A number of epidemiologic studies have reported an association between habitual short duration of sleep and obesity or diabetes. Either an inverse or U-shaped relationship between sleep duration and weight occurred, depending on the study. It is unclear whether chronic sleep deprivation and impaired sleep quality have a causal relationship with obesity, or whether they represent coincidental secular trends. The effect of chronic sleep deprivation on resting energy expenditure (REE) is not well characterized. It has been suggested that sleep deprivation through activation of the stress system and the sympathetic nervous system could lead to increased energy expenditure.
The aim of this randomized, prospective, long-term, intervention trial was to investigate the relationship between sleep, REE, and stress hormones in a population of obese adult men and premenopausal women. The study was conducted at a tertiary, referral research clinical center. The presence and severity of sleep apnea, as well as sleep duration and quality, were determined in 126 obese individuals with the following characteristics: either sex: 30 males and 96 females; age: 40.5 ± 6.9 years; body mass index: 38.6 ± 6.5 kg/m2; sleep duration, 360 ± 50 min/night; and sleep efficiency, 79.5% ± 7.5%.
The primary study outcome measures included (1) REE and respiratory quotient (RQ) assessed by indirect calorimetry, (2) sleep duration and sleep efficiency assessed by actigraphy, (3) daytime sleepiness assessed by the Epworth Sleepiness Scale questionnaire, (4) sleep quality estimated by questionnaires, and (5) sleep apnea evaluated by respiratory disturbance index (RDI). Levels of plasma adrenocorticotropic hormone (ACTH), serum cortisol, 24-hour urinary free cortisol (UFC), and 24-hour catecholamines were also measured. Resting energy expenditure was adjusted for fat mass, age, and sex.
Respiratory disturbance index was directly correlated with adjusted REE (r = 0.307, P = 0.003) and RQ (r = 0.377, P = 0.001). There was an inverse correlation between sleep efficiency and RQ (r = −0.200, P = 0.033). Several sex differences were noted. The relationship between RDI and REE was stronger in men than in women (P = 0.035). Among women, there was a positive correlation between serum cortisol and the adjusted REE (r = 0.407, P = 0.001), and the Epworth sleepiness score tended to be inversely correlated with adjusted REE (r = −0.190, P = 0.086). Respiratory quotient was positively related to RDI in women (P < 0.004), whereas subjective sleep time was related to RQ in men (P < 0.008). Three parameters (RDI, serum cortisol, and urinary norepinephrine) were positively related to REE in a linear regression model (P < 0.05), whereas only serum cortisol was also directly correlated with adjusted REE (P < 0.007).
These data suggest that poor sleep quality is associated with increased REE. The higher RQ in patients with sleep apnea indicates a shift from fat toward carbohydrate oxidation and activation of the stress system.
Section on Neuroendocrinology of Obesity (L.d.J., X.Z., M.S.M., S.M.Z., G. Cizza.), Diabetes, Endocrinology, and Obesity Branch/National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD; Department of Endocrinology and Kidney (P.P.), University Hospital of Pisa, Pisa, Italy; and Department of Laboratory Medicine (G. Csako), Clinical Center, National Institutes of Health, Bethesda, MD