Association between objectively measured sleep duration, adiposity and weight loss history

Background: An association between sleep duration and obesity has been suggested in several studies, but most have relied on self-reported sleep data and BMI as the only measure of adiposity. Moreover, a relationship between weight loss history and attained sleep duration has not previously been thoroughly explored. Design: The study consisted of 1202 participants from the UK, Denmark and Portugal, participants of the NoHoW study (should we mention it?) who had achieved a verified weight loss of ≥5 % and had a BMI of ≥25 kg/m2 prior to losing weight. Information was available on sleep duration (collected during a 14-day period using the Fitbit Charge 2TM), adiposity measures, weight loss history and several potential confounding factors. Analysis of covariance was conducted to assess the associations between sleep duration and body mass index (BMI), fat mass index (FMI), fat free mass index (FFMI) and waist hip ratio (WHR). Linear regression or analysis of covariance was applied to assess whether weight loss history (twelve-month weight loss, frequency of prior weight loss attempts and average duration of weight loss maintenance after prior weight loss attempts) was associated with attained sleep duration. Results: We found an association between sleep duration and BMI (P<0.001, controlling for ??? If you have space I think you should include them), with the highest BMI observed in the group of participants sleeping < 6 hours a day [34.0 kg/m2 (95% CI: 31.8-36.1)]. Less difference in BMI was detected between the remaining groups, with the lowest BMI observed among participants sleeping 8-9 hours a day [29.4 kg/m2 (95% CI: 28.8-29.9)]. Similar results were found for FMI (P=0.008) and FFMI (P<0.001). We found no association between sleep duration and WHR. Likewise, we found no evidence of associations between weight loss history and attained sleep duration. I have read the results now... It is surprising to see that the FFMI associations present the same pattern as the BMI and FMI. Wouldn't be possible to use a percentage of Fat Mass and Fat Free Mass, instead of the FMI and FFMI. It seems counter-intuitive to observe a similar pattern between a measure of fat and a measure of muscle (I'm oversimplifying here, I know). I suspect that this is showing that the bigger participants - who have more BMI, FM and FFM in kg/m2- are the ones who sleep less. But we may be missing the participants who have more FM relative to their weight - whom I suspect are the ones with worst sleep patterns - and the ones who have more FFM relative to their weight - whom I suspect will have the best sleep pattern. Sorry if I'm making a mess out of this... Conclusion: These results build on the current evidence suggesting that a short sleep duration is associated with a higher BMI with at least 5% WL in the last 12 months. Most previous research has suggested that this association is primarily due to an association between sleep and adiposity, but our results suggests that the association also involves lean mass. We found no evidence of association between weight loss history and attained sleep duration.
Refereed journal