Irregular sleep duration was associated with a higher risk for diabetes in middle-aged to older adults in a new UK Biobank study.
The analysis of more than 84,000 participants with 7-day accelerometry data suggested that individuals with the most irregular sleep duration patterns had a 34% higher risk for diabetes compared with their peers who had more consistent sleep patterns.
“It’s recommended to have 7-9 hours of nightly sleep, but what is not considered much in policy guidelines or at the clinical level is how regularly that’s needed,” Sina Kianersi, PhD, of Brigham and Women’s Hospital in Boston, Massachusetts, said in an interview. “What our study added is that it’s not just the duration but keeping it consistent. Patients can reduce their risk of diabetes by maintaining their 7-9 hours of sleep, not just for 1 night but throughout life.”
The study was published online in Diabetes Care.
Modifiable Lifestyle Factor
Researchers analyzed data from 84,421 UK Biobank participants who were free of diabetes when they provided accelerometer data in 2013-2015 and who were followed for a median of 7.5 years (622,080 person-years).
Participants had an average age of 62 years, 57% were women, 97% were White individuals, and 50% were employed in non–shift work jobs.
Sleep duration variability was quantified by the within-person standard deviation (SD) of 7-night accelerometer-measured sleep duration.
Participants with higher sleep duration SD were younger and more likely to be women, shift workers, or current smokers; those who reported definite “evening” chronotype (natural preference of the body to sleep at a certain time); those having lower socioeconomic status, higher body mass index, and shorter mean sleep duration; and were less likely to be White individuals.
In addition, a family history of diabetes and of depression was more prevalent among these participants.
A total of 2058 incident diabetes cases occurred during follow-up.
After adjustment for age, sex, and race, compared with a sleep duration SD ≤ 30 minutes, the hazard ratio (HR) was 1.15 for 31-45 minutes, 1.28 for 46-60 minutes, 1.54 for 61-90 minutes, and 1.59 for ≥ 91 minutes.
After the initial adjustment, individuals with a sleep duration SD of > 60 vs ≤ 60 minutes had a 34% higher diabetes risk. However, further adjustment for lifestyle, comorbidities, environmental factors, and adiposity attenuated the association — ie, the HR comparing sleep duration SD of > 60 vs ≤ 60 minutes was 1.11.
Furthermore, researchers found that the association between sleep duration and diabetes was stronger among individuals with lower diabetes polygenic risk score.
“One possible explanation for this finding is that the impact of sleep irregularity on diabetes risk may be less noticeable in individuals with a high genetic predisposition, where genetic factors dominate,” Dr. Kianersi said. “However, it is important to note that these sleep-gene interaction effects were not consistently observed across different measures and gene-related variables. This is something that remains to be further studied.”
Nevertheless, he added, “I want to emphasize that the association between irregular sleep duration and increased diabetes risk was evident across all levels of diabetes polygenic risk scores.”
The association also was stronger with longer sleep duration. The authors suggested that longer sleep duration “might reduce daylight exposure, which could, in turn, give rise to circadian disruption.”
Overall, Dr. Kianersi said, “Our study identified a modifiable lifestyle factor that can help lower the risk of developing type 2 diabetes.”
The study had several limitations. There was a time lag of a median of 5 years between sleep duration measurements and covariate assessments, which might bias lifestyle behaviors that may vary over time. In addition, a single 7-day sleep duration measurement may not capture long-term sleep patterns. A constrained random sampling approach was used to select participants, raising the potential of selection bias.