Decoding how macronutrients influence depressive symptoms: Combined insights from regression, nonlinear modeling, and machine learning: a cross-sectional study using NHANES data

Dietary management is an essential component of comprehensive depression care;however, studies simultaneously evaluating multiple macronutrients and their associations with depressive symptoms are limited. Data from National Health and Nutrition Examination Survey 2005 to 2023 were analyzed. Multiple linear regression, logistic regression, restricted cubic spline models, and machine learning were applied to assess the relationships between macronutrient intake and depression levels and depressive symptoms. A total of 30,285 participants (mean age 48 years; 47% male) were included. Individuals with different depression levels exhibited significant differences in dietary patterns. After multivariable adjustment, higher protein intake was consistently associated with lower depression levels ( β  = –1.159, P  < .001) and reduced depressive symptoms risk (odds ratio [OR] = 0.414, P  < .001). In contrast, higher fat intake was linked to increased depression levels ( β  = 0.437, P  < .001) and risk (OR = 1.274, P  < .001). Carbohydrates showed a modest increase in depressive symptoms risk (OR = 1.052, P  = .0105). Further decomposition revealed that sugar intake markedly increased depressive symptoms risk and levels, while dietary fiber exerted strong protective effects ( β  = –2.341; OR = 0.173, both P  < .001). Higher saturated fat intake was associated with higher depressive symptom levels, and 3 fatty acids showed notable associations: octanoic and eicosatetraenoic acids were associated with lower depressive symptoms risk, while dodecanoic acid was associated with higher risk. Machine learning analyses identified proteins, fibers, sugars, and these fatty acids as key predictors. SHapley Additive exPlanations values highlighted protein and fiber intake as the most influential factors. This observational study found that higher dietary fiber and protein intake were associated with lower odds of clinically significant depressive symptoms, suggesting that high-fiber/high-protein dietary patterns may play a beneficial role in mental health. However, causal inferences cannot be drawn from this cross-sectional design; prospective studies are needed to clarify the association and its underlying mechanisms.