Trends in short sleep duration in the National Health and Interview Survey 2004-2018: A Decomposition Analysis

Since the early 2000s, it is increasingly common that people have short sleep durations (<=6 per 24-hour), making this a high-prevalence public health issue, especially among Black individuals. We investigate how trends in short sleep duration are influenced by changes in population aging, socioeconomic factors (e.g., education, labor force participation, marital status), and health conditions and behaviors (e.g., pain, smoking, drinking, obesity, psychological distress). We use the pooled cross-sectional data from the 2004-2018 National Health and Interview Survey, a large and nationally representative study. We partition overall trends in short sleep duration into (a) compositional effects due to distributional changes in age structure, socioeconomic and health factors and (b) unexplained rate effects. Subsequently, we compare the relative contribution of each factor to the total compositional effects among Black and White participants. Results demonstrate that greater education is associated with greater prevalence of short sleep duration among Black Americans, but reduced prevalence among White Americans. For both racial groups, population aging contributes relatively little to temporal patterns in short sleep duration; by contrast, higher distress and lower labor force participation in recent years are associated with the increasing prevalence of short sleep duration.