We found that sleep problems are highly prevalent in the United States. More than 90% of NHANES respondents reported any examined sleep problem and 10% to 40% reported any given problem. Inadequate sleep, frequent daytime sleepiness, apnea, leg symptoms, and nocturia were all more common among those with diabetes than those without diabetes. These associations persisted after adjustment for conditions that could contribute to poor sleep and were generally strongest among respondents who were younger and female. Although frequent sleeping pill use and severe sleep deprivation were both common, the prevalence of these problems did not differ by diabetes status.
The association of diabetes with sleep apnea, which was independent of obesity — a risk factor for both conditions — has been shown previously (10
), as has our observed association between diabetes and restless legs symptoms (19
), which may have neuropathic origins or be related to commonly used medications such as second-generation antidepressants and antihistamines. However, less is known about other sleep problems among people with diabetes. We found that nocturia and excessive sleepiness were also independently associated with diabetes. Nocturia was the most common sleep problem reported in this population, and the association with diabetes persisted even after adjustment for diuretic use. Nocturia is often not considered among the common sleep problems, but it may have deleterious effects on both sleep and quality of life (20
), which was not assessed in this study. Inadequate sleep and excess sleepiness are similarly under-recognized (21
) and detrimental to physical and mental health.
Although we have shown that diabetes is independently associated with increased risk of several sleep problems, the cross-sectional study design precludes causal inference, despite the magnitude and dose-response nature of the observed associations. Additionally, while we found that duration of diabetes was positively associated with increased risk of apnea, nocturia, leg symptoms, and daytime sleepiness — suggesting a possible temporal relationship between diabetes and subsequent sleep problems — we cannot definitively establish the directionality of the association. Previous research examining potential causal links between diabetes and sleep problems has been inconclusive. For example, a small study showed that forced sleep deprivation in healthy young men led to decreased leptin levels and increased appetite (22
); another cross-sectional analysis of participants in the Sleep Heart Health Study showed that impaired glucose tolerance was more common among people with habitual sleep restriction (23
), suggesting that sleep problems may lead to diabetes via physiologic mechanisms. However, the prevalence of sleep problems among patients with type 1 diabetes was found to be similar to that seen in type 2 diabetes (24
). Given the difference in etiology, these results suggest that diabetes may also lead to sleep problems.
Other limitations of note include the self-report of sleep problems, which is subject to recall and detection bias. Data for neuropathy or chronic pain, pruritis, caffeine use, neck circumference, and 24-hour blood pressure were not available in the surveys analyzed. Restless legs symptoms may not be limited to leg cramps or jerks. Sleep studies were not performed. Although we were able to show that adjustment for income did not affect the association between diabetes and sleep problems, we did not have data on poverty-related stressors. Such stressors are thought to increase the "allostatic load" and both interfere with sleep and increase risk for chronic diseases such as diabetes (25
However, our study also augments previous research on diabetes and sleep problems in several ways. It is a large, nationally representative study examining a comprehensive range of sleep problems, regardless of diagnosis. We were able to compare prevalence among people with diabetes to that of similar people without diabetes, to determine whether diabetes had an independent effect on sleep problems after adjustment for other factors strongly related to both diabetes and sleep, including age, sex, obesity, cardiovascular disease, depression, kidney disease, alcohol use, and income.
In conclusion, people with diabetes are more likely to have sleep problems than their counterparts of the same age, sex, and race/ethnicity without diabetes, regardless of several behaviors and comorbid conditions that could interfere with sleep. Because sleep problems and diabetes are both associated with poor quality of life (5
), further research is warranted. Such efforts should examine the possibly dual directionality of the association of diabetes and sleep problems and explore possible causal inference and associated outcomes such as quality of life, ideally in a longitudinal cohort. Our results suggest that diabetes patients should be screened for these treatable sleep problems (sleep apnea, insomnia, restless legs, and nocturia). Recognition and treatment of sleep problems, particularly in the disproportionately affected younger and female diabetes populations, could considerably improve quality of life and, possibly, clinical outcomes among these patients.