(P14-004-23) Derived Dietary Patterns Using Reduced Rank Regression in Postmenopausal Women and Risk of Mortality: A European Population-Based Study

Objectives: Due to the menopause, women might experience various complications related to cardiometabolic, bone, and mental health. Diet can affect these complications. We aimed at extracting dietary patterns (DPs) based on the factors associated with menopause complications in postmenopausal women. Further, we investigated the association between the DPs adherence scores and risk of all-cause and CVD mortality.

Methods: We generated DPs using reduced rank regression among 1814 postmenopausal women participating in the Rotterdam Study, a population-based study in the Netherlands. This approach identifies linear functions of predictor variables that explain the maximum variation in the response variables. We used the intake of 25 food groups as predictors and several indicators related to lipid profile, insulin resistance, systolic blood pressure (SBP), anthropometry (e.g. lean mass index (LMI), fat mass percentage (FM%)), bone mineral density (BMD), cognitive function (G-factor), sleep quality and depression as responses. The associations between the DPs adherence scores and risk of mortality were determined by Cox proportional hazard models.

Results: The first extracted DP (DP1) explained 2.95% of variation in all responses and was characterized by the higher loadings of vegetables, whole grain, alcoholic beverages and coffee. DP1 was positively correlated to G-factor and BMD, of which explained 12.11% and 4.13% of their variations and negatively to SBP, of which explained 5.62% of its variation. The second DP (DP2) explained 1.54% of variations in all responses and was identified by the higher loadings of meat products and ready-meals and a lower loading of sweets. DP2 was positively correlated to FM%, LMI, and BMD, of which it explained 5.45%, 3.47%, and 3.29% of their variations. No association was observed between the DPs adherence scores and risk of all-cause or CVD mortality.

Conclusions: We extracted two DPs based on a combination of health indicators related to menopause complications among postmenopausal women; however, no association was found between the adherence to these DPs and risk of mortality. We highlight the importance of nutritional researches exclusively among postmenopausal women, an underrepresented population, and understanding the role of diet through menopause complications.

Funding Sources: Nutricia Research Foundation