INTRODUCTION: Sex hormones shape biological sex differences and alter the onset and severity of sleep and metabolic diseases in a sex-specific manner. To better understand relationships and underlying mechanisms, we develop summary proteomics and metabolomics scores for sex hormones and investigate their associations with sleep and metabolic disorders.

METHODS: We used proteome- (n= 3680) and metabolome- wide (n= 1649) data from the baseline exam of the Multi-Ethnic Study of Atherosclerosis (MESA) cohort to develop female- and male-specific omics scores for sex hormones including total (Total T), bioavailable (Bio T), and free (Free T) testosterone, estradiol (E2) and sex hormone binding protein (SHBG). Each omics dataset was randomly split assigning 80% of participants to a training dataset and the remaining 20% to a test dataset. We applied linear regression with bootstrap standard errors, adjusting for age, BMI, self-reported race/ ethnicity and study site, to identify sex hormone-associated proteins and metabolites (i.e FDR< .05). Lasso penalized regression was then used to select independent features, from which weighted protein (ProtS) and metabolite scores (MetS) were constructed as weighted sums, and examined in the validation dataset. Subsequently, we conducted sex-stratified association analysis of the validated omics scores using data from MESA baseline, exams 4 (proteomics) and 5 (proteomics, metabolomics) with sleep and metabolic phenotypes, timepoints where sex hormones were not measured.

RESULTS: All constructed omics scores were significantly associated with their corresponding hormones in the test dataset. Higher omics scores of SHBG and lower omics scores of Free T were associated with lower diabetes risk in both sexes; and higher E2 scores with higher incident hypertension risk only in men. In males, Total T had protective diabetes associations, whereas in females they were linked to greater risk. Similarly, higher ProtS-Free T and lower ProtS-SHBG were associated with increased risk for OSA in both sexes. Finally, higher E2 scores were associated with higher risk of insomnia only in males.

CONCLUSIONS: Summary omics-based scores reveal sex-specific cross-sectional associations with sleep and incident metabolic disorders. These findings highlight the potential of these omics proxies to improve risk stratification and generate insights into mechanisms linking sex hormones with disease.

Integrative approaches that simultaneously model multi-omics data have gained increasing popularity because they provide holistic system biology views of multiple or all components in a biological system of interest. Canonical correlation analysis (CCA) is a correlation-based integrative method designed to extract latent features shared between multiple assays by finding the linear combinations of features-referred to as canonical variables (CVs)-within each assay that achieve maximal across-assay correlation. Although widely acknowledged as a powerful approach for multi-omics data, CCA has not been systematically applied to multi-omics data in large cohort studies, which has only recently become available. Here, we adapted sparse multiple CCA (SMCCA), a widely-used derivative of CCA, to proteomics and methylomics data from the Multi-Ethnic Study of Atherosclerosis (MESA) and Jackson Heart Study (JHS). To tackle challenges encountered when applying SMCCA to MESA and JHS, our adaptations include the incorporation of the Gram-Schmidt (GS) algorithm with SMCCA to improve orthogonality among CVs, and the development of Sparse Supervised Multiple CCA (SSMCCA) to allow supervised integration analysis for more than two assays. Effective application of SMCCA to the two real datasets reveals important findings. Applying our SMCCA-GS to MESA and JHS, we identified strong associations between blood cell counts and protein abundance, suggesting that adjustment of blood cell composition should be considered in protein-based association studies. Importantly, CVs obtained from two independent cohorts also demonstrate transferability across the cohorts. For example, proteomic CVs learned from JHS, when transferred to MESA, explain similar amounts of blood cell count phenotypic variance in MESA, explaining 39.0%   50.0% variation in JHS and 38.9%   49.1% in MESA. Similar transferability was observed for other omics-CV-trait pairs. This suggests that biologically meaningful and cohort-agnostic variation is captured by CVs. We anticipate that applying our SMCCA-GS and SSMCCA on various cohorts would help identify cohort-agnostic biologically meaningful relationships between multi-omics data and phenotypic traits.