(P19-012-23) Dietary Intake of Omega-3 Fatty Acids Measured by 24HR Differs From Omega-3 in Erythrocyte Membranes in Brazilian Middle-Aged and Older Adults

Objectives: To compare the reported dietary intake of omega-3 fatty acids assessed by two 24h recalls (24HR) and the omega-3 index (O3I - defined as the proportion of eicosapentaenoic (EPA) and docosahexaenoic (DHA) acids) measured in erythrocyte membranes of Brazilian middle-aged and older adults.

Methods: The study included 192 individuals of both sexes, aged 45-75 years old from the 2015 Health Survey of São Paulo with a Focus on Nutrition, a cross-sectional population-based study of residents of São Paulo, Brazil. The O3I was obtained by adding the EPA and DHA percentages in the fatty acids composition of erythrocyte membranes, using a gas chromatograph (Shimadzu, CG-2010, Kyoto, Japan). Dietary intake data were collected through two 24HR on nonconsecutive days, adjusted for usual dietary intake using the multiple source method (MSM). Dietary and biomarkers of omega-3 fatty acids were compared using the Spearman correlation, and multiple linear regression (adjusted for confounders).

Results: The O3I presented a mean percentage of 4.4% with no significant differences between sexes (p=0.947). Only 2.1% (n=4) of the individuals had an O3I above 8%, while in 38% (n=73), the O3I was below 4%. Moreover, the mean reported dietary intake of omega-3 fatty acids was 57mg/day, with 23% (n=44) of the population reporting an intake below 5mg/day. There was no significant correlation between dietary intake and O3I (R=0.028, p=0.698), with results remaining similar after the removal of the individuals who did not report any intake of omega-3 (0mg in both 24HR) or it was below 5mg (R=-0.077, p=0.353). After adjusting for age, sex, total fat, and energy intake in multiple linear regression models, the reported dietary intake of omega-3 explained only 1.5% of the variance in O3I, and the association was not statistically significant.

Conclusions: The proportion of individuals with an O3I above the cut-off associated with cardioprotection (≥8%) was small, while there was a high prevalence of individuals with a low O3I (≤4%). Reported dietary intake of omega-3 fatty acids in 24HR was low in the population. However, it was not associated with the omega-3 biomarker (O3I). In studies investigating omega-3 intake with health outcomes, the 24HR may lead to bias in omega-3 fatty acids dietary assessment.

Funding Sources: São Paulo Research Foundation (grant numbers 2020/019451-9 and 2017/05125-7)