Association Between Caffeine Intake and Depressive Symptoms: A Cross-sectional Analysis Using the NHANES 2017-2018 Database

Main Article Content

Marion Feddersen Welkner
Jorgelina Stegmann
Miray Nilufer Cimsit Kemahli
Diego Perez Paez
Hanna Bennewart
Laura Amaral de Lara Resende
Fernando Espinoza
Paula Aranguren
Paula Bertagnoli
Sarah Burgos
Nikolaos Paschos
Carlos Roa
Natália Puschnick Gimenez
Letícia Cabreira dos Santos
Lissa Raquel Sandoval Castillo
Karla Andrade Ruiz
Siddiha Poovelil Shahulhameed
Paulina Troncoso Escudero
Priscila Sales Barroso
Yulissa Hernández Frías
Daniela Vicinansa Monaco Ferreira
Ala Hassan
Henrique Balseiro Zin
Alice Campanholo de Oliveira
Jae Kyung Jo
Fernando Jose Gomez Juarez
Roy Rosado

Abstract

Background: Depression is a leading global cause of disability, and identifying modifiable lifestyle factors such as caffeine intake may help mitigate its burden. Although caffeine consumption has been proposed as protective against depressive symptoms, existing evidence remains inconsistent and often insufficiently adjusted for key confounders.


Objective: To examine the association between caffeine intake and depressive symptoms among U.S. adults using data from the 2017–2018 National Health and Nutrition Examination Survey (NHANES).


Methods: This cross-sectional study was conducted and reported according to STROBE guidelines. Adult NHANES participants were included. Daily caffeine intake (mg/day), assessed via 24-hour dietary recall, was categorized as low, moderate, or high. Depressive symptoms were measured using the Patient Health Questionnaire-9 (PHQ-9) and classified into five clinically meaningful severity levels. Weighted ordered logistic regression models were used to evaluate associations between caffeine intake and depressive symptom severity, adjusting for demographic, socioeconomic, lifestyle, and clinical covariates, including age, sex, education, body mass index (BMI), marital status, family income, ethnicity, smoking status, sedentary behavior, thyroid disease, diabetes, and cardiovascular disease.


Results: The analytic sample included 4,503 adults (mean age 47.3 years; 51.8% female). Caffeine intake was low in 47%, moderate in 27%, and high in 26% of participants. In weighted univariable analyses, both moderate (OR 0.73; 95% CI 0.60–0.90; p=0.006) and were associated with lower odds of more severe depressive symptoms compared with low intake. After multivariable adjustment, only high caffeine intake remained independently associated with lower depressive symptom severity (OR 0.62; 95% CI 0.48–0.80; p=0.001). In adjusted models, female sex, obesity, unmarried status, current smoking, thyroid disease, and cardiovascular disease were independently associated with higher odds of depressive symptoms, whereas higher educational attainment and higher family income were independently associated with lower odds of depressive symptoms.


Conclusions: High caffeine intake was independently associated with lower depressive symptom severity in U.S. adults, even after accounting for a broad range of confounders. However, several demographic, socioeconomic, and health-related factors exerted stronger and opposing influences on depressive symptoms. These findings suggest that while caffeine consumption may play a modest protective role, addressing broader social and health determinants is likely more impactful for depression prevention and management. Longitudinal studies are needed to clarify causality and temporal relationships.

Article Details

How to Cite
Association Between Caffeine Intake and Depressive Symptoms: A Cross-sectional Analysis Using the NHANES 2017-2018 Database. (2026). Principles and Practice of Clinical Research, 11(4). https://doi.org/10.21801/ppcrj.2025.114.10
Section
Full Length Articles

How to Cite

Association Between Caffeine Intake and Depressive Symptoms: A Cross-sectional Analysis Using the NHANES 2017-2018 Database. (2026). Principles and Practice of Clinical Research, 11(4). https://doi.org/10.21801/ppcrj.2025.114.10

Most read articles by the same author(s)