Association Between PHQ-9 (Patient Health Questionnaire-9) Depression Scores and Cardiometabolic Control in Older Primary Care Patients: A Retrospective Cohort Study
Abstract Background Depression is common among older adults and may affect cardiometabolic health through behavioral and biological pathways. The relationship between routinely collected Patient Health Questionnaire-9 (PHQ-9) scores and cardiometabolic measures in usual primary care is incompletely characterized. The primary objective is to evaluate associations between PHQ-9 scores and body mass index (BMI), blood pressure, hemoglobin A1c (HbA1c), and lipid measures among older adults receiving primary care. Methods This is a retrospective electronic health record cohort study including adults aged 65 years or older with at least 1 documented PHQ-9 assessment at affiliated primary care sites between January 1, 2021, and December 31, 2025. PHQ-9 score was modeled continuously. BMI and blood pressure were obtained from the PHQ-9-associated encounter; laboratory values were matched to PHQ-9 assessments within ± 30 days. Linear mixed-effects models accounted for repeated observations and adjusted for age, sex, race, ethnicity, smoking status, and season. Results Among 1,089 patients contributing 1,552 PHQ-9 observations, the median PHQ-9 score was 0.0 (interquartile range [IQR], 0.0–2.0); 85.6% of observations reflected none/minimal depressive symptoms. Each 1-point higher PHQ-9 score was associated with 1.64 mg/dL higher triglycerides (95% confidence interval [CI], 0.27 to 3.01; p = 0.019). This association persisted in analyses using log-transformed triglycerides and a ± 90-day laboratory-matching window. In the repeated-measures triglyceride subgroup, both within-patient PHQ-9 deviations and patient mean PHQ-9 scores were associated with higher triglyceride levels. PHQ-9 score was not significantly associated with BMI, systolic or diastolic blood pressure, HbA1c, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, or total cholesterol. Conclusions In this older primary care cohort, depressive symptom burden was not broadly associated with cardiometabolic measures but was consistently associated with triglyceride levels. Because the effect size was modest, depressive symptoms were predominantly none/minimal, and multiple outcomes were assessed without multiplicity adjustment. This finding should be considered hypothesis-generating.
Authors
- Traci Marquis-Eydman (ORCID: https://orcid.org/0000-0003-3905-990X)
- Ingrid Philibert (ORCID: https://orcid.org/0000-0002-7221-955X)
- Amber Parson
- Julianne Sohn
- Shyam Maisuria (ORCID: https://orcid.org/0009-0000-4246-3052)
- Reynaldo Zamora
- Alayna Santarosa
- Jenna Phreaner
Institutions
- Quinnipiac University (US)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-13
- DOI
- https://doi.org/10.21203/rs.3.rs-11543460/v1
- Primary Topic
- Cardiac Health and Mental Health
- Type
- preprint