LAUSR

We thank Tomoyuki Kawada for his comments in response to our recent article in the Journal of Diabetes entitled “Short sleep duration and longer daytime napping are associated with non-alcoholic fatty liver disease in Chinese adults,” in which we assessed the association between sleep habit and nonalcoholic fatty liver disease (NAFLD). Kawada commented on the conclusions and confounding factors in the article. We appreciate the opportunity to discuss these issues. According to current categories of nightly sleep duration (≤6.0, 6.1–7, 7.1–8, 8.1–9, and >9 h [reference group]), shorter sleep durations tend to be associated with higher risk of prevalent NAFLD. In addition, when fatty liver index (FLI)-diagnosed NAFLD (i.e. FLI ≥60) was used as the outcome, an increased risk of prevalent NAFLD was only associated with 6.1–7 and <6 h sleep duration. We did not specify the “short” nightly sleep duration that increased the risk of NAFLD in the study’s conclusion because a different study population structure could generate different nightly sleep patterns, including sleep duration. Therefore, considering the effects of ethnicity and population structure on sleep patterns, whether there is a negative correlation between sleep duration and prevalent NAFLD needs to be investigated in different populations. Previous studies have provided some insight into the potential roles of inflammatory cytokines, such as interleukin (IL)-6, enhanced by sleep disturbance, in the pathogenesis of NAFLD. However, we did not provide any data regarding IL-6 in our study, but we do speculate that inflammatory cytokines may be an intermediate factor associated with sleep duration and NAFLD. We agree with Kawada that it is important to consider ethnicity when evaluating the risk factors for NAFLD. The population in our study was confined to community-based middle-aged and elderly Chinese with similar customs and lifestyles. As noted by Kawada, the relationship between NAFLD and type 2 diabetes mellitus (T2DM) is bidirectional. These conditions share insulin resistance as a common pathophysiological mechanism. A metaanalysis has indicated an association between sleep duration and diabetes. To reduce the effect of diabetes, we conducted stratified analysis according to insulin resistance status (yes/no). However, the results showed that the associations between short sleep durations and NAFLD were generally similar across subgroups. We agree with Kawada that there may be a synergistic effect between prolonged napping and short nightly sleep duration. Specifically, the association between daytime napping and NAFLD remained after adjustment for nightly sleep duration and other covariates, suggesting an independent contribution of daytime napping to NAFLD, as reported in our article. In conclusion, given the existence of confounding factors, future prospective or mechanistic studies are warranted to elucidate the associations of nightly sleep duration and daytime napping with NAFLD.