Sleep health associations with serum metabolites in healthy adults

Abstract

Study objectives: Short and long sleep duration as well as poor sleep quality have been linked to higher prevalence of metabolic disorders. However, it is still unclear how diverse sleep variables relate to different metabolic pathways. This study examines how different features of sleep health relate to serum metabolites.

Methods: The study used data from 197 healthy individuals aged 20-79 (Females n = 103) from the IronAge study performed at Karolinska Institutet in Sweden. Sleep variables were assessed with the Karolinska Sleep Questionnaire, where the following variables were computed: sleep duration, sleep debt, midpoint, social jetlag (i.e., the discrepancy between midpoint on free and workdays), napping frequency and sleep quality. Morning fasting blood samples were collected and ¹H NMR spectroscopy was utilized for metabolomic analysis. The metabolites were categorized according to their major metabolic pathways: amino acid, lipid, carbohydrate, energy and gut microbiota. Linear regressions were performed to examine the relationship between each sleep variable and metabolite.

Results: Sleep duration, midpoint of sleep on free days, social jetlag and chronotype associated with eight metabolites at a significance level of p<0.01. Notably, midpoint associated with most metabolites spanning multiple pathways. A later midpoint was associated with higher levels of metabolites in the lipid pathway, and lower levels in the amino acid and energy pathway.

Conclusion: These observations indicate that sleep timing features, midpoint and social jetlag, have a stronger relationship with morning metabolism than other sleep health dimensions. Following replication in larger samples, these complex relationships may hold potential for health promotion.

Keywords: Karolinska sleep questionnaire; Metabolomics; Sleep health; Sleep timing.

Fig. 1

Heatmap of pairwise correlations among sleep variables and serum metabolites significant at p < 0.01. From the series of robust linear regressions, four sleep variables (sleep duration, MSF, social jetlag and chronotype) were associated with 8 metabolites at a significance level of p < 0.01. Left) Spearman correlation coefficients between the sleep variables. Right) Pearson correlation coefficients between the metabolites associated with sleep variables at p < 0.01. Pairwise correlations significant at p < 0.05∗, p < 0.01∗∗, p < 0.001∗∗∗. MSF = Mid-sleep on free days.