Omega-6 polyunsaturated fatty acids and adiposity: a cross-sectional and longitudinal analysis in the UK Biobank

The relationship of omega-6 polyunsaturated fatty acids (PUFAs), especially linoleic acid (LA), with adiposity remains uncertain (1) . Small-scale human clinical trials suggest that a greater dietary LA intake may favourably influence body composition (1–2) , but evidence from observational studies have been mixed (1, 3–4) . In addition, limited observational research has evaluated changes over time, as opposed to cross-sectional relationships (4) . To address these gaps in knowledge, we examined the associations of circulating LA, non-LA omega-6, and total omega-6 PUFAs with adiposity outcomes in the UK Biobank. Our analysis were conducted in the UK Biobank. Using multivariable linear models, we evaluated cross-sectional and longitudinal associations between omega-6 fatty acid levels (total omega-6, LA, and non-LA omega-6 [defined as total omega-6 minus LA)] and waist circumference (WC), weight, and whole-body fat mass (FM). FA levels were determined by nuclear magnetic resonance and assessed as per interquintile range (90th – 10th percentile) and by quintiles. Models were adjusted for potential confounders, including relevant demographics, lifestyle and medical history. Interaction terms were evaluated in separate models to test for potential age or sex modification of the LA–outcome relationships. We evaluated a total of 272,587 individuals for the cross-sectional analysis and 58,335 individuals for the longitudinal analysis. Nearly half were female, most were Caucasian, and with just about half reporting a college education. Cross-sectionally, higher circulating LA was inversely associated with WC, weight, and FM. Individuals in the highest quintile of LA had significantly smaller WC [–11.04 (–11.17, –10.91) cm], lower weight [–11.77 (–11.92, –11.62) kg], and lower FM [–7.87 (–7.97, –7.77) kg] vs. individuals in the lowest quintile. Associations for total omega-6 generally mirrored those of LA. Conversely, non-LA omega-6 was positively associated with WC when comparing the highest to lowest quintile [1.46 (1.32, 1.61) cm], weight [2.41 (2.25, 2.58) kg], and FM [1.81 (1.69, 1.92) kg]. Longitudinal analyses largely corroborate these findings, with average annual changes in WC, weight, and FM inversely associated with LA and positively associated with non-LA omega-6. Age- and sex-stratified analyses revealed heterogeneity, with stronger inverse associations for younger individuals and women (all p<0.001). We found that higher circulating LA, but not non-LA omega-6, were consistently linked to a lower WC, weight, and FM in both cross-sectional and longitudinal analyses. Our results lend support to current dietary recommendations to promote LA-rich oils. Divergent associations between LA and non-LA omega-6 suggest that omega-6 PUFAs should not be treated as a homogenous group. Further investigations are therefore needed to clarify the distinct health effects of individual non-LA omega-6.