Robert Finn

Major Finding: Consumption of sugar-sweetened beverages contributed to an estimated 130,000 new cases of diabetes and 14,000 new cases of coronary heart disease between 2000 and 2010.

Data Source: Computer simulation based on the Coronary Heart Disease Policy Model.

Disclosures: Supported by a grant from the American Heart Association Western States Affiliate.

SAN FRANCISCO — The increase in the consumption of sugar-sweetened beverages between 1990 and 2000 contributed to 130,000 new cases of diabetes and 14,000 new cases of coronary heart disease between 2000 and 2010, according to estimates from a computer model of the U.S. population.

In addition, the rising consumption of sugar-sweetened beverages, which include soda, sports drinks, and fruit drinks, led to an estimated 1.4 million additional life-years burdened by diabetes and 50,000 additional life-years burdened by coronary heart disease in the first decade of the 21st century.

To derive those estimates, Dr. Litsa K. Lambrakos of the University of California, San Francisco, and her colleagues used data from the 1990-2000 National Health and Nutrition Examination Survey (NHANES) on consumption of sugar-sweetened beverages. She combined that with the Coronary Heart Disease Policy Model, a computer simulation of heart disease in U.S. adults aged 35-84 years.

According to that model, the relative risk of incident diabetes related to the daily consumption of sugar-sweetened beverages was 1.32 after adjusting for body mass index. Dr. Lambrakos presented the findings during a poster session at a conference sponsored by the American Heart Association.

The estimated increase in coronary heart disease related to the increased consumption of sugar-sweetened beverages would have generated an additional $300-$500 million in health care costs between 2000 and 2010.

“The numbers about excess health care costs are very conservative, because they only account for health care costs attributed to coronary heart disease,” Dr. Lambrakos said in an interview. “We know we have an increase in diabetes as well that we can attribute to soft drink consumption. And those costs—the cost of caring for and treating patients with diabetes—is a very large number as well.”

The investigators also analyzed how a 1 cent per ounce tax on sugar-sweetened beverages might have limited coronary heart disease costs, had it been implemented in the year 2000. Based on economic studies, the computer model assumed that such a tax would decrease consumption by 10%. This would translate to a savings of $170 million in health care costs over 10 years.

Commenting on the study findings, the American Heart Association issued this statement:

“The AHA acknowledges the importance of limiting intake of added sugars, including sugar-sweetened beverages. The association is still evaluating the research to determine which strategies accomplish this best, comparing more punitive strategies like taxation with more positive incentives like subsidies or lowering prices for healthy foods. The AHA will continue to monitor the best available research to more fully understand the connection between taxation policy and consumption trends, and ensure that our public policy positions reflect the best available science…. [R]obust evaluation should be part of any tax measures that are passed and advocates for broader nutrition policy efforts that make healthy foods more affordable and accessible to all consumers and bring food pricing and subsidies in line with federal dietary guidelines and AHA nutrition recommendations.”

Asked what message primary care physicians should take from the findings, Dr. Lambrakos said that “what we're talking about here is primary prevention…. It's important for the general public and physicians to understand that these drinks really shouldn't be considered a staple of the American diet.”

The relative risk of diabetes related to daily consumption of sugary beverages was 1.32 after adjusting for BMI.

Source DR. LAMBRAKOS

Major Finding: Consumption of sugar-sweetened beverages contributed to an estimated 130,000 new cases of diabetes and 14,000 new cases of coronary heart disease between 2000 and 2010.

Data Source: Computer simulation based on the Coronary Heart Disease Policy Model.

Disclosures: Supported by a grant from the American Heart Association Western States Affiliate.

SAN FRANCISCO — The increase in the consumption of sugar-sweetened beverages between 1990 and 2000 contributed to 130,000 new cases of diabetes and 14,000 new cases of coronary heart disease between 2000 and 2010, according to estimates from a computer model of the U.S. population.

In addition, the rising consumption of sugar-sweetened beverages, which include soda, sports drinks, and fruit drinks, led to an estimated 1.4 million additional life-years burdened by diabetes and 50,000 additional life-years burdened by coronary heart disease in the first decade of the 21st century.

To derive those estimates, Dr. Litsa K. Lambrakos of the University of California, San Francisco, and her colleagues used data from the 1990-2000 National Health and Nutrition Examination Survey (NHANES) on consumption of sugar-sweetened beverages. She combined that with the Coronary Heart Disease Policy Model, a computer simulation of heart disease in U.S. adults aged 35-84 years.

According to that model, the relative risk of incident diabetes related to the daily consumption of sugar-sweetened beverages was 1.32 after adjusting for body mass index. Dr. Lambrakos presented the findings during a poster session at a conference sponsored by the American Heart Association.

The estimated increase in coronary heart disease related to the increased consumption of sugar-sweetened beverages would have generated an additional $300-$500 million in health care costs between 2000 and 2010.

“The numbers about excess health care costs are very conservative, because they only account for health care costs attributed to coronary heart disease,” Dr. Lambrakos said in an interview. “We know we have an increase in diabetes as well that we can attribute to soft drink consumption. And those costs—the cost of caring for and treating patients with diabetes—is a very large number as well.”

The investigators also analyzed how a 1 cent per ounce tax on sugar-sweetened beverages might have limited coronary heart disease costs, had it been implemented in the year 2000. Based on economic studies, the computer model assumed that such a tax would decrease consumption by 10%. This would translate to a savings of $170 million in health care costs over 10 years.

Commenting on the study findings, the American Heart Association issued this statement:

“The AHA acknowledges the importance of limiting intake of added sugars, including sugar-sweetened beverages. The association is still evaluating the research to determine which strategies accomplish this best, comparing more punitive strategies like taxation with more positive incentives like subsidies or lowering prices for healthy foods. The AHA will continue to monitor the best available research to more fully understand the connection between taxation policy and consumption trends, and ensure that our public policy positions reflect the best available science…. [R]obust evaluation should be part of any tax measures that are passed and advocates for broader nutrition policy efforts that make healthy foods more affordable and accessible to all consumers and bring food pricing and subsidies in line with federal dietary guidelines and AHA nutrition recommendations.”

Asked what message primary care physicians should take from the findings, Dr. Lambrakos said that “what we're talking about here is primary prevention…. It's important for the general public and physicians to understand that these drinks really shouldn't be considered a staple of the American diet.”

The relative risk of diabetes related to daily consumption of sugary beverages was 1.32 after adjusting for BMI.

Source DR. LAMBRAKOS

Major Finding: Consumption of sugar-sweetened beverages contributed to an estimated 130,000 new cases of diabetes and 14,000 new cases of coronary heart disease between 2000 and 2010.

Data Source: Computer simulation based on the Coronary Heart Disease Policy Model.

Disclosures: Supported by a grant from the American Heart Association Western States Affiliate.

SAN FRANCISCO — The increase in the consumption of sugar-sweetened beverages between 1990 and 2000 contributed to 130,000 new cases of diabetes and 14,000 new cases of coronary heart disease between 2000 and 2010, according to estimates from a computer model of the U.S. population.

In addition, the rising consumption of sugar-sweetened beverages, which include soda, sports drinks, and fruit drinks, led to an estimated 1.4 million additional life-years burdened by diabetes and 50,000 additional life-years burdened by coronary heart disease in the first decade of the 21st century.

To derive those estimates, Dr. Litsa K. Lambrakos of the University of California, San Francisco, and her colleagues used data from the 1990-2000 National Health and Nutrition Examination Survey (NHANES) on consumption of sugar-sweetened beverages. She combined that with the Coronary Heart Disease Policy Model, a computer simulation of heart disease in U.S. adults aged 35-84 years.

According to that model, the relative risk of incident diabetes related to the daily consumption of sugar-sweetened beverages was 1.32 after adjusting for body mass index. Dr. Lambrakos presented the findings during a poster session at a conference sponsored by the American Heart Association.

The estimated increase in coronary heart disease related to the increased consumption of sugar-sweetened beverages would have generated an additional $300-$500 million in health care costs between 2000 and 2010.

“The numbers about excess health care costs are very conservative, because they only account for health care costs attributed to coronary heart disease,” Dr. Lambrakos said in an interview. “We know we have an increase in diabetes as well that we can attribute to soft drink consumption. And those costs—the cost of caring for and treating patients with diabetes—is a very large number as well.”

The investigators also analyzed how a 1 cent per ounce tax on sugar-sweetened beverages might have limited coronary heart disease costs, had it been implemented in the year 2000. Based on economic studies, the computer model assumed that such a tax would decrease consumption by 10%. This would translate to a savings of $170 million in health care costs over 10 years.

Commenting on the study findings, the American Heart Association issued this statement:

“The AHA acknowledges the importance of limiting intake of added sugars, including sugar-sweetened beverages. The association is still evaluating the research to determine which strategies accomplish this best, comparing more punitive strategies like taxation with more positive incentives like subsidies or lowering prices for healthy foods. The AHA will continue to monitor the best available research to more fully understand the connection between taxation policy and consumption trends, and ensure that our public policy positions reflect the best available science…. [R]obust evaluation should be part of any tax measures that are passed and advocates for broader nutrition policy efforts that make healthy foods more affordable and accessible to all consumers and bring food pricing and subsidies in line with federal dietary guidelines and AHA nutrition recommendations.”

Asked what message primary care physicians should take from the findings, Dr. Lambrakos said that “what we're talking about here is primary prevention…. It's important for the general public and physicians to understand that these drinks really shouldn't be considered a staple of the American diet.”

The relative risk of diabetes related to daily consumption of sugary beverages was 1.32 after adjusting for BMI.

Source DR. LAMBRAKOS

Major Finding: Women who drank more than 4 cups of coffee daily were 56% less likely to develop type 2 diabetes than were those who drank no coffee, a significant difference. After adjustment for SHBG, the difference was not significant.

Data Source: Nested case-control study of 359 women with incident type 2 diabetes and matched controls from the Women's Health Study.

Disclosures: Study supported by a grant from the National Institutes of Health.

SAN FRANCISCO — Sex hormone–binding globulin may be the key to the protective effect of coffee consumption against development of type 2 diabetes, according to an analysis of the Women's Health Study.

Women who drank at least 4 cups of coffee per day were less than half as likely to develop diabetes than were those who drank no coffee, and after adjustment for sex hormone–binding globulin (SHBG), the interaction disappeared.

It has been known for some time that women who drink coffee are significantly less likely to develop type 2 diabetes than are those who do not, and that the relationship between coffee consumption and diabetes is much less pronounced in men.

SHBG is a glycoprotein with a high affinity for testosterone and estradiol. SHBG levels tend to be substantially higher in women than in men, Atsushi Goto, a doctoral candidate at the University of California, Los Angeles, said at a conference sponsored by the American Heart Association. Previous studies have shown that variations in the genes controlling SHBG have a strong association with the development of diabetes and that coffee consumption increases plasma levels of SHBG.

To study this association, Mr. Goto and his colleagues used data from the Women's Health Study, in which nearly 40,000 women were followed for a median of 10 years. During that time, 359 of the women developed diabetes. The investigators matched those women by age, race, and time of blood draw with 359 women who had not developed the disease.

After adjustment for age, smoking, alcohol consumption, physical activity, past use of hormone therapy, total energy intake, fiber intake, body mass index, and plasma testosterone and estradiol levels, the investigators found that women who drank at least 4 cups of caffeinated coffee (500 mg caffeine) daily had significantly higher mean SHBG levels than did nondrinkers: 27.3 nmol/L versus 24.5 nmol/L. Decaffeinated coffee was not significantly associated with SHBG levels.

Furthermore, when controlling for all of the above factors plus education levels and family history of type 2 diabetes, the investigators found that women who drank at least 4 cups of caffeinated coffee daily were 56% less likely to develop diabetes than were nondrinkers. However, when the investigators additionally controlled for plasma SHBG levels, the decrease in risk associated with coffee consumption became nonsignificant. This suggests that it is SHBG that mediates the decrease in risk of developing type 2 diabetes, Mr. Goto commented.

Major Finding: Women who drank more than 4 cups of coffee daily were 56% less likely to develop type 2 diabetes than were those who drank no coffee, a significant difference. After adjustment for SHBG, the difference was not significant.

Data Source: Nested case-control study of 359 women with incident type 2 diabetes and matched controls from the Women's Health Study.

Disclosures: Study supported by a grant from the National Institutes of Health.

SAN FRANCISCO — Sex hormone–binding globulin may be the key to the protective effect of coffee consumption against development of type 2 diabetes, according to an analysis of the Women's Health Study.

Women who drank at least 4 cups of coffee per day were less than half as likely to develop diabetes than were those who drank no coffee, and after adjustment for sex hormone–binding globulin (SHBG), the interaction disappeared.

It has been known for some time that women who drink coffee are significantly less likely to develop type 2 diabetes than are those who do not, and that the relationship between coffee consumption and diabetes is much less pronounced in men.

SHBG is a glycoprotein with a high affinity for testosterone and estradiol. SHBG levels tend to be substantially higher in women than in men, Atsushi Goto, a doctoral candidate at the University of California, Los Angeles, said at a conference sponsored by the American Heart Association. Previous studies have shown that variations in the genes controlling SHBG have a strong association with the development of diabetes and that coffee consumption increases plasma levels of SHBG.

To study this association, Mr. Goto and his colleagues used data from the Women's Health Study, in which nearly 40,000 women were followed for a median of 10 years. During that time, 359 of the women developed diabetes. The investigators matched those women by age, race, and time of blood draw with 359 women who had not developed the disease.

After adjustment for age, smoking, alcohol consumption, physical activity, past use of hormone therapy, total energy intake, fiber intake, body mass index, and plasma testosterone and estradiol levels, the investigators found that women who drank at least 4 cups of caffeinated coffee (500 mg caffeine) daily had significantly higher mean SHBG levels than did nondrinkers: 27.3 nmol/L versus 24.5 nmol/L. Decaffeinated coffee was not significantly associated with SHBG levels.

Furthermore, when controlling for all of the above factors plus education levels and family history of type 2 diabetes, the investigators found that women who drank at least 4 cups of caffeinated coffee daily were 56% less likely to develop diabetes than were nondrinkers. However, when the investigators additionally controlled for plasma SHBG levels, the decrease in risk associated with coffee consumption became nonsignificant. This suggests that it is SHBG that mediates the decrease in risk of developing type 2 diabetes, Mr. Goto commented.

Major Finding: Women who drank more than 4 cups of coffee daily were 56% less likely to develop type 2 diabetes than were those who drank no coffee, a significant difference. After adjustment for SHBG, the difference was not significant.

Data Source: Nested case-control study of 359 women with incident type 2 diabetes and matched controls from the Women's Health Study.

Disclosures: Study supported by a grant from the National Institutes of Health.

SAN FRANCISCO — Sex hormone–binding globulin may be the key to the protective effect of coffee consumption against development of type 2 diabetes, according to an analysis of the Women's Health Study.

Women who drank at least 4 cups of coffee per day were less than half as likely to develop diabetes than were those who drank no coffee, and after adjustment for sex hormone–binding globulin (SHBG), the interaction disappeared.

It has been known for some time that women who drink coffee are significantly less likely to develop type 2 diabetes than are those who do not, and that the relationship between coffee consumption and diabetes is much less pronounced in men.

SHBG is a glycoprotein with a high affinity for testosterone and estradiol. SHBG levels tend to be substantially higher in women than in men, Atsushi Goto, a doctoral candidate at the University of California, Los Angeles, said at a conference sponsored by the American Heart Association. Previous studies have shown that variations in the genes controlling SHBG have a strong association with the development of diabetes and that coffee consumption increases plasma levels of SHBG.

To study this association, Mr. Goto and his colleagues used data from the Women's Health Study, in which nearly 40,000 women were followed for a median of 10 years. During that time, 359 of the women developed diabetes. The investigators matched those women by age, race, and time of blood draw with 359 women who had not developed the disease.

After adjustment for age, smoking, alcohol consumption, physical activity, past use of hormone therapy, total energy intake, fiber intake, body mass index, and plasma testosterone and estradiol levels, the investigators found that women who drank at least 4 cups of caffeinated coffee (500 mg caffeine) daily had significantly higher mean SHBG levels than did nondrinkers: 27.3 nmol/L versus 24.5 nmol/L. Decaffeinated coffee was not significantly associated with SHBG levels.

Furthermore, when controlling for all of the above factors plus education levels and family history of type 2 diabetes, the investigators found that women who drank at least 4 cups of caffeinated coffee daily were 56% less likely to develop diabetes than were nondrinkers. However, when the investigators additionally controlled for plasma SHBG levels, the decrease in risk associated with coffee consumption became nonsignificant. This suggests that it is SHBG that mediates the decrease in risk of developing type 2 diabetes, Mr. Goto commented.

Major Finding: Women randomized to a low-fat or a low-carbohydrate diet considered appropriate (based on a pattern of three single nucleotide polymorphisms) lost about 6 kg, compared with about 1-1.5 kg among women randomized to diets judged as inappropriate.

Data Source: Data from 133 participants in the A TO Z Weight Loss Study.

Disclosures: Research support was provided by Interleukin Genetics. Dr. Nelson had no financial conflicts.

SAN FRANCISCO — Genotypes may identify women likely to lose weight on a low-carbohydrate diet and those likely to do better on a low-fat diet, based on data from 133 participants in the A TO Z Weight Loss Study.

Women who have one pattern of single nucleotide polymorphisms (SNPs) lost five times as much weight on the Atkins diet, compared with those who did not have that pattern.

Similarly, women with a different SNP genotype lost five times as much weight on the Ornish diet, Mindy Dopler Nelson, Ph.D., of Stanford (Calif.) University, said at a conference sponsored by the American Heart Association.

In the original A TO Z study, 311 women were randomized to one of four popular diets. Ranging on a continuum from low carbohydrate to low fat, they were the Atkins diet, the Zone diet, the LEARN diet, and the Ornish diet.

On average, the women lost weight on all four diets; the only significant difference was that they tended to lose somewhat more weight on the Atkins diet than on the Ornish diet (JAMA 2007;297:969–77).

“Within each of the diet groups, there are women who had lost over 15 kg … as well as people who gained 5 kg,” Dr. Nelson said in an interview. “When you look at the averages you don't see the differences, but when you look at each individual participant you see some variability.”

Some time after the conclusion of that trial, Interleukin Genetics approached Stanford researchers and suggested they use the company's proprietary SNP test to assess responders and nonresponders to particular diets.

In previous studies, the company had found polymorphisms in three genes—fatty acid binding protein, PPAR-gamma, and the beta-2 adrenergic receptor—that appeared to predict a person's response to diets.

Among the 133 women from the original study who agreed to provide DNA samples from swabs of the inner cheek, 31 had been in the Atkins group, 32 in the Zone group, 34 in the LEARN group, and 36 in the Ornish group.

There were no statistically significant baseline differences among the groups in measures such as body mass index, blood pressure, or levels of cholesterol, insulin, and glucose.

The company's test showed that 79 of the women had genotypes designated as low-carb appropriate, and 54 had genotypes designated as low-fat appropriate.

The interaction between genotype and diet was statistically significant, with striking differences among the women in the lowest-carb and lowest-fat diets.

Among the women on the Atkins diet, those designated as low-carb appropriate lost an average of just under 6 kg during 12 months, while those designated as low-carb inappropriate lost about 1 kg.

Among the women on the Ornish diet, those designated as low-fat appropriate lost an average of more than 6 kg during 12 months, while those designated as low-fat inappropriate lost an average of about 1.5 kg.

Thus, in each of those groups, women who had been randomized to what was designated as the appropriate diet lost about five times as much weight as those randomized to the apparently inappropriate diet.

Among women on the Zone or LEARN diets, which involve intermediate levels of carbohydrates and fat, women with low-carb and low-fat genotypes did not have statistically significant differences in weight loss.

Dr. Nelson acknowledged that the trial was relatively small, and that the findings need to be confirmed in a larger trial in a more heterogeneous population.

Nevertheless, the results do provide some guidance to people who are trying to lose weight, she said.

“I would suggest that if somebody is discouraged by the weight loss that they're having on whatever particular diet they're following, they may just want to consider changing the distribution of their macronutrients,” Dr. Nelson said.

“So maybe if you're doing a higher-carbohydrate diet and you're not seeing your weight loss, give up some of the more processed carbohydrates, keep the healthier ones in there, and see if shifting to the lower-carbohydrate diet will help with weight loss,” she added.

Dr. Nelson said that Interleukin Genetics supported her study, but neither she nor Stanford University has any financial involvement in the company's products.

Major Finding: Women randomized to a low-fat or a low-carbohydrate diet considered appropriate (based on a pattern of three single nucleotide polymorphisms) lost about 6 kg, compared with about 1-1.5 kg among women randomized to diets judged as inappropriate.

Data Source: Data from 133 participants in the A TO Z Weight Loss Study.

Disclosures: Research support was provided by Interleukin Genetics. Dr. Nelson had no financial conflicts.

SAN FRANCISCO — Genotypes may identify women likely to lose weight on a low-carbohydrate diet and those likely to do better on a low-fat diet, based on data from 133 participants in the A TO Z Weight Loss Study.

Women who have one pattern of single nucleotide polymorphisms (SNPs) lost five times as much weight on the Atkins diet, compared with those who did not have that pattern.

Similarly, women with a different SNP genotype lost five times as much weight on the Ornish diet, Mindy Dopler Nelson, Ph.D., of Stanford (Calif.) University, said at a conference sponsored by the American Heart Association.

In the original A TO Z study, 311 women were randomized to one of four popular diets. Ranging on a continuum from low carbohydrate to low fat, they were the Atkins diet, the Zone diet, the LEARN diet, and the Ornish diet.

On average, the women lost weight on all four diets; the only significant difference was that they tended to lose somewhat more weight on the Atkins diet than on the Ornish diet (JAMA 2007;297:969–77).

“Within each of the diet groups, there are women who had lost over 15 kg … as well as people who gained 5 kg,” Dr. Nelson said in an interview. “When you look at the averages you don't see the differences, but when you look at each individual participant you see some variability.”

Some time after the conclusion of that trial, Interleukin Genetics approached Stanford researchers and suggested they use the company's proprietary SNP test to assess responders and nonresponders to particular diets.

In previous studies, the company had found polymorphisms in three genes—fatty acid binding protein, PPAR-gamma, and the beta-2 adrenergic receptor—that appeared to predict a person's response to diets.

Among the 133 women from the original study who agreed to provide DNA samples from swabs of the inner cheek, 31 had been in the Atkins group, 32 in the Zone group, 34 in the LEARN group, and 36 in the Ornish group.

There were no statistically significant baseline differences among the groups in measures such as body mass index, blood pressure, or levels of cholesterol, insulin, and glucose.

The company's test showed that 79 of the women had genotypes designated as low-carb appropriate, and 54 had genotypes designated as low-fat appropriate.

The interaction between genotype and diet was statistically significant, with striking differences among the women in the lowest-carb and lowest-fat diets.

Among the women on the Atkins diet, those designated as low-carb appropriate lost an average of just under 6 kg during 12 months, while those designated as low-carb inappropriate lost about 1 kg.

Among the women on the Ornish diet, those designated as low-fat appropriate lost an average of more than 6 kg during 12 months, while those designated as low-fat inappropriate lost an average of about 1.5 kg.

Thus, in each of those groups, women who had been randomized to what was designated as the appropriate diet lost about five times as much weight as those randomized to the apparently inappropriate diet.

Among women on the Zone or LEARN diets, which involve intermediate levels of carbohydrates and fat, women with low-carb and low-fat genotypes did not have statistically significant differences in weight loss.

Dr. Nelson acknowledged that the trial was relatively small, and that the findings need to be confirmed in a larger trial in a more heterogeneous population.

Nevertheless, the results do provide some guidance to people who are trying to lose weight, she said.

“I would suggest that if somebody is discouraged by the weight loss that they're having on whatever particular diet they're following, they may just want to consider changing the distribution of their macronutrients,” Dr. Nelson said.

“So maybe if you're doing a higher-carbohydrate diet and you're not seeing your weight loss, give up some of the more processed carbohydrates, keep the healthier ones in there, and see if shifting to the lower-carbohydrate diet will help with weight loss,” she added.

Dr. Nelson said that Interleukin Genetics supported her study, but neither she nor Stanford University has any financial involvement in the company's products.

Major Finding: Women randomized to a low-fat or a low-carbohydrate diet considered appropriate (based on a pattern of three single nucleotide polymorphisms) lost about 6 kg, compared with about 1-1.5 kg among women randomized to diets judged as inappropriate.

Data Source: Data from 133 participants in the A TO Z Weight Loss Study.

Disclosures: Research support was provided by Interleukin Genetics. Dr. Nelson had no financial conflicts.

SAN FRANCISCO — Genotypes may identify women likely to lose weight on a low-carbohydrate diet and those likely to do better on a low-fat diet, based on data from 133 participants in the A TO Z Weight Loss Study.

Women who have one pattern of single nucleotide polymorphisms (SNPs) lost five times as much weight on the Atkins diet, compared with those who did not have that pattern.

Similarly, women with a different SNP genotype lost five times as much weight on the Ornish diet, Mindy Dopler Nelson, Ph.D., of Stanford (Calif.) University, said at a conference sponsored by the American Heart Association.

In the original A TO Z study, 311 women were randomized to one of four popular diets. Ranging on a continuum from low carbohydrate to low fat, they were the Atkins diet, the Zone diet, the LEARN diet, and the Ornish diet.

On average, the women lost weight on all four diets; the only significant difference was that they tended to lose somewhat more weight on the Atkins diet than on the Ornish diet (JAMA 2007;297:969–77).

“Within each of the diet groups, there are women who had lost over 15 kg … as well as people who gained 5 kg,” Dr. Nelson said in an interview. “When you look at the averages you don't see the differences, but when you look at each individual participant you see some variability.”

Some time after the conclusion of that trial, Interleukin Genetics approached Stanford researchers and suggested they use the company's proprietary SNP test to assess responders and nonresponders to particular diets.

In previous studies, the company had found polymorphisms in three genes—fatty acid binding protein, PPAR-gamma, and the beta-2 adrenergic receptor—that appeared to predict a person's response to diets.

Among the 133 women from the original study who agreed to provide DNA samples from swabs of the inner cheek, 31 had been in the Atkins group, 32 in the Zone group, 34 in the LEARN group, and 36 in the Ornish group.

There were no statistically significant baseline differences among the groups in measures such as body mass index, blood pressure, or levels of cholesterol, insulin, and glucose.

The company's test showed that 79 of the women had genotypes designated as low-carb appropriate, and 54 had genotypes designated as low-fat appropriate.

The interaction between genotype and diet was statistically significant, with striking differences among the women in the lowest-carb and lowest-fat diets.

Among the women on the Atkins diet, those designated as low-carb appropriate lost an average of just under 6 kg during 12 months, while those designated as low-carb inappropriate lost about 1 kg.

Among the women on the Ornish diet, those designated as low-fat appropriate lost an average of more than 6 kg during 12 months, while those designated as low-fat inappropriate lost an average of about 1.5 kg.

Thus, in each of those groups, women who had been randomized to what was designated as the appropriate diet lost about five times as much weight as those randomized to the apparently inappropriate diet.

Among women on the Zone or LEARN diets, which involve intermediate levels of carbohydrates and fat, women with low-carb and low-fat genotypes did not have statistically significant differences in weight loss.

Dr. Nelson acknowledged that the trial was relatively small, and that the findings need to be confirmed in a larger trial in a more heterogeneous population.

Nevertheless, the results do provide some guidance to people who are trying to lose weight, she said.

“I would suggest that if somebody is discouraged by the weight loss that they're having on whatever particular diet they're following, they may just want to consider changing the distribution of their macronutrients,” Dr. Nelson said.

“So maybe if you're doing a higher-carbohydrate diet and you're not seeing your weight loss, give up some of the more processed carbohydrates, keep the healthier ones in there, and see if shifting to the lower-carbohydrate diet will help with weight loss,” she added.

Dr. Nelson said that Interleukin Genetics supported her study, but neither she nor Stanford University has any financial involvement in the company's products.

Major Finding: Men who were obese at age 50 years but did not have metabolic syndrome or insulin resistance had 10-15 times the chance of developing type 2 diabetes as did normal-weight men over 10 and 20 years.

Data Source: Longitudinal study of 934 men.

Disclosures: None reported.

SAN FRANCISCO — Men who are obese but “metabolically healthy” have no protection against developing type 2 diabetes, according to a large longitudinal study.

Compared with men of normal weight, men who have a body mass index greater than 30 kg/m

Other studies have shown that 25%–30% of individuals who are obese do not meet criteria for metabolic syndrome or insulin resistance. Earlier studies appeared to indicate that those individuals were unlikely to develop diabetes or cardiovascular disease. But those studies were hampered by relatively short period of follow-up.

“Our conclusion from this study … is that metabolically healthy obesity is not very healthy, that it is not a benign condition,” said Dr. Lars Lind of the University of Uppsala, Sweden, the study's lead investigator.

The investigators used data from the Uppsala Longitudinal Study of Adult Men, a cohort of men born during 1920-1924. At age 50 years 1,758 of these men were available for study; that number declined to 1,420 at age 60 and 934 at age 70.

The investigators defined normal weight as a BMI less than 25 kg/m

During 10 years of follow-up, 124 of the men developed diabetes, and that number increased to 169 after 20 years.

After correction for age, smoking, and LDL cholesterol level, obese men with and without metabolic syndrome had 10 times the risk of developing diabetes as did normal-weight men after 10 years. Overweight men without metabolic syndrome had a threefold increase in risk. All those increases in risk were statistically significant.

After 20 years, obese men without metabolic syndrome had 15 times the risk of developing diabetes and overweight men had 4 times the risk.

The situation with men who were not insulin resistant was similar. After 10 years, obese men with and without insulin resistance had a 15-fold increase in the risk of developing diabetes, and overweight men had a 3-fold increase in risk.

After 20 years, a statistically significant difference in risk appeared between obese men with and without insulin resistance. Men without insulin resistance were 15 times as likely to develop diabetes, while men with insulin resistance had a 30-fold increase in risk. But the 15-fold increase in risk among men without insulin resistance at baseline was still significantly elevated.

Metabolically healthy obesity is not a benign condtion, researchers concluded.

Source ©geronimo/Fotolia.com

Major Finding: Men who were obese at age 50 years but did not have metabolic syndrome or insulin resistance had 10-15 times the chance of developing type 2 diabetes as did normal-weight men over 10 and 20 years.

Data Source: Longitudinal study of 934 men.

Disclosures: None reported.

SAN FRANCISCO — Men who are obese but “metabolically healthy” have no protection against developing type 2 diabetes, according to a large longitudinal study.

Compared with men of normal weight, men who have a body mass index greater than 30 kg/m

Other studies have shown that 25%–30% of individuals who are obese do not meet criteria for metabolic syndrome or insulin resistance. Earlier studies appeared to indicate that those individuals were unlikely to develop diabetes or cardiovascular disease. But those studies were hampered by relatively short period of follow-up.

“Our conclusion from this study … is that metabolically healthy obesity is not very healthy, that it is not a benign condition,” said Dr. Lars Lind of the University of Uppsala, Sweden, the study's lead investigator.

The investigators used data from the Uppsala Longitudinal Study of Adult Men, a cohort of men born during 1920-1924. At age 50 years 1,758 of these men were available for study; that number declined to 1,420 at age 60 and 934 at age 70.

The investigators defined normal weight as a BMI less than 25 kg/m

During 10 years of follow-up, 124 of the men developed diabetes, and that number increased to 169 after 20 years.

After correction for age, smoking, and LDL cholesterol level, obese men with and without metabolic syndrome had 10 times the risk of developing diabetes as did normal-weight men after 10 years. Overweight men without metabolic syndrome had a threefold increase in risk. All those increases in risk were statistically significant.

After 20 years, obese men without metabolic syndrome had 15 times the risk of developing diabetes and overweight men had 4 times the risk.

The situation with men who were not insulin resistant was similar. After 10 years, obese men with and without insulin resistance had a 15-fold increase in the risk of developing diabetes, and overweight men had a 3-fold increase in risk.

After 20 years, a statistically significant difference in risk appeared between obese men with and without insulin resistance. Men without insulin resistance were 15 times as likely to develop diabetes, while men with insulin resistance had a 30-fold increase in risk. But the 15-fold increase in risk among men without insulin resistance at baseline was still significantly elevated.

Metabolically healthy obesity is not a benign condtion, researchers concluded.

Source ©geronimo/Fotolia.com

Major Finding: Men who were obese at age 50 years but did not have metabolic syndrome or insulin resistance had 10-15 times the chance of developing type 2 diabetes as did normal-weight men over 10 and 20 years.

Data Source: Longitudinal study of 934 men.

Disclosures: None reported.

SAN FRANCISCO — Men who are obese but “metabolically healthy” have no protection against developing type 2 diabetes, according to a large longitudinal study.

Compared with men of normal weight, men who have a body mass index greater than 30 kg/m

Other studies have shown that 25%–30% of individuals who are obese do not meet criteria for metabolic syndrome or insulin resistance. Earlier studies appeared to indicate that those individuals were unlikely to develop diabetes or cardiovascular disease. But those studies were hampered by relatively short period of follow-up.

“Our conclusion from this study … is that metabolically healthy obesity is not very healthy, that it is not a benign condition,” said Dr. Lars Lind of the University of Uppsala, Sweden, the study's lead investigator.

The investigators used data from the Uppsala Longitudinal Study of Adult Men, a cohort of men born during 1920-1924. At age 50 years 1,758 of these men were available for study; that number declined to 1,420 at age 60 and 934 at age 70.

The investigators defined normal weight as a BMI less than 25 kg/m

During 10 years of follow-up, 124 of the men developed diabetes, and that number increased to 169 after 20 years.

After correction for age, smoking, and LDL cholesterol level, obese men with and without metabolic syndrome had 10 times the risk of developing diabetes as did normal-weight men after 10 years. Overweight men without metabolic syndrome had a threefold increase in risk. All those increases in risk were statistically significant.

After 20 years, obese men without metabolic syndrome had 15 times the risk of developing diabetes and overweight men had 4 times the risk.

The situation with men who were not insulin resistant was similar. After 10 years, obese men with and without insulin resistance had a 15-fold increase in the risk of developing diabetes, and overweight men had a 3-fold increase in risk.

After 20 years, a statistically significant difference in risk appeared between obese men with and without insulin resistance. Men without insulin resistance were 15 times as likely to develop diabetes, while men with insulin resistance had a 30-fold increase in risk. But the 15-fold increase in risk among men without insulin resistance at baseline was still significantly elevated.

Metabolically healthy obesity is not a benign condtion, researchers concluded.

Source ©geronimo/Fotolia.com

A new “inside-out” hypothesis on the pathogenesis of rheumatoid arthritis suggests that joint damage may arise from within adjacent bone marrow, rather than—or in addition to—arising from outside via the synovial membrane, according to Dr. Georg Schett and Dr. Gary S. Firestein and published online in the journal Annals of the Rheumatic Diseases.

The conventional “outside-in” hypothesis holds that the primary pathogenic event in RA is the alteration of a synovial membrane. The altered membrane recruits immune cells, resulting in an onslaught of inflammation, cell accumulation from unbalanced proliferation and cell death, and perhaps a synovial immune response.

Either of two scenarios could lead to alterations in the synovial membrane. In one scenario, there's a confluence of environmental and genetic factors and the breakdown of tolerance. Alternatively, the synovial membrane could be changed by systemic processes.

The inside-out hypothesis holds that lesions within the bone marrow could begin to destroy the inner cortical bone surface, eventually opening pathways to the synovium. Mesenchymal elements could migrate through these cortical pores, stimulating joint inflammation, wrote Dr. Schett, professor of internal medicine, rheumatology, immunology, and oncology at the University of Erlangen-Nuremberg (Germany), and Dr. Firestein, professor of medicine of the University of California, San Diego.

Lesions within the bone marrow have been observed with MRI in the earliest stages of the disease. Microscopically, these lesions are sites where bone marrow fat has been replaced by inflammatory tissue dominated by lymphocytes. Other studies have demonstrated that these lesions are associated with structural damage in joints. “Different forms of arthritis may preferentially use either the outside-in or inside-out mechanism.…Preference for one of these two mechanisms may better explain some of the clinical differences of the various forms of arthritis.” (Arch. Rheum. Dis. 2010 March 18 [doi:10.1136/ard.2009.121657

Disclosures: The investigators stated that they had no conflicts of interest.

A new “inside-out” hypothesis on the pathogenesis of rheumatoid arthritis suggests that joint damage may arise from within adjacent bone marrow, rather than—or in addition to—arising from outside via the synovial membrane, according to Dr. Georg Schett and Dr. Gary S. Firestein and published online in the journal Annals of the Rheumatic Diseases.

The conventional “outside-in” hypothesis holds that the primary pathogenic event in RA is the alteration of a synovial membrane. The altered membrane recruits immune cells, resulting in an onslaught of inflammation, cell accumulation from unbalanced proliferation and cell death, and perhaps a synovial immune response.

Either of two scenarios could lead to alterations in the synovial membrane. In one scenario, there's a confluence of environmental and genetic factors and the breakdown of tolerance. Alternatively, the synovial membrane could be changed by systemic processes.

The inside-out hypothesis holds that lesions within the bone marrow could begin to destroy the inner cortical bone surface, eventually opening pathways to the synovium. Mesenchymal elements could migrate through these cortical pores, stimulating joint inflammation, wrote Dr. Schett, professor of internal medicine, rheumatology, immunology, and oncology at the University of Erlangen-Nuremberg (Germany), and Dr. Firestein, professor of medicine of the University of California, San Diego.

Lesions within the bone marrow have been observed with MRI in the earliest stages of the disease. Microscopically, these lesions are sites where bone marrow fat has been replaced by inflammatory tissue dominated by lymphocytes. Other studies have demonstrated that these lesions are associated with structural damage in joints. “Different forms of arthritis may preferentially use either the outside-in or inside-out mechanism.…Preference for one of these two mechanisms may better explain some of the clinical differences of the various forms of arthritis.” (Arch. Rheum. Dis. 2010 March 18 [doi:10.1136/ard.2009.121657

Disclosures: The investigators stated that they had no conflicts of interest.

A new “inside-out” hypothesis on the pathogenesis of rheumatoid arthritis suggests that joint damage may arise from within adjacent bone marrow, rather than—or in addition to—arising from outside via the synovial membrane, according to Dr. Georg Schett and Dr. Gary S. Firestein and published online in the journal Annals of the Rheumatic Diseases.

The conventional “outside-in” hypothesis holds that the primary pathogenic event in RA is the alteration of a synovial membrane. The altered membrane recruits immune cells, resulting in an onslaught of inflammation, cell accumulation from unbalanced proliferation and cell death, and perhaps a synovial immune response.

Either of two scenarios could lead to alterations in the synovial membrane. In one scenario, there's a confluence of environmental and genetic factors and the breakdown of tolerance. Alternatively, the synovial membrane could be changed by systemic processes.

The inside-out hypothesis holds that lesions within the bone marrow could begin to destroy the inner cortical bone surface, eventually opening pathways to the synovium. Mesenchymal elements could migrate through these cortical pores, stimulating joint inflammation, wrote Dr. Schett, professor of internal medicine, rheumatology, immunology, and oncology at the University of Erlangen-Nuremberg (Germany), and Dr. Firestein, professor of medicine of the University of California, San Diego.

Lesions within the bone marrow have been observed with MRI in the earliest stages of the disease. Microscopically, these lesions are sites where bone marrow fat has been replaced by inflammatory tissue dominated by lymphocytes. Other studies have demonstrated that these lesions are associated with structural damage in joints. “Different forms of arthritis may preferentially use either the outside-in or inside-out mechanism.…Preference for one of these two mechanisms may better explain some of the clinical differences of the various forms of arthritis.” (Arch. Rheum. Dis. 2010 March 18 [doi:10.1136/ard.2009.121657

Disclosures: The investigators stated that they had no conflicts of interest.

Major Finding: Women with depression were 54% more likely to develop overweight/obesity within 5 years than were those who were not depressed, and women who were overweight or obese were 27% more likely to develop depression within 5 years than were those who were not.

Data Source: A study of 5,031 men and women aged 45-84 years at baseline.

Disclosures: The investigator stated that she had no disclosures.

SAN FRANCISCO — Depression might lead to overweight and obesity, and overweight and obesity might also lead to depression, but only in women, according to a longitudinal study of 5,031 participants in the Multi-Ethnic Study of Atherosclerosis presented at a conference sponsored by the American Heart Association.

Numerous studies have demonstrated associations between depression and the development of type 2 diabetes and cardiovascular disease.

According to lead investigator Rosemay A. Remigio-Baker, a doctoral candidate at Johns Hopkins Bloomberg School of Public Health, Baltimore, overweight and obesity might provide the link connecting depression with diabetes and cardiovascular disease.

Participants entered the study between 2000 and 2002, when they were 45-84 years of age. Investigators followed them for 5 years. None of the participants had diabetes at baseline.

The investigators defined overweight as a body mass index of 25 kg/m

To see whether depression was associated with the later development of overweight, the investigators restricted their analysis to the 1,496 individuals whose baseline BMI was less than 25 kg/m

After controlling for age, ethnicity, education, income, smoking status, daily caloric intake, exercise, and levels of interleukin-6 and C-reactive protein, the investigators found that women with depression were 54% more likely to develop overweight or obesity within 5 years than were those without depression. The hazard ratio was statistically significant. The investigators found no statistically significant association between depression and incident overweight among men.

To see whether overweight/obesity was associated with the development of depression, the investigators restricted their analysis to the 3,801 participants without depression at baseline. At baseline, 65% of those women and 70% of those men were overweight or obese.

After controlling for the same set of covariates listed above, the investigators found that women who were overweight or obese were 27% more likely to develop depression within 5 years than were those who were not. The hazard ratio was statistically significant. Once again, the investigators found no statistically significant association between overweight/obesity and incident depression among men.

She stated that she had no disclosures.

Major Finding: Women with depression were 54% more likely to develop overweight/obesity within 5 years than were those who were not depressed, and women who were overweight or obese were 27% more likely to develop depression within 5 years than were those who were not.

Data Source: A study of 5,031 men and women aged 45-84 years at baseline.

Disclosures: The investigator stated that she had no disclosures.

SAN FRANCISCO — Depression might lead to overweight and obesity, and overweight and obesity might also lead to depression, but only in women, according to a longitudinal study of 5,031 participants in the Multi-Ethnic Study of Atherosclerosis presented at a conference sponsored by the American Heart Association.

Numerous studies have demonstrated associations between depression and the development of type 2 diabetes and cardiovascular disease.

According to lead investigator Rosemay A. Remigio-Baker, a doctoral candidate at Johns Hopkins Bloomberg School of Public Health, Baltimore, overweight and obesity might provide the link connecting depression with diabetes and cardiovascular disease.

Participants entered the study between 2000 and 2002, when they were 45-84 years of age. Investigators followed them for 5 years. None of the participants had diabetes at baseline.

The investigators defined overweight as a body mass index of 25 kg/m

To see whether depression was associated with the later development of overweight, the investigators restricted their analysis to the 1,496 individuals whose baseline BMI was less than 25 kg/m

After controlling for age, ethnicity, education, income, smoking status, daily caloric intake, exercise, and levels of interleukin-6 and C-reactive protein, the investigators found that women with depression were 54% more likely to develop overweight or obesity within 5 years than were those without depression. The hazard ratio was statistically significant. The investigators found no statistically significant association between depression and incident overweight among men.

To see whether overweight/obesity was associated with the development of depression, the investigators restricted their analysis to the 3,801 participants without depression at baseline. At baseline, 65% of those women and 70% of those men were overweight or obese.

After controlling for the same set of covariates listed above, the investigators found that women who were overweight or obese were 27% more likely to develop depression within 5 years than were those who were not. The hazard ratio was statistically significant. Once again, the investigators found no statistically significant association between overweight/obesity and incident depression among men.

She stated that she had no disclosures.

Major Finding: Women with depression were 54% more likely to develop overweight/obesity within 5 years than were those who were not depressed, and women who were overweight or obese were 27% more likely to develop depression within 5 years than were those who were not.

Data Source: A study of 5,031 men and women aged 45-84 years at baseline.

Disclosures: The investigator stated that she had no disclosures.

SAN FRANCISCO — Depression might lead to overweight and obesity, and overweight and obesity might also lead to depression, but only in women, according to a longitudinal study of 5,031 participants in the Multi-Ethnic Study of Atherosclerosis presented at a conference sponsored by the American Heart Association.

Numerous studies have demonstrated associations between depression and the development of type 2 diabetes and cardiovascular disease.

According to lead investigator Rosemay A. Remigio-Baker, a doctoral candidate at Johns Hopkins Bloomberg School of Public Health, Baltimore, overweight and obesity might provide the link connecting depression with diabetes and cardiovascular disease.

Participants entered the study between 2000 and 2002, when they were 45-84 years of age. Investigators followed them for 5 years. None of the participants had diabetes at baseline.

The investigators defined overweight as a body mass index of 25 kg/m

To see whether depression was associated with the later development of overweight, the investigators restricted their analysis to the 1,496 individuals whose baseline BMI was less than 25 kg/m

After controlling for age, ethnicity, education, income, smoking status, daily caloric intake, exercise, and levels of interleukin-6 and C-reactive protein, the investigators found that women with depression were 54% more likely to develop overweight or obesity within 5 years than were those without depression. The hazard ratio was statistically significant. The investigators found no statistically significant association between depression and incident overweight among men.

To see whether overweight/obesity was associated with the development of depression, the investigators restricted their analysis to the 3,801 participants without depression at baseline. At baseline, 65% of those women and 70% of those men were overweight or obese.

After controlling for the same set of covariates listed above, the investigators found that women who were overweight or obese were 27% more likely to develop depression within 5 years than were those who were not. The hazard ratio was statistically significant. Once again, the investigators found no statistically significant association between overweight/obesity and incident depression among men.

She stated that she had no disclosures.

Major Finding: Women randomized to a low-fat or a low-carbohydrate diet considered appropriate (based on a pattern of three single nucleotide polymorphisms) lost about 6 kg, compared with about 1-1.5 kg among women randomized to diets judged as inappropriate.

Data Source: Data from 133 participants in the A TO Z Weight Loss Study.

Disclosures: Research support was provided by Interleukin Genetics. Dr. Nelson had no financial conflicts.

SAN FRANCISCO — Genotypes may identify women likely to lose weight on a low-carbohydrate diet and those likely to do better on a low-fat diet, based on data from 133 participants in the A TO Z Weight Loss Study.

Women with one pattern of single nucleotide polymorphisms (SNPs) lost five times as much weight on the Atkins diet, compared with those who did not have that pattern. Similarly, women with a different SNP genotype lost five times as much weight on the Ornish diet, Mindy Dopler Nelson, Ph.D., of Stanford (Calif.) University, said at a conference sponsored by the American Heart Association.

In the original A TO Z study, 311 women were randomized to one of four popular diets. Ranging on a continuum from low carbohydrate to low fat, they were the Atkins diet, the Zone diet, the LEARN diet, and the Ornish diet. On average, the women lost weight on all four diets; the only significant difference was that they tended to lose somewhat more weight on the Atkins diet than on the Ornish diet (JAMA 2007;297:969-77).

“Within each of the diet groups, there are women who had lost over 15 kg … as well as people who gained 5 kg,” Dr. Nelson said in an interview. “When you look at the averages you don't see the differences, but when you look at each individual participant you see some variability.”

Some time after the conclusion of that trial, a company called Interleukin Genetics approached Stanford researchers and suggested they use the company's proprietary SNP test to assess responders and nonresponders to particular diets. In previous studies, the company had found polymorphisms in three genes—fatty acid binding protein, PPAR-gamma, and the beta₂ adrenergic receptor—that appeared to predict a person's response to diets.

Among the 133 women from the original study who agreed to provide DNA samples from swabs of the inner cheek, 31 had been in the Atkins group, 32 in the Zone group, 34 in the LEARN group, and 36 in the Ornish group. There were no statistically significant baseline differences among the groups in measures such as body mass index, blood pressure, or levels of cholesterol, insulin, and glucose.

The company's test showed that 79 of the women had genotypes designated as low-carb appropriate, and 54 had genotypes designated as low-fat appropriate.

The interaction between genotype and diet was statistically significant, with striking differences among the women in the lowest-carb and lowest-fat diets. Among the women on the Atkins diet, those designated as low-carb appropriate lost an average of just under 6 kg during 12 months, while those designated as low-carb inappropriate lost about 1 kg.

Among the women on the Ornish diet, those designated as low-fat appropriate lost an average of more than 6 kg during 12 months, while those designated as low-fat inappropriate lost an average of about 1.5 kg.

Thus, in each of those groups, women who had been randomized to what was designated as the appropriate diet lost about five times as much weight as those randomized to the apparently inappropriate diet.

Among women on the Zone or LEARN diets, which involve intermediate levels of carbohydrates and fat, women with low-carb and low-fat genotypes did not have statistically significant differences in weight loss.

Dr. Nelson acknowledged that the trial was relatively small, and that the findings need to be confirmed in a larger trial in a more heterogeneous population. Nevertheless, the results do provide some guidance to people who are trying to lose weight, she said.

Taxing Pizza, Soda Proposed to Fight Obesity Epidemic

Taxing soda and restaurant pizza could discourage U.S. adults from consuming those foods, helping them reduce long-term weight gain and insulin resistance, according to a 20-year longitudinal study.

The results bolster the argument that taxes on fast food and sweetened beverages could reduce obesity and improve overall health in the United States, the authors said (Arch. Intern. Med. 2010;170:420-6).

The team pulled data from the Coronary Artery Risk Development in Young Adults (CARDIA) study, which has tracked various factors reported by adults since 1985, including their diets and the prices they paid for foods. The new analysis focused on relationships between dietary changes and the prices reported by 5,115 participants at 0, 7, and 20 years.

The study found that a 10% increase in the price of soda (roughly 20 cents/1-L bottle) resulted in a 3% decline in the probability of consuming soda and a decrease in the amount of soda consumed. Pizza followed a similar trend. A $1 increase in the price of both soda and pizza was associated with even greater changes in total energy intake, body weight, and individuals' homeostasis model assessment of insulin resistance scores, the study found.

As a result, the authors estimated, an 18% tax on soda and fast food could cut energy intake among young to middle-aged adults by about 56 kcal per day. At the population level, this reduction could lead to about 5 fewer pounds in weight gain per person per year “and significant reductions” in the chronic disease risks.

In an editorial, Dr. Mitchell Katz, director of San Francisco department of public health, and Dr. Rajiv Bhatia, medical director of San Francisco's division of occupational and environmental health, advocated that “agricultural subsidies be used to make healthful foods such as locally grown vegetables, fruits, and whole grains less expensive.”

The study was conducted by researchers at the University of North Carolina at Chapel Hill, the University of Alabama at Birmingham, the University of Minnesota, Minneapolis, and the University of Oslo, none of whom reported any conflicts. Dr. Bhatia reported no conflicts, and Dr. Katz received consulting payments from the hospital management company Health Management Associates Inc.

—Jane Anderson

Major Finding: Women randomized to a low-fat or a low-carbohydrate diet considered appropriate (based on a pattern of three single nucleotide polymorphisms) lost about 6 kg, compared with about 1-1.5 kg among women randomized to diets judged as inappropriate.

Data Source: Data from 133 participants in the A TO Z Weight Loss Study.

Disclosures: Research support was provided by Interleukin Genetics. Dr. Nelson had no financial conflicts.

SAN FRANCISCO — Genotypes may identify women likely to lose weight on a low-carbohydrate diet and those likely to do better on a low-fat diet, based on data from 133 participants in the A TO Z Weight Loss Study.

Women with one pattern of single nucleotide polymorphisms (SNPs) lost five times as much weight on the Atkins diet, compared with those who did not have that pattern. Similarly, women with a different SNP genotype lost five times as much weight on the Ornish diet, Mindy Dopler Nelson, Ph.D., of Stanford (Calif.) University, said at a conference sponsored by the American Heart Association.

In the original A TO Z study, 311 women were randomized to one of four popular diets. Ranging on a continuum from low carbohydrate to low fat, they were the Atkins diet, the Zone diet, the LEARN diet, and the Ornish diet. On average, the women lost weight on all four diets; the only significant difference was that they tended to lose somewhat more weight on the Atkins diet than on the Ornish diet (JAMA 2007;297:969-77).

“Within each of the diet groups, there are women who had lost over 15 kg … as well as people who gained 5 kg,” Dr. Nelson said in an interview. “When you look at the averages you don't see the differences, but when you look at each individual participant you see some variability.”

Some time after the conclusion of that trial, a company called Interleukin Genetics approached Stanford researchers and suggested they use the company's proprietary SNP test to assess responders and nonresponders to particular diets. In previous studies, the company had found polymorphisms in three genes—fatty acid binding protein, PPAR-gamma, and the beta₂ adrenergic receptor—that appeared to predict a person's response to diets.

Among the 133 women from the original study who agreed to provide DNA samples from swabs of the inner cheek, 31 had been in the Atkins group, 32 in the Zone group, 34 in the LEARN group, and 36 in the Ornish group. There were no statistically significant baseline differences among the groups in measures such as body mass index, blood pressure, or levels of cholesterol, insulin, and glucose.

The company's test showed that 79 of the women had genotypes designated as low-carb appropriate, and 54 had genotypes designated as low-fat appropriate.

The interaction between genotype and diet was statistically significant, with striking differences among the women in the lowest-carb and lowest-fat diets. Among the women on the Atkins diet, those designated as low-carb appropriate lost an average of just under 6 kg during 12 months, while those designated as low-carb inappropriate lost about 1 kg.

Among the women on the Ornish diet, those designated as low-fat appropriate lost an average of more than 6 kg during 12 months, while those designated as low-fat inappropriate lost an average of about 1.5 kg.

Thus, in each of those groups, women who had been randomized to what was designated as the appropriate diet lost about five times as much weight as those randomized to the apparently inappropriate diet.

Among women on the Zone or LEARN diets, which involve intermediate levels of carbohydrates and fat, women with low-carb and low-fat genotypes did not have statistically significant differences in weight loss.

Dr. Nelson acknowledged that the trial was relatively small, and that the findings need to be confirmed in a larger trial in a more heterogeneous population. Nevertheless, the results do provide some guidance to people who are trying to lose weight, she said.

Taxing Pizza, Soda Proposed to Fight Obesity Epidemic

Taxing soda and restaurant pizza could discourage U.S. adults from consuming those foods, helping them reduce long-term weight gain and insulin resistance, according to a 20-year longitudinal study.

The results bolster the argument that taxes on fast food and sweetened beverages could reduce obesity and improve overall health in the United States, the authors said (Arch. Intern. Med. 2010;170:420-6).

The team pulled data from the Coronary Artery Risk Development in Young Adults (CARDIA) study, which has tracked various factors reported by adults since 1985, including their diets and the prices they paid for foods. The new analysis focused on relationships between dietary changes and the prices reported by 5,115 participants at 0, 7, and 20 years.

The study found that a 10% increase in the price of soda (roughly 20 cents/1-L bottle) resulted in a 3% decline in the probability of consuming soda and a decrease in the amount of soda consumed. Pizza followed a similar trend. A $1 increase in the price of both soda and pizza was associated with even greater changes in total energy intake, body weight, and individuals' homeostasis model assessment of insulin resistance scores, the study found.

As a result, the authors estimated, an 18% tax on soda and fast food could cut energy intake among young to middle-aged adults by about 56 kcal per day. At the population level, this reduction could lead to about 5 fewer pounds in weight gain per person per year “and significant reductions” in the chronic disease risks.

In an editorial, Dr. Mitchell Katz, director of San Francisco department of public health, and Dr. Rajiv Bhatia, medical director of San Francisco's division of occupational and environmental health, advocated that “agricultural subsidies be used to make healthful foods such as locally grown vegetables, fruits, and whole grains less expensive.”

The study was conducted by researchers at the University of North Carolina at Chapel Hill, the University of Alabama at Birmingham, the University of Minnesota, Minneapolis, and the University of Oslo, none of whom reported any conflicts. Dr. Bhatia reported no conflicts, and Dr. Katz received consulting payments from the hospital management company Health Management Associates Inc.

—Jane Anderson

Major Finding: Women randomized to a low-fat or a low-carbohydrate diet considered appropriate (based on a pattern of three single nucleotide polymorphisms) lost about 6 kg, compared with about 1-1.5 kg among women randomized to diets judged as inappropriate.

Data Source: Data from 133 participants in the A TO Z Weight Loss Study.

Disclosures: Research support was provided by Interleukin Genetics. Dr. Nelson had no financial conflicts.

SAN FRANCISCO — Genotypes may identify women likely to lose weight on a low-carbohydrate diet and those likely to do better on a low-fat diet, based on data from 133 participants in the A TO Z Weight Loss Study.

Women with one pattern of single nucleotide polymorphisms (SNPs) lost five times as much weight on the Atkins diet, compared with those who did not have that pattern. Similarly, women with a different SNP genotype lost five times as much weight on the Ornish diet, Mindy Dopler Nelson, Ph.D., of Stanford (Calif.) University, said at a conference sponsored by the American Heart Association.

In the original A TO Z study, 311 women were randomized to one of four popular diets. Ranging on a continuum from low carbohydrate to low fat, they were the Atkins diet, the Zone diet, the LEARN diet, and the Ornish diet. On average, the women lost weight on all four diets; the only significant difference was that they tended to lose somewhat more weight on the Atkins diet than on the Ornish diet (JAMA 2007;297:969-77).

“Within each of the diet groups, there are women who had lost over 15 kg … as well as people who gained 5 kg,” Dr. Nelson said in an interview. “When you look at the averages you don't see the differences, but when you look at each individual participant you see some variability.”

Some time after the conclusion of that trial, a company called Interleukin Genetics approached Stanford researchers and suggested they use the company's proprietary SNP test to assess responders and nonresponders to particular diets. In previous studies, the company had found polymorphisms in three genes—fatty acid binding protein, PPAR-gamma, and the beta₂ adrenergic receptor—that appeared to predict a person's response to diets.

Among the 133 women from the original study who agreed to provide DNA samples from swabs of the inner cheek, 31 had been in the Atkins group, 32 in the Zone group, 34 in the LEARN group, and 36 in the Ornish group. There were no statistically significant baseline differences among the groups in measures such as body mass index, blood pressure, or levels of cholesterol, insulin, and glucose.

The company's test showed that 79 of the women had genotypes designated as low-carb appropriate, and 54 had genotypes designated as low-fat appropriate.

The interaction between genotype and diet was statistically significant, with striking differences among the women in the lowest-carb and lowest-fat diets. Among the women on the Atkins diet, those designated as low-carb appropriate lost an average of just under 6 kg during 12 months, while those designated as low-carb inappropriate lost about 1 kg.

Among the women on the Ornish diet, those designated as low-fat appropriate lost an average of more than 6 kg during 12 months, while those designated as low-fat inappropriate lost an average of about 1.5 kg.

Thus, in each of those groups, women who had been randomized to what was designated as the appropriate diet lost about five times as much weight as those randomized to the apparently inappropriate diet.

Among women on the Zone or LEARN diets, which involve intermediate levels of carbohydrates and fat, women with low-carb and low-fat genotypes did not have statistically significant differences in weight loss.

Dr. Nelson acknowledged that the trial was relatively small, and that the findings need to be confirmed in a larger trial in a more heterogeneous population. Nevertheless, the results do provide some guidance to people who are trying to lose weight, she said.

Taxing Pizza, Soda Proposed to Fight Obesity Epidemic

Taxing soda and restaurant pizza could discourage U.S. adults from consuming those foods, helping them reduce long-term weight gain and insulin resistance, according to a 20-year longitudinal study.

The results bolster the argument that taxes on fast food and sweetened beverages could reduce obesity and improve overall health in the United States, the authors said (Arch. Intern. Med. 2010;170:420-6).

The team pulled data from the Coronary Artery Risk Development in Young Adults (CARDIA) study, which has tracked various factors reported by adults since 1985, including their diets and the prices they paid for foods. The new analysis focused on relationships between dietary changes and the prices reported by 5,115 participants at 0, 7, and 20 years.

The study found that a 10% increase in the price of soda (roughly 20 cents/1-L bottle) resulted in a 3% decline in the probability of consuming soda and a decrease in the amount of soda consumed. Pizza followed a similar trend. A $1 increase in the price of both soda and pizza was associated with even greater changes in total energy intake, body weight, and individuals' homeostasis model assessment of insulin resistance scores, the study found.

As a result, the authors estimated, an 18% tax on soda and fast food could cut energy intake among young to middle-aged adults by about 56 kcal per day. At the population level, this reduction could lead to about 5 fewer pounds in weight gain per person per year “and significant reductions” in the chronic disease risks.

In an editorial, Dr. Mitchell Katz, director of San Francisco department of public health, and Dr. Rajiv Bhatia, medical director of San Francisco's division of occupational and environmental health, advocated that “agricultural subsidies be used to make healthful foods such as locally grown vegetables, fruits, and whole grains less expensive.”

The study was conducted by researchers at the University of North Carolina at Chapel Hill, the University of Alabama at Birmingham, the University of Minnesota, Minneapolis, and the University of Oslo, none of whom reported any conflicts. Dr. Bhatia reported no conflicts, and Dr. Katz received consulting payments from the hospital management company Health Management Associates Inc.

—Jane Anderson

Major Finding: HIV infection is associated with an 80% increase in the risk of incident lung cancer after controlling for smoking and other known risk factors.

Data Source: Data on 3,707 HIV-positive individuals and 5,980 HIV-negative controls.

Disclosures: The researchers said they had no disclosures.

The increased risk for lung cancer associated with HIV infection appears to be more modest than previous studies had indicated.

The finding is based on a study of nearly 14,000 veterans who were followed for a median of 8 years. After controlling for smoking status, age, chronic obstructive pulmonary disease, and race and ethnicity, HIV infection was associated with a 1.8-fold increase in the risk of lung cancer, Dr. Keith Sigel said at the Conference on Retroviruses and Opportunistic Infections.

Previous studies had suggested that HIV infection was associated with a 2.5-fold to 3.6-fold increase in the risk of lung cancer.

Dr. Sigel of the Mount Sinai School of Medicine, New York, and his colleagues used data from the Veterans Aging Cohort Study (VACS), a “virtual” cohort of 33,420 HIV-positive individuals and 66,840 HIV-negative controls matched by age, race, gender, and site.

The VACS data set does not include information on smoking status. In order to control for smoking status, the investigators combined the VACS data set with data from the 1999 Large Health Survey of Veteran Enrollees. The merged dataset included 3,707 HIV-infected individuals and 9,980 healthy controls.

The median age of individuals in the cohort was 48 years, and 98% were male. Minorities were well represented: 41% were white, 39% were black, 10% were Hispanic, and 10% were of other races or ethnicities.

There were some significant baseline differences between HIV-infected and uninfected individuals in the cohort. HIV-infected individuals were more likely to be current daily smokers (32% vs. 28%), were more likely to be drug abusers (16% vs. 10%), and were more likely to have lung cancer at the beginning of the study (16 vs. 38 individuals). Individuals with prevalent lung cancer were excluded from the study.

The unadjusted absolute incidence of lung cancer was 26 cases/10,000 person-years among individuals who were HIV positive and 15 cases/10,000 person-years among individuals who were HIV negative.

Although the observed 80% increase in the risk of lung cancer associated with HIV infection was statistically significant, several other independent predictors conferred much larger risks of lung cancer. Current daily smoking was associated with a 9.8-fold increase in risk, and current occasional smoking was associated with a 3.4-fold increase in risk. Also, Hispanic ethnicity was associated with a 60% decrease in the risk of incident lung cancer.

“Our results represent a more modest risk than previous adjusted analyses, which may reflect differences in methods in our case identification, the presence of matched HIV controls for comparison within our cohort, or potentially greater precision allowed by our sample size,” Dr. Sigel said.

Major Finding: HIV infection is associated with an 80% increase in the risk of incident lung cancer after controlling for smoking and other known risk factors.

Data Source: Data on 3,707 HIV-positive individuals and 5,980 HIV-negative controls.

Disclosures: The researchers said they had no disclosures.

The increased risk for lung cancer associated with HIV infection appears to be more modest than previous studies had indicated.

The finding is based on a study of nearly 14,000 veterans who were followed for a median of 8 years. After controlling for smoking status, age, chronic obstructive pulmonary disease, and race and ethnicity, HIV infection was associated with a 1.8-fold increase in the risk of lung cancer, Dr. Keith Sigel said at the Conference on Retroviruses and Opportunistic Infections.

Previous studies had suggested that HIV infection was associated with a 2.5-fold to 3.6-fold increase in the risk of lung cancer.

Dr. Sigel of the Mount Sinai School of Medicine, New York, and his colleagues used data from the Veterans Aging Cohort Study (VACS), a “virtual” cohort of 33,420 HIV-positive individuals and 66,840 HIV-negative controls matched by age, race, gender, and site.

The VACS data set does not include information on smoking status. In order to control for smoking status, the investigators combined the VACS data set with data from the 1999 Large Health Survey of Veteran Enrollees. The merged dataset included 3,707 HIV-infected individuals and 9,980 healthy controls.

The median age of individuals in the cohort was 48 years, and 98% were male. Minorities were well represented: 41% were white, 39% were black, 10% were Hispanic, and 10% were of other races or ethnicities.

There were some significant baseline differences between HIV-infected and uninfected individuals in the cohort. HIV-infected individuals were more likely to be current daily smokers (32% vs. 28%), were more likely to be drug abusers (16% vs. 10%), and were more likely to have lung cancer at the beginning of the study (16 vs. 38 individuals). Individuals with prevalent lung cancer were excluded from the study.

The unadjusted absolute incidence of lung cancer was 26 cases/10,000 person-years among individuals who were HIV positive and 15 cases/10,000 person-years among individuals who were HIV negative.

Although the observed 80% increase in the risk of lung cancer associated with HIV infection was statistically significant, several other independent predictors conferred much larger risks of lung cancer. Current daily smoking was associated with a 9.8-fold increase in risk, and current occasional smoking was associated with a 3.4-fold increase in risk. Also, Hispanic ethnicity was associated with a 60% decrease in the risk of incident lung cancer.

“Our results represent a more modest risk than previous adjusted analyses, which may reflect differences in methods in our case identification, the presence of matched HIV controls for comparison within our cohort, or potentially greater precision allowed by our sample size,” Dr. Sigel said.

Major Finding: HIV infection is associated with an 80% increase in the risk of incident lung cancer after controlling for smoking and other known risk factors.

Data Source: Data on 3,707 HIV-positive individuals and 5,980 HIV-negative controls.

Disclosures: The researchers said they had no disclosures.

The increased risk for lung cancer associated with HIV infection appears to be more modest than previous studies had indicated.

The finding is based on a study of nearly 14,000 veterans who were followed for a median of 8 years. After controlling for smoking status, age, chronic obstructive pulmonary disease, and race and ethnicity, HIV infection was associated with a 1.8-fold increase in the risk of lung cancer, Dr. Keith Sigel said at the Conference on Retroviruses and Opportunistic Infections.

Previous studies had suggested that HIV infection was associated with a 2.5-fold to 3.6-fold increase in the risk of lung cancer.

Dr. Sigel of the Mount Sinai School of Medicine, New York, and his colleagues used data from the Veterans Aging Cohort Study (VACS), a “virtual” cohort of 33,420 HIV-positive individuals and 66,840 HIV-negative controls matched by age, race, gender, and site.

The VACS data set does not include information on smoking status. In order to control for smoking status, the investigators combined the VACS data set with data from the 1999 Large Health Survey of Veteran Enrollees. The merged dataset included 3,707 HIV-infected individuals and 9,980 healthy controls.

The median age of individuals in the cohort was 48 years, and 98% were male. Minorities were well represented: 41% were white, 39% were black, 10% were Hispanic, and 10% were of other races or ethnicities.

There were some significant baseline differences between HIV-infected and uninfected individuals in the cohort. HIV-infected individuals were more likely to be current daily smokers (32% vs. 28%), were more likely to be drug abusers (16% vs. 10%), and were more likely to have lung cancer at the beginning of the study (16 vs. 38 individuals). Individuals with prevalent lung cancer were excluded from the study.

The unadjusted absolute incidence of lung cancer was 26 cases/10,000 person-years among individuals who were HIV positive and 15 cases/10,000 person-years among individuals who were HIV negative.

Although the observed 80% increase in the risk of lung cancer associated with HIV infection was statistically significant, several other independent predictors conferred much larger risks of lung cancer. Current daily smoking was associated with a 9.8-fold increase in risk, and current occasional smoking was associated with a 3.4-fold increase in risk. Also, Hispanic ethnicity was associated with a 60% decrease in the risk of incident lung cancer.

“Our results represent a more modest risk than previous adjusted analyses, which may reflect differences in methods in our case identification, the presence of matched HIV controls for comparison within our cohort, or potentially greater precision allowed by our sample size,” Dr. Sigel said.

Major Finding: Among people with AIDS and non–AIDS-defining cancers, cancer-attributable mortality rose from 72% to 87% between 1980 and 2006.

Data Source: Records of 372,364 people with AIDS along with links to corresponding cancer registry records.

Disclosures: The researchers had no conflicts to report.

Although AIDS-related mortality has declined since highly active antiretroviral therapy became widely available in the mid-1990s, there has been a corresponding increase in the incidence of cancer among people with AIDS, according to a study of more than 300,000 individuals.

The rates of the AIDS-defining cancers—Kaposi's sarcoma, non-Hodgkin's lymphoma, and cervical cancer—remain high in people with AIDS compared with the general population, but people with AIDS are also at high risk of several types of non–AIDS-defining cancers.

“Among those who die with AIDS and cancer, cancer now accounts for the vast majority of all deaths,” said Edgar P. Simard, Ph.D., of the National Cancer Institute. “And in the entire population, non–AIDS-defining cancers represent an increasing fraction of all deaths.”

Dr. Simard and his colleagues used data from 372,364 people diagnosed with AIDS in the United States between 1980 and 2006 and linked those with corresponding cancer registry records. The investigators divided their analyses into three eras: 1980–1989, when there was little in the way of effective AIDS treatment; 1990–1995, when one- and two-drug regimens were typical; and 1996–2006, when highly active antiretroviral therapy (HAART) became widely used.

Since numerous studies have already been done on cancer in the 2 years following AIDS diagnosis, the investigators focused on cancer risk 3–5 years after AIDS onset, Dr. Simard said at the Conference on Retroviruses and Opportunistic Infections.

As expected, the incidence of AIDS-defining cancers was very high during years 3–5 after diagnosis. Compared with the general population, people with AIDS had 5,321 times the risk of developing Kaposi's sarcoma, 32 times the risk of developing non-Hodgkin's lymphoma, and 5.6 times the risk of developing cervical cancer, he reported.

People with AIDS also had significant increases in the risk of developing four different non–AIDS-defining cancers. They had a 27-fold increase in the risk of anal cancer, a 9.1-fold increase in the risk of Hodgkin's lymphoma, a 3.7-fold increase in the risk of liver cancer, and a 3.0-fold increase in the risk of lung cancer. Overall, people with AIDS had a statistically significant 70% increase in the risk of developing any non–AIDS-defining cancer.

The cumulative incidence of AIDS-defining cancers declined significantly. At 60 months following diagnosis, the cumulative incidence was 8.7% for the 1980–1989 era, 6.4% for 1990–1995, and 2.1% for 1996–2006, Dr. Simard said.

Yet the cumulative incidence of the four non–AIDS-defining cancers increased significantly. The 60-month cumulative incidence of lung cancer rose from 0.14% in 1980–1989 to 0.28% in 1990–1995 and to 0.37% in 1996–2006. The cumulative incidence of Hodgkin's lymphoma rose from 0.04% in 1980–1989 to 0.10% in 1990–1995 to 0.17% in the 1996–2006 era, he said.

The cancer-attributable mortality of both AIDS-defining and non–AIDS-defining cancers increased significantly from the earliest to the latest treatment eras. Among AIDS-defining cancers, the cancer-attributable mortality rose from 69% to 88%, and among non–AIDS-defining cancers, the cancer-attributable mortality rose from 72% to 87%.

Cancer prevention and treatment “will become increasingly important as survival from AIDS increases and the population continues to age,” he said.

Major Finding: Among people with AIDS and non–AIDS-defining cancers, cancer-attributable mortality rose from 72% to 87% between 1980 and 2006.

Data Source: Records of 372,364 people with AIDS along with links to corresponding cancer registry records.

Disclosures: The researchers had no conflicts to report.

Although AIDS-related mortality has declined since highly active antiretroviral therapy became widely available in the mid-1990s, there has been a corresponding increase in the incidence of cancer among people with AIDS, according to a study of more than 300,000 individuals.

The rates of the AIDS-defining cancers—Kaposi's sarcoma, non-Hodgkin's lymphoma, and cervical cancer—remain high in people with AIDS compared with the general population, but people with AIDS are also at high risk of several types of non–AIDS-defining cancers.

“Among those who die with AIDS and cancer, cancer now accounts for the vast majority of all deaths,” said Edgar P. Simard, Ph.D., of the National Cancer Institute. “And in the entire population, non–AIDS-defining cancers represent an increasing fraction of all deaths.”

Dr. Simard and his colleagues used data from 372,364 people diagnosed with AIDS in the United States between 1980 and 2006 and linked those with corresponding cancer registry records. The investigators divided their analyses into three eras: 1980–1989, when there was little in the way of effective AIDS treatment; 1990–1995, when one- and two-drug regimens were typical; and 1996–2006, when highly active antiretroviral therapy (HAART) became widely used.

Since numerous studies have already been done on cancer in the 2 years following AIDS diagnosis, the investigators focused on cancer risk 3–5 years after AIDS onset, Dr. Simard said at the Conference on Retroviruses and Opportunistic Infections.

As expected, the incidence of AIDS-defining cancers was very high during years 3–5 after diagnosis. Compared with the general population, people with AIDS had 5,321 times the risk of developing Kaposi's sarcoma, 32 times the risk of developing non-Hodgkin's lymphoma, and 5.6 times the risk of developing cervical cancer, he reported.

People with AIDS also had significant increases in the risk of developing four different non–AIDS-defining cancers. They had a 27-fold increase in the risk of anal cancer, a 9.1-fold increase in the risk of Hodgkin's lymphoma, a 3.7-fold increase in the risk of liver cancer, and a 3.0-fold increase in the risk of lung cancer. Overall, people with AIDS had a statistically significant 70% increase in the risk of developing any non–AIDS-defining cancer.

The cumulative incidence of AIDS-defining cancers declined significantly. At 60 months following diagnosis, the cumulative incidence was 8.7% for the 1980–1989 era, 6.4% for 1990–1995, and 2.1% for 1996–2006, Dr. Simard said.

Yet the cumulative incidence of the four non–AIDS-defining cancers increased significantly. The 60-month cumulative incidence of lung cancer rose from 0.14% in 1980–1989 to 0.28% in 1990–1995 and to 0.37% in 1996–2006. The cumulative incidence of Hodgkin's lymphoma rose from 0.04% in 1980–1989 to 0.10% in 1990–1995 to 0.17% in the 1996–2006 era, he said.

The cancer-attributable mortality of both AIDS-defining and non–AIDS-defining cancers increased significantly from the earliest to the latest treatment eras. Among AIDS-defining cancers, the cancer-attributable mortality rose from 69% to 88%, and among non–AIDS-defining cancers, the cancer-attributable mortality rose from 72% to 87%.

Cancer prevention and treatment “will become increasingly important as survival from AIDS increases and the population continues to age,” he said.

Major Finding: Among people with AIDS and non–AIDS-defining cancers, cancer-attributable mortality rose from 72% to 87% between 1980 and 2006.

Data Source: Records of 372,364 people with AIDS along with links to corresponding cancer registry records.

Disclosures: The researchers had no conflicts to report.

Although AIDS-related mortality has declined since highly active antiretroviral therapy became widely available in the mid-1990s, there has been a corresponding increase in the incidence of cancer among people with AIDS, according to a study of more than 300,000 individuals.

The rates of the AIDS-defining cancers—Kaposi's sarcoma, non-Hodgkin's lymphoma, and cervical cancer—remain high in people with AIDS compared with the general population, but people with AIDS are also at high risk of several types of non–AIDS-defining cancers.

“Among those who die with AIDS and cancer, cancer now accounts for the vast majority of all deaths,” said Edgar P. Simard, Ph.D., of the National Cancer Institute. “And in the entire population, non–AIDS-defining cancers represent an increasing fraction of all deaths.”

Dr. Simard and his colleagues used data from 372,364 people diagnosed with AIDS in the United States between 1980 and 2006 and linked those with corresponding cancer registry records. The investigators divided their analyses into three eras: 1980–1989, when there was little in the way of effective AIDS treatment; 1990–1995, when one- and two-drug regimens were typical; and 1996–2006, when highly active antiretroviral therapy (HAART) became widely used.

Since numerous studies have already been done on cancer in the 2 years following AIDS diagnosis, the investigators focused on cancer risk 3–5 years after AIDS onset, Dr. Simard said at the Conference on Retroviruses and Opportunistic Infections.

As expected, the incidence of AIDS-defining cancers was very high during years 3–5 after diagnosis. Compared with the general population, people with AIDS had 5,321 times the risk of developing Kaposi's sarcoma, 32 times the risk of developing non-Hodgkin's lymphoma, and 5.6 times the risk of developing cervical cancer, he reported.

People with AIDS also had significant increases in the risk of developing four different non–AIDS-defining cancers. They had a 27-fold increase in the risk of anal cancer, a 9.1-fold increase in the risk of Hodgkin's lymphoma, a 3.7-fold increase in the risk of liver cancer, and a 3.0-fold increase in the risk of lung cancer. Overall, people with AIDS had a statistically significant 70% increase in the risk of developing any non–AIDS-defining cancer.

The cumulative incidence of AIDS-defining cancers declined significantly. At 60 months following diagnosis, the cumulative incidence was 8.7% for the 1980–1989 era, 6.4% for 1990–1995, and 2.1% for 1996–2006, Dr. Simard said.

Yet the cumulative incidence of the four non–AIDS-defining cancers increased significantly. The 60-month cumulative incidence of lung cancer rose from 0.14% in 1980–1989 to 0.28% in 1990–1995 and to 0.37% in 1996–2006. The cumulative incidence of Hodgkin's lymphoma rose from 0.04% in 1980–1989 to 0.10% in 1990–1995 to 0.17% in the 1996–2006 era, he said.

The cancer-attributable mortality of both AIDS-defining and non–AIDS-defining cancers increased significantly from the earliest to the latest treatment eras. Among AIDS-defining cancers, the cancer-attributable mortality rose from 69% to 88%, and among non–AIDS-defining cancers, the cancer-attributable mortality rose from 72% to 87%.

Cancer prevention and treatment “will become increasingly important as survival from AIDS increases and the population continues to age,” he said.

Major Finding: Women who were randomized to a low-fat or a low-carbohydrate diet considered appropriate (based on a pattern of three single nucleotide polymorphisms) lost about 6 kg vs. about 1-1.5 kg among women randomized to diets judged as inappropriate.

Data Source: Data from 133 women who participated in the A TO Z Weight Loss Study.

Disclosures: Research support was provided by Interleukin Genetics. Dr. Nelson had no financial conflicts.

SAN FRANCISCO — Genotypes may identify women who are likely to lose weight on a low-carbohydrate diet and those who are likely to do better on a low-fat diet, based on data from 133 participants in the A TO Z Weight Loss Study.

The women who had one pattern of single nucleotide polymorphisms (SNPs) lost about five times as much weight on the Atkins diet, compared with the women who did not have that pattern. Similarly, the women who had a different SNP genotype lost about five times as much weight on the Ornish diet as did the women who did not have that pattern, Mindy Dopler Nelson, Ph.D., of Stanford (Calif.) University, reported at a conference sponsored by the American Heart Association.

In the original A TO Z study, 311 women were randomized to one of four popular weight-loss diets. Ranging on a continuum from low carbohydrate to low fat, they were the Atkins diet, the Zone diet, the LEARN diet, and the Ornish diet. On average, the women lost weight on all four of the diets; the only significant difference was that they tended to lose somewhat more weight on the Atkins diet than on the Ornish diet (JAMA 2007;297:969-77).

“Within each of the diet groups, there are women who had lost over 15 kg … as well as people who gained 5 kg,” Dr. Nelson explained in an interview. “When you look at the averages you don't see the differences, but when you look at each individual participant you see some variability.”

Some time after the conclusion of the A TO Z study, a company called Interleukin Genetics approached Stanford researchers and suggested that they use the company's proprietary SNP test to assess responders and nonresponders to particular diets. In previous studies, the company had found polymorphisms in three genes—those coding for fatty acid binding protein, PPAR-gamma, and the beta-2 adrenergic receptor—that appeared to predict a person's responses to various weight-loss diets.

Among the 133 women from the original study who agreed to provide DNA samples from swabs of the inner cheek, 31 had been in the Atkins group, 32 in the Zone group, 34 in the LEARN group, and 36 in the Ornish group. There were no statistically significant baseline differences among the four groups in measures such as body mass index and blood pressure, or in levels of cholesterol, insulin, and glucose, Dr. Nelson reported.

The company's test showed that 79 of the women had genotypes that were designated as low-carb appropriate, and 54 had genotypes that were designated as low-fat appropriate.

The interaction between genotype and diet was statistically significant, with striking differences among the women in the lowest-carb and lowest-fat diets. Among the women on the Atkins diet, those who were designated as low-carb appropriate lost an average of just under 6 kg during 12 months, while those who were designated as low-carb inappropriate lost about 1 kg.

Among the women on the Ornish diet, those who were designated as low-fat appropriate lost an average of more than 6 kg during 12 months, while those who were designated as low-fat inappropriate lost an average of about 1.5 kg.

Thus, in each of those groups, women who had been randomized to what was designated as the appropriate diet lost about five times as much weight as those who had been randomized to the apparently inappropriate diet, Dr. Nelson reported.

Among women on the Zone or LEARN diets, which involve intermediate levels of carbohydrates and fat, women who had low-carb and low-fat genotypes did not have statistically significant differences in weight loss.

Dr. Nelson acknowledged that the trial was relatively small, and that the findings need to be confirmed in a larger trial in a more heterogeneous population. Nevertheless, the results do provide some guidance to people who are trying to lose weight, she said.

“I would suggest that if somebody is discouraged by the weight loss that they're having on whatever particular diet they're following, they may just want to consider changing the distribution of their macronutrients,” Dr. Nelson said. “So maybe if you're doing a higher-carbohydrate diet and you're not seeing your weight loss, give up some of the more processed carbohydrates, keep the healthier ones in there, and see if shifting to the lower-carbohydrate diet will help with weight loss.”

Major Finding: Women who were randomized to a low-fat or a low-carbohydrate diet considered appropriate (based on a pattern of three single nucleotide polymorphisms) lost about 6 kg vs. about 1-1.5 kg among women randomized to diets judged as inappropriate.

Data Source: Data from 133 women who participated in the A TO Z Weight Loss Study.

Disclosures: Research support was provided by Interleukin Genetics. Dr. Nelson had no financial conflicts.

SAN FRANCISCO — Genotypes may identify women who are likely to lose weight on a low-carbohydrate diet and those who are likely to do better on a low-fat diet, based on data from 133 participants in the A TO Z Weight Loss Study.

The women who had one pattern of single nucleotide polymorphisms (SNPs) lost about five times as much weight on the Atkins diet, compared with the women who did not have that pattern. Similarly, the women who had a different SNP genotype lost about five times as much weight on the Ornish diet as did the women who did not have that pattern, Mindy Dopler Nelson, Ph.D., of Stanford (Calif.) University, reported at a conference sponsored by the American Heart Association.

In the original A TO Z study, 311 women were randomized to one of four popular weight-loss diets. Ranging on a continuum from low carbohydrate to low fat, they were the Atkins diet, the Zone diet, the LEARN diet, and the Ornish diet. On average, the women lost weight on all four of the diets; the only significant difference was that they tended to lose somewhat more weight on the Atkins diet than on the Ornish diet (JAMA 2007;297:969-77).

“Within each of the diet groups, there are women who had lost over 15 kg … as well as people who gained 5 kg,” Dr. Nelson explained in an interview. “When you look at the averages you don't see the differences, but when you look at each individual participant you see some variability.”

Some time after the conclusion of the A TO Z study, a company called Interleukin Genetics approached Stanford researchers and suggested that they use the company's proprietary SNP test to assess responders and nonresponders to particular diets. In previous studies, the company had found polymorphisms in three genes—those coding for fatty acid binding protein, PPAR-gamma, and the beta-2 adrenergic receptor—that appeared to predict a person's responses to various weight-loss diets.

Among the 133 women from the original study who agreed to provide DNA samples from swabs of the inner cheek, 31 had been in the Atkins group, 32 in the Zone group, 34 in the LEARN group, and 36 in the Ornish group. There were no statistically significant baseline differences among the four groups in measures such as body mass index and blood pressure, or in levels of cholesterol, insulin, and glucose, Dr. Nelson reported.

The company's test showed that 79 of the women had genotypes that were designated as low-carb appropriate, and 54 had genotypes that were designated as low-fat appropriate.

The interaction between genotype and diet was statistically significant, with striking differences among the women in the lowest-carb and lowest-fat diets. Among the women on the Atkins diet, those who were designated as low-carb appropriate lost an average of just under 6 kg during 12 months, while those who were designated as low-carb inappropriate lost about 1 kg.

Among the women on the Ornish diet, those who were designated as low-fat appropriate lost an average of more than 6 kg during 12 months, while those who were designated as low-fat inappropriate lost an average of about 1.5 kg.

Thus, in each of those groups, women who had been randomized to what was designated as the appropriate diet lost about five times as much weight as those who had been randomized to the apparently inappropriate diet, Dr. Nelson reported.

Among women on the Zone or LEARN diets, which involve intermediate levels of carbohydrates and fat, women who had low-carb and low-fat genotypes did not have statistically significant differences in weight loss.

Dr. Nelson acknowledged that the trial was relatively small, and that the findings need to be confirmed in a larger trial in a more heterogeneous population. Nevertheless, the results do provide some guidance to people who are trying to lose weight, she said.

“I would suggest that if somebody is discouraged by the weight loss that they're having on whatever particular diet they're following, they may just want to consider changing the distribution of their macronutrients,” Dr. Nelson said. “So maybe if you're doing a higher-carbohydrate diet and you're not seeing your weight loss, give up some of the more processed carbohydrates, keep the healthier ones in there, and see if shifting to the lower-carbohydrate diet will help with weight loss.”

Major Finding: Women who were randomized to a low-fat or a low-carbohydrate diet considered appropriate (based on a pattern of three single nucleotide polymorphisms) lost about 6 kg vs. about 1-1.5 kg among women randomized to diets judged as inappropriate.

Data Source: Data from 133 women who participated in the A TO Z Weight Loss Study.

Disclosures: Research support was provided by Interleukin Genetics. Dr. Nelson had no financial conflicts.

SAN FRANCISCO — Genotypes may identify women who are likely to lose weight on a low-carbohydrate diet and those who are likely to do better on a low-fat diet, based on data from 133 participants in the A TO Z Weight Loss Study.

The women who had one pattern of single nucleotide polymorphisms (SNPs) lost about five times as much weight on the Atkins diet, compared with the women who did not have that pattern. Similarly, the women who had a different SNP genotype lost about five times as much weight on the Ornish diet as did the women who did not have that pattern, Mindy Dopler Nelson, Ph.D., of Stanford (Calif.) University, reported at a conference sponsored by the American Heart Association.

In the original A TO Z study, 311 women were randomized to one of four popular weight-loss diets. Ranging on a continuum from low carbohydrate to low fat, they were the Atkins diet, the Zone diet, the LEARN diet, and the Ornish diet. On average, the women lost weight on all four of the diets; the only significant difference was that they tended to lose somewhat more weight on the Atkins diet than on the Ornish diet (JAMA 2007;297:969-77).

“Within each of the diet groups, there are women who had lost over 15 kg … as well as people who gained 5 kg,” Dr. Nelson explained in an interview. “When you look at the averages you don't see the differences, but when you look at each individual participant you see some variability.”

Some time after the conclusion of the A TO Z study, a company called Interleukin Genetics approached Stanford researchers and suggested that they use the company's proprietary SNP test to assess responders and nonresponders to particular diets. In previous studies, the company had found polymorphisms in three genes—those coding for fatty acid binding protein, PPAR-gamma, and the beta-2 adrenergic receptor—that appeared to predict a person's responses to various weight-loss diets.

Among the 133 women from the original study who agreed to provide DNA samples from swabs of the inner cheek, 31 had been in the Atkins group, 32 in the Zone group, 34 in the LEARN group, and 36 in the Ornish group. There were no statistically significant baseline differences among the four groups in measures such as body mass index and blood pressure, or in levels of cholesterol, insulin, and glucose, Dr. Nelson reported.

The company's test showed that 79 of the women had genotypes that were designated as low-carb appropriate, and 54 had genotypes that were designated as low-fat appropriate.

The interaction between genotype and diet was statistically significant, with striking differences among the women in the lowest-carb and lowest-fat diets. Among the women on the Atkins diet, those who were designated as low-carb appropriate lost an average of just under 6 kg during 12 months, while those who were designated as low-carb inappropriate lost about 1 kg.

Among the women on the Ornish diet, those who were designated as low-fat appropriate lost an average of more than 6 kg during 12 months, while those who were designated as low-fat inappropriate lost an average of about 1.5 kg.

Thus, in each of those groups, women who had been randomized to what was designated as the appropriate diet lost about five times as much weight as those who had been randomized to the apparently inappropriate diet, Dr. Nelson reported.

Among women on the Zone or LEARN diets, which involve intermediate levels of carbohydrates and fat, women who had low-carb and low-fat genotypes did not have statistically significant differences in weight loss.

Dr. Nelson acknowledged that the trial was relatively small, and that the findings need to be confirmed in a larger trial in a more heterogeneous population. Nevertheless, the results do provide some guidance to people who are trying to lose weight, she said.

“I would suggest that if somebody is discouraged by the weight loss that they're having on whatever particular diet they're following, they may just want to consider changing the distribution of their macronutrients,” Dr. Nelson said. “So maybe if you're doing a higher-carbohydrate diet and you're not seeing your weight loss, give up some of the more processed carbohydrates, keep the healthier ones in there, and see if shifting to the lower-carbohydrate diet will help with weight loss.”