Socioeconomic and lifestyle factors predict the association between sleep health and depression

Objective: Sleep health and depression are interconnected multidimensional constructs, yet their shared determinants remain obscure. Understanding the role of socioeconomic/lifestyle factors in predicting sleep-related depression (SRD) is critical for preventive strategies. This study aimed to identify the key socioeconomic/lifestyle predictors of SRD in the general population and patients with clinical depression. Methods: To characterize SRD, we performed regularized canonical correlation analysis between sleep and depression to identify latent phenotypes of SRD in a general population subsample (GP1; n=87,405) from the UK Biobank. Subsequently, machine-learning predictive models were developed in GP1 to predict SRD using socioeconomic/lifestyle factors. The best-performing predictive model was subsequently validated in GP2 at both baseline and follow-up (GP2; n=5,187), and in clinical depression (n=7,454) to assess its generalizability. Complementary analyses were conducted to assess other latent phenotypes (i.e., depression-related sleep, non-SRD, non-depression-related sleep, overall sleep health, and overall depression). Results: A robust multivariate association was identified between sleep and depression in GP1 (canonical r = 0.42, PFDR < 0.001). Socioeconomic/lifestyle factors moderately predicted SRD (r = 0.25; 95% CI: [0.24, 0.25]; R2 = 0.06; 95% CI: [0.06, 0.06]; rMSE = 1.08; 95% CI: [1.08, 1.09]). The top predictors were less frequency of confiding in others, more sedentary television viewing, less vigorous physical activity, and passive smoking exposure. Out-of-sample validation of the predictive model showed similar patterns in GP2 at baseline, at follow-up, and in clinical depression subsamples. Similarly, less frequency of confiding in others and greater sedentary television viewing were the main predictors of other depression-related profiles, whereas more alcohol consumption frequency, less walking frequency, and less time spent outdoors in winter predicted poor sleep-related profiles. Conclusions: Our generalizable predictive model identifies critical modifiable predictors of the association between sleep health and depression that could serve as potential targets for personalized interventions.