n-3 index, the erythrocyte proportion of the EPA + DHA fatty acids is a clinical marker of age-related disease risk. It is unclear whether regular intake of α-linolenic acid (ALA), a plant-derived n-3 polyunsaturated fatty acid, raises n-3 index in older adults. Of the 356 participants at the Loma Linda, CA centre from the original study, a randomly selected subset (n 192) was included for this secondary analysis (mostly Caucasian women, mean age 69 years). Participants were assigned to either the walnut (15 % of daily energy from walnuts) or the control group (usual diet, no walnuts) for 2 years. Erythrocyte fatty acids were determined at baseline and 1-year following intervention. No differences were observed for erythrocyte EPA, but erythrocyte DHA decreased albeit modestly in the walnut group (–0·125 %) and slightly improved in the control group (0·17 %). The change in n-3 index between the walnut and control groups was significantly different only among fish consumers (those who ate fish ≥ once/month). Longitudinal analyses combining both groups showed significant inverse association between the 1-year changes of the n-3 index and fasting plasma TAG (ß = −10), total cholesterol (ß = −5·59) and plasma glucose (ß = −0·27). Consuming ALA-rich walnuts failed to improve n-3 index in elders. A direct source of EPA/DHA may be needed to achieve desirable n-3 index, as it is inversely associated with cardiometabolic risk. Nevertheless, incorporating walnuts as part of heart healthy diets is still encouraged.

The association between n-3 PUFA intake and type 2 diabetes (T2D) is unclear, and studies relating objective biomarkers of n-3 PUFA consumption to diabetic status remain limited. The aim of this study was to determine whether erythrocyte n-3 PUFA levels (n-3 index; n-3I) are associated with T2D in a cohort of older adults (n 608). To achieve this, the n-3I (erythrocyte %EPA+%DHA) was determined by GC and associated with fasting blood glucose; HbA1c; and plasma insulin. Insulin resistance (IR) was assessed using the homeostatic model assessment of insulin resistance (HOMA--IR). OR for T2D were calculated for each quartile of n-3I. In all, eighty-two type 2 diabetic (46·3 % female; 76·7 (sd 5·9) years) and 466 non-diabetic (57·9 % female; 77·8 (sd 7·1) years) individuals were included in the analysis. In overweight/obese (BMI≥27 kg/m2), the prevalence of T2D decreased across ascending n-3I quartiles: 1·0 (reference), 0·82 (95 % CI 0·31, 2·18), 0·56 (95 % CI 0·21, 1·52) and 0·22 (95 % CI 0·06, 0·82) (Ptrend=0·015). A similar but non-significant trend was seen in overweight men. After adjusting for BMI, no associations were found between n-3I and fasting blood glucose, HbA1c, insulin or HOMA-IR. In conclusion, higher erythrocyte n-3 PUFA status may be protective against the development of T2D in overweight women. Further research is warranted to determine whether dietary interventions that improve n-3 PUFA status can improve measures of IR, and to further elucidate sex-dependent differences.