Plasma Phospholipid Polyunsaturated Fatty Acid Associations with Neurocognition

Abstract

Neurocognitive deficits are implicated in major depressive disorder (MDD) and suicidal behavior, and cognitive function may be affected by blood levels of polyunsaturated fatty acids (PUFAs). Neuroprotective functions have been described for omega-3 (n-3) PUFAs, while omega-6 (n-6) PUFAs exhibit broadly opposing activities. Both classes of PUFAs are linked to MDD and suicidal behavior. However, few studies have investigated the relationships between PUFAs and neurocognitive function with respect to MDD or suicidal behavior. Among participants with MDD (n = 45) and healthy volunteers (HV, n = 30) we assessed performance on tasks of attentional capacity and executive function and its relationship to plasma phospholipid PUFA levels, expressed as a percentage of total plasma phospholipids, for eicosapentaenoic acid (EPA%), docosahexaenoic acid (DHA%), and arachidonic acid (AA%). Regression models tested the correlations between PUFA levels and task performance in three groups: MDD with a history of suicide attempt (SA, n = 20), MDD with no attempts (NA, n = 25), and HV. Interaction testing indicated a significant positive correlation of EPA% with continuous performance test scores in the NA group (F = 4.883, df = 2,72, p = 0.01), a measure of sustained attention. The AA% correlated negatively with performance on two executive function tasks, object alternation (beta = -3.97, z-score = -2.67, p = 0.008) and the Wisconsin card sort (beta = 0.80, t-score = -2.16, df = 69, p = 0.035), after adjustment for group and age, with no group effects. Our findings suggest a role for PUFA imbalance in attentional functioning and executive performance; however, no MDD-specific effect was observed.

Keywords: PUFA; attention; depression; executive function; neurocognition; omega-3 fatty acids; suicide.

Figure 1

Partial correlations between eicosapentaenoic acid as a percentage of total phospholipids (EPA%) and continuous performance task performance sensitivity index (CPT-IP d’). Shown are partial correlations between age-residualized, ln-transformed EPA% and CPT-IP d’ scores in the HV, NA, and SA groups. Correlations were estimated using Pearson’s correlation coefficient, r.

Figure 2

Object alternation performance and AA% levels. Group-wise levels of AA as a percentage of total plasma phospholipids (AA%) between participants who attained and failed to attain the object alternation test to criterion. AA% data are ln-transformed; apparent outliers (•) for subgroup boxplots were not considered outliers within the total sample distribution. In logistic regression models, AA% was a negative predictor of OAT performance (p = 0.008), covarying for group and age; however, there were no significant group interactions.