Unpacking the enigma of long sleep and cardiovascular disease in South African adults

Introduction: Low-income South African adults of African-origin report long sleep durations (8-10h per night). Both long sleep durations and low socio-economic status (SES) settings are associated with worse cardiovascular disease (CVD) risk. Since low SES individuals of African descent are disproportionately affected by non-communicable diseases, specifically CVD and present with striking gender differences between CVD risk factors, a better understanding of the relationship between long sleep and CVD risk in this population is required.

Methods: As part of the Modelling the Epidemiological Transition Study (METS)-Microbiome, habitual sleep was measured with actigraphy and sleep diaries for at least seven days. Participants (n=201: 107 women, 94 men, 23% employed, 39±7.9 years old) with at least 5 valid nights of actigraphy were used. Participants completed questionnaires detailing their demographics and lifestyle behaviours. CVD risk was determined using the BMI-modified Framingham 10-year CVD risk score. Actigraphy-derived sleep variables included sleep onset and offset, duration (onset to offset time), total sleep time, sleep efficiency, wake after sleep onset (WASO) and sleep fragmentation index (SFI).

Results: In the men, mean(±SD) sleep duration and sleep time were 9.5h±1.6h and 7.7h±1.5h, respectively; median(IQR) sleep onset was 22:30(21:23-23:10), sleep offset was 08:01(07:18-08:37), sleep efficiency was 80%(74-85%), WASO was 104min(77-136min) and SFI was 33.8(27.4-40.5). In the women, mean sleep duration and time were 8.7h±1.3h and 7.2h±1.3h, respectively; median sleep onset was 22:34(21:56-23:10), offset was 07:44(07:04-08:46), sleep efficiency was 82%(77-85%), WASO was 84min(68-105min) and SFI was 28.6(24.0-34.8). After adjusting for sleep apnoea, linear regression models indicated that CVD risk was higher among men with higher SFIs (β:0.22, 95% confidence interval (CI): 0.02-0.42,p=0.030) and more WASO (β:0.001, 95%CI: 0.00-0.01,p=0.010). Among the women, those with earlier sleep offset times had higher CVD risk (β:-1.02, 95%CI: -1.8- -0.19,p=0.015). No relationships were found between longer sleep duration (>9h) and CVD risk in either the men or women. Longer sleep duration (>9h) did, however, moderate the relationship between SFI and CVD risk such that men sleeping longer than 9h with greater sleep fragmentation had higher CVD risk (β: 0.02, 95%CI: 0.00-0.03,p=0.014) compared to those sleeping <9h. Longer sleep duration also moderated the relationship between sleep offset and CVD risk such that women who slept longer than 9h and had earlier sleep offsets had higher CVD risk (β:-1.06, 95%CI: -1.89- -0.22,p=0.014).

Conclusions: Data from this study show objectively-measured long, but poor quality and disturbed sleep among low-income South Africans of African-origin. Among the men, one pathway by which longer sleep duration increases CVD risk might be through fragmented sleep. Among the women, the increased CVD risk among longer sleepers with earlier wake-up times suggests a role for circadian misalignment in the development of CVD. Collectively, these results reinforce current thinking in the field which suggests that rather than focussing on isolated sleep variables, interactions between variables or composite sleep variables should be considered when examining relationships between sleep and health outcomes.