher, 3% had trace proteinuria or myoglobinuria, and 2% had a creatine kinase level of 500 U/L or higher.

After adjustment for age, sex, statin dose, and concomitant medications, a higher plasma HDL-triglyceride level before starting treatment was associated with a greater risk of myotoxicity, and especially the more severe forms, noted Dr. Bouhali.

Specifically, the log10-transformed HDL-triglyceride level in millimoles per liter was significantly associated with the development of myalgia (odds ratio, 2.9); a creatine kinase level of 300 U/L or higher (OR, 6.0); proteinuria or myoglobinuria (OR, 12.6); and a creatine kinase level of 500 U/L or higher (OR, 20.9).

In addition, a higher HDL-triglyceride level at baseline was associated with a borderline increase in the risk of any (muscular or nonmuscular) adverse effect (OR, 2.5; P = .065).

In contrast, the level of HDL cholesterol was not significantly associated with the likelihood of statin adverse effects.

The study’s findings should be validated and replicated in other populations, said Dr. Bouhali.

In addition, “further mechanistic studies are needed to explore the implication of HDL particle composition in the development of statin intolerance,” he said in comments provided by e-mail.

Dr. Bouhali reported that he had no relevant conflicts of interest.

VANCOUVER, B.C. – Information about a patient’s risk of developing statin-induced myotoxicity may come from a surprising source: the composition of his or her high-density lipoprotein particles, a study has shown.

In the observational study of 670 patients, those having higher levels of HDL particles rich in triglycerides before starting statins were 3-21 times more likely to develop various types of myotoxicity while on these medications.

“These results suggest an association between HDL-triglyceride concentration at baseline and markers of statin intolerance,” lead investigator Dr. Tarek Bouhali said at the annual World Congress on Heart Disease sponsored by the International Academy of Cardiology.

“Right now, we can’t tell what the mechanism is,” he commented. “But ... HDL is a heterogeneous group of lipoproteins and is not only cholesterol,” and in this setting, it appears that heterogeneity may be useful for risk stratification.

Skeletal muscle adverse events are the leading complication of statin therapy, according to Dr. Bouhali. “Muscular side effects are usually mild and reversible; however, knowledge of the underlying mechanism and risk factors is required for prompt identification and proper management for these side effects.”

Furthermore, he noted, understanding and preventing statin adverse effects are important because side effects can lead to both reduced compliance and increased health care costs.

The investigators collected fasting plasma samples from adult French Canadian patients who were about to start lipid-lowering therapy and were receiving care according to American Heart Association guidelines. The samples were ultracentrifuged to determine levels of cholesterol and triglyceride associated with HDL particles.

The study patients were 50 years old on average, and 63% were male. About 12% were taking statins in combination with other lipid-lowering medications (fibrates, bile acid sequestrants, nicotinic acid, or inhibitors of cholesterol absorption).

Overall, 22% of the patients had at least one adverse effect of any type (muscular or nonmuscular), reported Dr. Bouhali of the University of Montreal and the lead investigator with the ECOGENE-21 project.

In terms of the muscular adverse effects, 20% of patients developed myalgia, 7% had a creatine kinase level of 300 U/L or higher, 3% had trace proteinuria or myoglobinuria, and 2% had a creatine kinase level of 500 U/L or higher.

After adjustment for age, sex, statin dose, and concomitant medications, a higher plasma HDL-triglyceride level before starting treatment was associated with a greater risk of myotoxicity, and especially the more severe forms, noted Dr. Bouhali.

Specifically, the log10-transformed HDL-triglyceride level in millimoles per liter was significantly associated with the development of myalgia (odds ratio, 2.9); a creatine kinase level of 300 U/L or higher (OR, 6.0); proteinuria or myoglobinuria (OR, 12.6); and a creatine kinase level of 500 U/L or higher (OR, 20.9).

In addition, a higher HDL-triglyceride level at baseline was associated with a borderline increase in the risk of any (muscular or nonmuscular) adverse effect (OR, 2.5; P = .065).

In contrast, the level of HDL cholesterol was not significantly associated with the likelihood of statin adverse effects.

The study’s findings should be validated and replicated in other populations, said Dr. Bouhali.

In addition, “further mechanistic studies are needed to explore the implication of HDL particle composition in the development of statin intolerance,” he said in comments provided by e-mail.

Dr. Bouhali reported that he had no relevant conflicts of interest.

SEATTLE – Careful timing in measuring high-sensitivity C-reactive protein during the menstrual cycle can make all the difference in classifying young women’s risk of cardiovascular disease, new data show.

In a study of 259 healthy premenopausal women, high-sensitivity C-reactive protein (hs-CRP) levels fluctuated considerably over the course of the menstrual cycle, with the highest (and most variable) levels seen during menses and the lowest seen at ovulation, researchers reported at the annual meeting of the Society for Pediatric and Perinatal Epidemiologic Research.

The proportion of women classified as having a high or moderate risk for cardiovascular disease based on their levels of hs-CRP, a marker of chronic inflammation, was significantly greater when levels measured during menses were used (41%) than when levels at ovulation were used (29%).

“The take-home message here is that the measurement of CRP in clinical settings and in future research studies should be standardized to the menstrual cycle phase,” said lead investigator Audrey J. Gaskins, a postbaccalaureate fellow at the National Institute of Child Health and Human Development in Rockville, Md.

“Ideally, [one should] measure CRP around ovulation, when levels are lowest, but that is generally hard to time,” she commented. “So I would say any time other than menses, would be ideal.”

Several lines of evidence suggest that estrogen may modulate inflammation to a clinically relevant extent when it comes to cardiovascular outcomes, according to Ms. Gaskins.

“The risk of coronary events rises in women after menopause, and this corresponds to when endogenous estrogen levels decrease,” she explained. “Also, two recent studies have shown that in regularly menstruating women, there are more acute coronary events in the early follicular phase, when estrogen levels are lowest.”

Ms. Gaskins and her colleagues analyzed data from 259 healthy, normally menstruating women, aged an average of 27 years, and who were followed for up to two menstrual cycles in the BioCycle Study.

Serum samples collected at eight distinct time points during the menstrual cycle were assayed for levels of hormones and hs-CRP. Any hs-CRP values exceeding 10 mg/L were excluded under the assumption that they reflected acute illness.

Ms. Gaskins noted that the population was more diverse than those in previous studies. Some 59% of the women were white, 20% were black, and 21% were of other races. Although 61% had a body mass index in the normal range, 25% were overweight, 10% obese, and 3% underweight (percentages rounded). Seventy four percent were nulliparous, and 4% were smokers.

Study results showed that hs-CRP levels varied widely over the menstrual cycle, she reported. They were highest and also showed the greatest inter-individual variability during menses, and lowest at ovulation, with about a 1.6-fold difference in values between these two time points.

In adjusted models, hs-CRP was significantly associated both with estradiol across the menstrual cycle and with progesterone during the luteal phase. Specifically, hs-CRP levels fell by 24% with each 10-fold increase in estradiol level and increased by 19% with each 10-fold increase in luteal progesterone level.

In a final analysis, the investigators classified the women according to the American Heart Association risk classification system, whereby cardiovascular disease risk is considered high if hs-CRP level is greater than 3 mg/L and moderate if it is 1- 3 mg/L.

Although 32% of women had hs-CRP levels in the high-risk category at at least one time point during the menstrual cycle, only 2% consistently had levels in this category at all eight time points.

Some 41% of the women had hs-CRP levels that placed them in the high- or moderate-risk category during menses, whereas only 29% had high levels at ovulation (P less than .05). The percentages at all other time points, except for the midluteal time point, were also significantly lower than those at menses.

“This is the largest study by far to look at hs-CRP and reproductive hormones in healthy premenopausal women,” noted Ms. Gaskins.

“In regard to estradiol, our results support the hypothesis that estrogen might have anti-inflammatory effects,” she said. “In regard to progesterone, our results support an inflammatory role.”

Having information on other serum markers of inflammation, such as interleukin-6 and tumor necrosis factor-alpha, would have been helpful, she acknowledged. She also noted that owing to strict inclusion criteria, the study population might not apply to all women, even though it was more generalizable than those in previous studies.

Ms. Gaskins reported that she had no relevant conflicts of interest.

SEATTLE – Careful timing in measuring high-sensitivity C-reactive protein during the menstrual cycle can make all the difference in classifying young women’s risk of cardiovascular disease, new data show.

In a study of 259 healthy premenopausal women, high-sensitivity C-reactive protein (hs-CRP) levels fluctuated considerably over the course of the menstrual cycle, with the highest (and most variable) levels seen during menses and the lowest seen at ovulation, researchers reported at the annual meeting of the Society for Pediatric and Perinatal Epidemiologic Research.

The proportion of women classified as having a high or moderate risk for cardiovascular disease based on their levels of hs-CRP, a marker of chronic inflammation, was significantly greater when levels measured during menses were used (41%) than when levels at ovulation were used (29%).

“The take-home message here is that the measurement of CRP in clinical settings and in future research studies should be standardized to the menstrual cycle phase,” said lead investigator Audrey J. Gaskins, a postbaccalaureate fellow at the National Institute of Child Health and Human Development in Rockville, Md.

“Ideally, [one should] measure CRP around ovulation, when levels are lowest, but that is generally hard to time,” she commented. “So I would say any time other than menses, would be ideal.”

Several lines of evidence suggest that estrogen may modulate inflammation to a clinically relevant extent when it comes to cardiovascular outcomes, according to Ms. Gaskins.

“The risk of coronary events rises in women after menopause, and this corresponds to when endogenous estrogen levels decrease,” she explained. “Also, two recent studies have shown that in regularly menstruating women, there are more acute coronary events in the early follicular phase, when estrogen levels are lowest.”

Ms. Gaskins and her colleagues analyzed data from 259 healthy, normally menstruating women, aged an average of 27 years, and who were followed for up to two menstrual cycles in the BioCycle Study.

Serum samples collected at eight distinct time points during the menstrual cycle were assayed for levels of hormones and hs-CRP. Any hs-CRP values exceeding 10 mg/L were excluded under the assumption that they reflected acute illness.

Ms. Gaskins noted that the population was more diverse than those in previous studies. Some 59% of the women were white, 20% were black, and 21% were of other races. Although 61% had a body mass index in the normal range, 25% were overweight, 10% obese, and 3% underweight (percentages rounded). Seventy four percent were nulliparous, and 4% were smokers.

Study results showed that hs-CRP levels varied widely over the menstrual cycle, she reported. They were highest and also showed the greatest inter-individual variability during menses, and lowest at ovulation, with about a 1.6-fold difference in values between these two time points.

In adjusted models, hs-CRP was significantly associated both with estradiol across the menstrual cycle and with progesterone during the luteal phase. Specifically, hs-CRP levels fell by 24% with each 10-fold increase in estradiol level and increased by 19% with each 10-fold increase in luteal progesterone level.

In a final analysis, the investigators classified the women according to the American Heart Association risk classification system, whereby cardiovascular disease risk is considered high if hs-CRP level is greater than 3 mg/L and moderate if it is 1- 3 mg/L.

Although 32% of women had hs-CRP levels in the high-risk category at at least one time point during the menstrual cycle, only 2% consistently had levels in this category at all eight time points.

Some 41% of the women had hs-CRP levels that placed them in the high- or moderate-risk category during menses, whereas only 29% had high levels at ovulation (P less than .05). The percentages at all other time points, except for the midluteal time point, were also significantly lower than those at menses.

“This is the largest study by far to look at hs-CRP and reproductive hormones in healthy premenopausal women,” noted Ms. Gaskins.

“In regard to estradiol, our results support the hypothesis that estrogen might have anti-inflammatory effects,” she said. “In regard to progesterone, our results support an inflammatory role.”

Having information on other serum markers of inflammation, such as interleukin-6 and tumor necrosis factor-alpha, would have been helpful, she acknowledged. She also noted that owing to strict inclusion criteria, the study population might not apply to all women, even though it was more generalizable than those in previous studies.

Ms. Gaskins reported that she had no relevant conflicts of interest.

SEATTLE – Careful timing in measuring high-sensitivity C-reactive protein during the menstrual cycle can make all the difference in classifying young women’s risk of cardiovascular disease, new data show.

In a study of 259 healthy premenopausal women, high-sensitivity C-reactive protein (hs-CRP) levels fluctuated considerably over the course of the menstrual cycle, with the highest (and most variable) levels seen during menses and the lowest seen at ovulation, researchers reported at the annual meeting of the Society for Pediatric and Perinatal Epidemiologic Research.

The proportion of women classified as having a high or moderate risk for cardiovascular disease based on their levels of hs-CRP, a marker of chronic inflammation, was significantly greater when levels measured during menses were used (41%) than when levels at ovulation were used (29%).

“The take-home message here is that the measurement of CRP in clinical settings and in future research studies should be standardized to the menstrual cycle phase,” said lead investigator Audrey J. Gaskins, a postbaccalaureate fellow at the National Institute of Child Health and Human Development in Rockville, Md.

“Ideally, [one should] measure CRP around ovulation, when levels are lowest, but that is generally hard to time,” she commented. “So I would say any time other than menses, would be ideal.”

Several lines of evidence suggest that estrogen may modulate inflammation to a clinically relevant extent when it comes to cardiovascular outcomes, according to Ms. Gaskins.

“The risk of coronary events rises in women after menopause, and this corresponds to when endogenous estrogen levels decrease,” she explained. “Also, two recent studies have shown that in regularly menstruating women, there are more acute coronary events in the early follicular phase, when estrogen levels are lowest.”

Ms. Gaskins and her colleagues analyzed data from 259 healthy, normally menstruating women, aged an average of 27 years, and who were followed for up to two menstrual cycles in the BioCycle Study.

Serum samples collected at eight distinct time points during the menstrual cycle were assayed for levels of hormones and hs-CRP. Any hs-CRP values exceeding 10 mg/L were excluded under the assumption that they reflected acute illness.

Ms. Gaskins noted that the population was more diverse than those in previous studies. Some 59% of the women were white, 20% were black, and 21% were of other races. Although 61% had a body mass index in the normal range, 25% were overweight, 10% obese, and 3% underweight (percentages rounded). Seventy four percent were nulliparous, and 4% were smokers.

Study results showed that hs-CRP levels varied widely over the menstrual cycle, she reported. They were highest and also showed the greatest inter-individual variability during menses, and lowest at ovulation, with about a 1.6-fold difference in values between these two time points.

In adjusted models, hs-CRP was significantly associated both with estradiol across the menstrual cycle and with progesterone during the luteal phase. Specifically, hs-CRP levels fell by 24% with each 10-fold increase in estradiol level and increased by 19% with each 10-fold increase in luteal progesterone level.

In a final analysis, the investigators classified the women according to the American Heart Association risk classification system, whereby cardiovascular disease risk is considered high if hs-CRP level is greater than 3 mg/L and moderate if it is 1- 3 mg/L.

Although 32% of women had hs-CRP levels in the high-risk category at at least one time point during the menstrual cycle, only 2% consistently had levels in this category at all eight time points.

Some 41% of the women had hs-CRP levels that placed them in the high- or moderate-risk category during menses, whereas only 29% had high levels at ovulation (P less than .05). The percentages at all other time points, except for the midluteal time point, were also significantly lower than those at menses.

“This is the largest study by far to look at hs-CRP and reproductive hormones in healthy premenopausal women,” noted Ms. Gaskins.

“In regard to estradiol, our results support the hypothesis that estrogen might have anti-inflammatory effects,” she said. “In regard to progesterone, our results support an inflammatory role.”

Having information on other serum markers of inflammation, such as interleukin-6 and tumor necrosis factor-alpha, would have been helpful, she acknowledged. She also noted that owing to strict inclusion criteria, the study population might not apply to all women, even though it was more generalizable than those in previous studies.

Ms. Gaskins reported that she had no relevant conflicts of interest.