The associations of anxiety and depression with metabolic syndrome (MetS) are not consistent across studies. Anxiety and depression are highly correlated and traditional methods don’t take the structure of this correlation into account. Our aim is to disentangle the relationship of these emotional conditions with MetS, using bifactor models, modelling both general and specific aspects between anxiety and depression.
Bifactor models were tested using the baseline data from the Brazilian Longitudinal Study of Adult Health (n = 13,584). Anxiety and depression were accessed with the Clinical Interview Schedule – Revised. MetS was measured through assessment of its continuous components.
A bifactor S-1 model better represent the 14 CIS-R indicators, composed by an internalizing factor corresponding to depressive symptoms, anxiety, worry and the shared variance of all remaining CIS-R indicators, and also by residual variance explained by a somatic (e.g., fatigue and pain) and fear (e.g., panic and phobias) specific factors. Internalizing spectrum (β = 0.116; p < 0.001) and the fear specific factor (β = 0.060; p = 0.008) were associated with MetS after adjusting for confounders, whereas somatic specific factor was unlikely to be associated with MetS (β = 0.002; p = 0.934).
Anxiety and depression indicators were associated with MetS via a shared internalizing factor and also by a residual fear factor, but not by somatic residual factor. This finding has potential implications about shared biological and behavioral mechanisms that may link emotional conditions with MetS in adults.