ABSTRACT

Background:

Coffee is among the most popular daily beverages in the United States. Importantly, coffee consumption has been associated with a lower risk of multiple health outcomes including a reduction in adiposity. DXA is a means to assess body fat and distribution.

Objectives:

The aim of this study was to examine the relation between coffee consumption and DXA-assessed adiposity and adiposity distribution.

Methods:

Cross-sectional data from the NHANES were used. Participants were adults aged 20-69 y from the 2003-2004 and 2005-2006 waves. Information on coffee consumption was assessed through the FFQ (categorized as no coffee, 0 to <0.25 cup/d, 0.25 to <1 cup/d, 1 cup/d, 2-3 cups/d, or ≥4 cups/d). Both caffeinated and decaffeinated coffee consumption were included. Trunk fat and total fat percentage were measured via whole-body DXA scans. The association between coffee consumption and body fat was investigated using age-adjusted and multivariable-adjusted linear regression models which accounted for sample weights.

Results:

Higher coffee consumption was associated with significantly lower total body fat percentage and trunk body fat percentage in a dose-response manner (all P values < 0.05) among women. Although this dose-response relation was nonsignificant among men, men aged 20-44 y who drank 2-3 cups/d had 1.3% (95% CI: -2.7%, 0.1%) less total fat and 1.8% (95% CI: -3.3%, -0.4%) less trunk fat than those who did not consume coffee. Furthermore, the association between coffee consumption and body fat percentage exhibited for both caffeinated and decaffeinated coffee among women (all P for trend < 0.001) but not among men (all P for trend > 0.05).

Conclusions:

The present study found a significant association between higher coffee consumption and lower DXA-measured adiposity. Moreover, a gender difference in this association in the general US adult population was also observed.