Physical activity, sedentary behaviour and incidence of obstructive sleep apnoea in three prospective US cohorts

Abstract

Background: Reduced physical activity and increased sedentary behaviour may independently contribute to the development of obstructive sleep apnoea (OSA) through increased adiposity, inflammation, insulin resistance and body fluid retention. However, epidemiological evidence remains sparse and is primarily limited to cross-sectional studies.

Methods: We prospectively followed 50 332 women from the Nurses' Health Study (2002-2012), 68 265 women from the Nurses' Health Study II (1995-2013) and 19 320 men from the Health Professionals Follow-up Study (1996-2012). Recreational physical activity (quantified by metabolic equivalent of task (MET)-h per week) and sitting time spent watching TV and at work/away from home were assessed by questionnaires every 2-4 years. Physician-diagnosed OSA was identified by validated self-report. Cox models were used to estimate hazard ratios (HRs) and 95% confidence intervals for OSA incidence associated with physical activity and sedentary behaviour.

Results: During 2 004 663 person-years of follow-up, we documented 8733 incident OSA cases. After adjusting for potential confounders, the pooled HR for OSA comparing participants with ≥36.0 versus <6.0 MET-h per week of physical activity was 0.46 (95% CI 0.43-0.50; p_trend<0.001). Compared with participants spending <4.0 h per week sitting watching TV, the multivariable-adjusted HR was 1.78 (95% CI 1.60-1.98) for participants spending ≥28.0 h per week (p_trend<0.001). The comparable HR was 1.49 (95% CI 1.38-1.62) for sitting hours at work/away from home (p_trend<0.001). With additional adjustment for several metabolic factors, including body mass index and waist circumference, the associations with physical activity and sitting hours at work/away from home were attenuated but remained significant (p_trend<0.001), whereas the association with sitting hours watching TV was no longer statistically significant (p_trend=0.18).

Conclusions: Higher levels of physical activity and fewer sedentary hours were associated with lower OSA incidence. The potential mediating role of metabolic factors in the association between sedentary behaviour and OSA incidence may depend on the type of sedentary behaviour. Our results suggest that promoting an active lifestyle may reduce OSA incidence.

Figure 1.. Association of physical activity and sedentary behavior with risk of incident obstructive sleep apnea according to age and lifestyle factors

Hazard ratio represents (A) per 9 MET-hours/week increment of total physical activity, (B) per 7 hours/week increment of sitting watching TV and (C) per 7 hours/week increment of sitting at work/away from home. Multivariate analyses were stratified by cohort (NHS, NHSII, and HPFS), age (in months) and questionnaire cycle; adjusted for race/ethnicity (White, non-White), smoking (never, past, current), menopausal status (premenopausal, postmenopausal) (women only), duration of postmenopausal hormone therapy by type (never, <5, 5 to <10, ≥10 years for estrogen-only and estrogen+progestin hormone therapy separately) (women only), alcohol consumption (none, 1-4.9, 5-14.9, 15-29.9, ≥30 g/day), Alternate Healthy Eating Index (in quintiles) and habitual sleep duration (≤5, 6, 7, 8, ≥9 hours/day). Likelihood ratio test was used to calculate the P for interaction. *The sex-specific cutoff point of waist circumference was 88 cm for women and 102 cm for men.

Figure 1.. Association of physical activity and sedentary behavior with risk of incident obstructive sleep apnea according to age and lifestyle factors

Hazard ratio represents (A) per 9 MET-hours/week increment of total physical activity, (B) per 7 hours/week increment of sitting watching TV and (C) per 7 hours/week increment of sitting at work/away from home. Multivariate analyses were stratified by cohort (NHS, NHSII, and HPFS), age (in months) and questionnaire cycle; adjusted for race/ethnicity (White, non-White), smoking (never, past, current), menopausal status (premenopausal, postmenopausal) (women only), duration of postmenopausal hormone therapy by type (never, <5, 5 to <10, ≥10 years for estrogen-only and estrogen+progestin hormone therapy separately) (women only), alcohol consumption (none, 1-4.9, 5-14.9, 15-29.9, ≥30 g/day), Alternate Healthy Eating Index (in quintiles) and habitual sleep duration (≤5, 6, 7, 8, ≥9 hours/day). Likelihood ratio test was used to calculate the P for interaction. *The sex-specific cutoff point of waist circumference was 88 cm for women and 102 cm for men.