Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Jul 1.
Published in final edited form as: Int J Geriatr Psychiatry. 2013 Dec 11;29(7):747–757. doi: 10.1002/gps.4058

Omega-3 Fatty Acid Biomarkers and Subsequent Depressive Symptoms

Jane E Persons 1, Jennifer G Robinson 1,2, Eric M Ammann 1, William H Coryell 3, Mark A Espeland 4, William S Harris 5, JoAnn E Manson 6, Jess G Fiedorowicz 1,2,3
PMCID: PMC4048630  NIHMSID: NIHMS553293  PMID: 24338726

Abstract

Objective

We sought to determine the relationship between the omega-3 fatty acid content of red blood cell membranes (RBC), in particular docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA), and baseline and new-onset depressive symptoms in postmenopausal women. We secondarily sought to characterize the association between dietary omega-3 fatty acid intake and depressive symptomatology.

Methods

Study participants included 7,086 members of the Women's Health Initiative Memory Study (aged 63–81) who had an assessment of RBC omega-3 fatty acid concentrations at the baseline screening visit. Depressive symptoms at baseline and follow-up were characterized using the Burnam 8-item scale for depressive disorders (CES-D/DIS short form), and secondarily additionally inferred by antidepressant medication use.

Results

In multivariable-adjusted models, our primary exposure, RBC DHA+EPA, was not related to depressive symptoms by any measure at baseline or follow-up, nor were RBC total omega-3, DHA, or EPA (all p>0.2). In contrast, dietary intake of omega-3 was positively associated with depressive symptoms at baseline (adjusted OR 1.082, 95% C.I. 1.004–1.166; p=0.04 for dietary DHA+EPA and Burnam Score ≥ 0.06), although this generally did not persist at follow-up.

Conclusion

No relationship between RBC omega-3 levels and subsequent depressive symptoms was evident, and associations between dietary omega-3 and depressive symptoms were variable. Biomarkers of omega-3 status do not appear to be related to risk of new depression in post-menopausal women.

Keywords: Omega-3 Fatty Acids, Biological Markers, Diet, Eicosapentaenoic Acid, Depression, Epidemiologic Studies

Introduction

There is evidence to suggest that women tend towards higher biological concentration of DHA compared to men, due to the effects of estrogen (Burdge and Wootton, 2002, Salisbury et al., 2011). This information, along with evidence of higher brain concentration of DHA compared to other body tissues, suggests relevance for DHA content in the CNS in post-menopausal depression. In some – but not all – observational studies, low levels of omega-3 fatty acids have been shown to be associated with depressive symptomatology (Lin et al., 2010). A 31-year follow-up study of the Northern Finland 1966 Birth Cohort showed an increased risk of depression among female subjects with low fish intake (Timonen et al., 2004); however, several other large longitudinal studies found no such associations - a prospective analysis of older women (50–77 years of age at baseline) from the Nurses' Health Study found no association between omega-3 dietary intake and incident depression (Lucas et al., 2011), nor did a 13-year follow-up of individuals from the SU.VI.MAX study (Kesse-Guyot et al., 2012). Omega-3 fatty acid supplementation has, however, been shown to be effective in the treatment of depression in several small clinical trials (Freeman et al., 2011, Rondanelli et al., 2011, Sinn et al., 2012, Sublette et al., 2011, Tajalizadekhoob et al., 2011). Further, a recent meta-analysis determined EPA-rich oils appear to be more effective in treating depression than DHA-rich oils (Sublette et al., 2011).

Given the inconsistency of the literature and the heterogeneous nature of depression, additional research is warranted to better understand the role of omega-3 fatty acids in specific populations with depression. To further examine this issue, we evaluated the relationship between omega-3 fatty acid levels and depression in postmenopausal women participating in the Women's Health Initiative Memory Study (WHIMS). Using longitudinal data from the WHIMS cohort (aged 63 to 81 years at baseline), we assessed the relationship between baseline levels of red blood cell (RBC) membrane DHA+EPA - a validated biomarker of tissue omega-3 status (Harris et al., 2004), and subsequent onset of depressive symptoms. We predicted that the risk of developing symptoms of depression in post-menopause decreases as erythrocyte membrane concentration of DHA+EPA increases. We additionally sought to determine whether a similar association exists between depression and omega-3 dietary intake as assessed by food frequency questionnaire.

Methods

Sample

Subjects were 7,086 participants of the Women's Health Initiative Memory Study, a WHI ancillary study, for whom baseline erythrocyte fatty acid concentration values were available. Baseline screening for the Women's Health Initiative began in 1993 and occurred over the course of six months, consisting of three visits in total. Blood samples used for fatty acid analyses were collected at the first WHI screening visit, on average 2.7 months prior to randomization to hormone therapy (estrogen-only or estrogen plus progestin) or placebo.

Assessment of Depressive Symptoms

Depressive symptoms were assessed via the Burnam 8-item scale for depressive disorders- a combined Center for Epidemiologic Studies/Diagnostic Interview Schedule short form (CES-D/DIS short form). Baseline depressive symptoms were assessed at the second screening visit and follow-up assessment occurred at study closeout, a mean (sd) of 90 (7) months after the baseline assessment. We selected a priori for primary analysis a score of 0.06 as the established cut-point for presence of depressive symptoms, in accordance to a prior study by Bertone-Johnson et al. (Bertone-Johnson et al., 2011). The CES-D/DIS short form was selected over the Geriatric Depression Scale (GDS), which was measured in only a small subset of participants and not concurrently. A previous study determined the Burnam Scale to provide an accurate assessment of presence of depressive symptoms (sensitivity 74%, specificity 87%), when evaluated against the Structured Clinical Interview for DSM-IV (Tuunainen et al., 2001).

For the assessment of new depression at follow-up, prevalent cases at baseline were excluded from analysis. New depression at follow-up was used in lieu of incident depression, as it cannot be determined based on the available data whether these are true incident cases or if the participant has experienced a previous depression episode outside of the data collection time points.

In a secondary analysis, we expanded our primary outcome definition to include participants with the primary Burnam threshold (0.06) or new antidepressant use, which was not considered in our primary definition of depressive symptomatology due to the potential for antidepressant medications to be prescribed off-label for the treatment of vasomotor symptoms following menopause. As a sensitivity analysis, we expanded our primary outcome definition to include participants with a Burnam score of 0.009 or higher (Burnam et al., 1988). Case counts for all analyses are available in Table 1.

Table 1.

Case counts by Depression Measure

Prevalent Depression New Depression at Follow-up

Depression Measure n (yes/no) % n (yes/no) %
Burnam Score 0.06 542/6544 7.65 319/6225 4.87
Burnam Score 0.06 or Antidepressant 967/6119 13.65 800/5319 13.07
Burnam Score 0.009 1350/5736 19.05 603/5133 10.51

Fatty Acid Measurement

For both RBC membrane and dietary intake levels, our a priori primary exposure indicator was DHA + EPA. Secondary analyses were conducted with total omega-3 [combined alpha-linolenic (ALA), docosapentaenoic acid, DHA and EPA], DHA-only, and EPA-only. Subjects included in the analysis had a mean (sd) RBC concentration of 5.26 (1.52) % for DHA+EPA, 4.56 (1.27) % for DHA-only, 0.71 (0.35) % for EPA-only, and 7.96 (1.69) % for total omega-3.

Omega-3 fatty acid content in the central nervous system may be relevant in the development of depression in post-menopausal women - fatty acid concentrations in the red blood cell membrane, while not capturing CNS concentrations, are likely to correspond (Connor et al., 1990) and can be obtained by routine blood draw, a procedure far less invasive than brain tissue biopsy; as such, we selected RBC omega-3 fatty acid levels as our primary measurement. Blood samples were collected by the respective WHI field centers at the first screening visit and stored at −80 °C by Fisher BioServices in Rockville, MD, the central repository for WHI blood samples. It was determined that the blood samples available for analyses had undergone some fatty acid oxidation as a result of short-term exposure to temperatures of −20 °C during sample aliquoting (Pottala et al., 2012a). Multiple imputation was used to correct for sample degradation to provide a more accurate depiction of RBC fatty acid composition.

Dietary fatty acid intake was assessed via the Women's Health Initiative Food Frequency Questionnaire. This diet assessment tool was originally created for use in the low-fat dietary modification arm of the WHI clinical trial, and as such was designed to provide an accurate measure of dietary fat intake (Patterson et al., 1999). Data on omega-3 fatty acid supplement use was not available. Subjects included in the analysis had a mean (sd) dietary intake of 0.11 (0.11) grams/day for DHA+EPA, 0.07 (0.07) grams/day for DHA-only, 0.04 (0.04) grams/day for EPA-only, and 1.38 (0.81) grams/day for total omega-3.

Statistical Analysis

All statistical analyses were performed using SAS 9.3 (SAS Institute Inc., Cary, NC). In order to account for the effects of degradation due to storage at −20 degrees, each RBC fatty acid value was replaced with ten imputed probable values, derived using methods described in greater detail elsewhere (Pottala et al., 2012a). For the purpose of baseline reporting, fatty acid values represent the mean imputed value from the screening visit blood sample. The multiple imputations were used in logistic regression analyses described below. Even after imputation, 3.1% of the WHIMS study sample had blood samples that remained technically unacceptable, defined as having a highly unsaturated to saturated fatty acid ratio of 0.52 or below, and were excluded from the analysis (Pottala et al., 2012a).

Participants were contrasted on a variety of sociodemographic and clinical variables by case status, using chi-square tests for categorical variables and the Wilcoxon rank-sum test for continuous variables. The covariates considered for inclusion are those that are associated with depression or omega-3 status based on WHIMS data or the available literature. Logistic regression was used to assess the relationships between fatty acid measures and new depression at follow-up. Regression analysis was conducted first with only age and WHI hormone treatment assignment as covariates (crude model), and repeated with more extensive adjustment. In addition to age and WHI hormone treatment assignment, the multivariable-adjusted model also included: U.S. region, marital status, race, education, income, obesity, high blood pressure, high cholesterol, smoking status, bilateral oophorectomy, arthritis, cardiovascular disease, diabetes, prescription narcotic use, physical activity, and alcohol consumption. Analysis of new depression at follow-up also adjusted for baseline Burnam score and number of days between screening and follow-up. A reduced model included: U.S. region, marital status, race, education, and income, in addition to age and WHI hormone treatment assignment; however, the results of these analyses did not appreciably differ from the fully adjusted models, and for clarity of presentation are not reported.

To assess for potential non-linear effects, our primary new-onset depression models were reevaluated using tertile categories for each erythrocyte membrane and dietary DHA+EPA, DHA, EPA, and total omega-3, with the lowest tertile as the reference category. Additionally, we modeled each RBC membrane omega-3 fatty acid measurement for a linear + quadratic effect; however the results of such were not contributory, and thus are not shown.

Secondarily, we examined the association between depressive symptoms, both present at baseline and new at follow-up, and fatty acid intake as assessed by food frequency questionnaire, to render comparison between methods of measurement of omega-3 status. Dietary data was not used for primary analyses on the assumption that biological concentrations were more reliable and relevant.

Results

Prevalent Depression at Baseline

Descriptive statistics for the study cohort by depression status can be seen in greater detail in Table 2. Participants reporting depression at baseline were more likely to have lower educational and income status, live in the south and northeast, be divorced or widowed, have participated in the estrogen-alone hormone trial, smoke, drink less alcohol, engage in less physical activity, and have lower RBC total omega-3 levels and higher dietary total omega-3 intake, relative to participants below the Burnam score cut-point for depressive symptoms.

Table 2.

Demographic and Selected Risk Factors by Baseline Depression

Depression (as indicated by Burnam score >0.06)
Characteristics Missing n Yes No Total p-value

0 7086 542 (7.7%) 6544 (92.3%) 100%
Age 0 7086 70 70 70 0.252
Ethnicity 15 0.059
   American Indian 26 3 (0.6%) 23 (0.3%) 0.4%
   Asian 120 8 (1.5%) 112 (1.7%) 1.7%
   African-American 497 53 (9.8%) 444 (6.8%) 7.0%
   Hispanic 166 18 (3.3%) 148 (2.3%) 2.4%
   White 6164 450 (83.5%) 5714 (87.5%) 87.2%
   Other 98 7 (1.3%) 91 (1.4%) 1.4%
School past high school? 20 <0.001
   Yes 4958 338 (62.8%) 4620 (70.8%) 70.2%
   No 2108 200 (37.2%) 1908 (29.2%) 29.8%
Income 429 <0.001
   <$50,000 5190 429 (85.8%) 4761 (77.3%) 78.0%
   $50,000+ 1467 71 (14.2%) 1396 (22.7%) 22.0%
U.S. region 0 0.014
   Northeast 1928 171 (31.5%) 1757 (26.8%) 27.2%
   South 1507 124 (22.9%) 1383 (21.1%) 21.3%
   Midwest 1682 103 (19.0%) 1579 (24.1%) 23.7%
   West 1969 144 (26.6%) 1825 (27.9%) 27.8%
Marital status 16 <0.001
   Never married 234 13 (2.4%) 221 (3.4%) 3.3%
   Divorced 889 81 (14.9%) 808 (12.4%) 12.6%
   Widowed 2195 220 (40.6%) 1975 (30.4%) 31.1%
   Married 3689 225 (41.5%) 3464 (52.9%) 52.2%
   S.O, not married 63 3 (0.5%) 60 (0.9%) 0.89%
Obese 0.057
   Yes 41 2368 201 (37.4%) 2167 (33.3%) 33.6%
   No 4677 336 (62.6%) 4341 (66.7%) 66.4%
High blood pressure 0.780
   Yes 0 2392 180 (33.2%) 2212 (33.8%) 33.8%
   No 4694 362 (66.8%) 4332 (66.2%) 66.2%
High Cholesterol
   Yes 90 1274 104 (19.4%) 1170 (18.1%) 18.2% 0.449
   No 5722 431 (81.9%) 5291 (80.6%) 81.8%
Hormone treatment 0 0.003
   E-alone intervention 1387 124 (22.8%) 1263 (19.3%) 19.6%
   E-alone control 1399 129 (23.8%) 1270 (19.4%) 19.7%
   E+P-intervention 2118 146 (26.9%) 1972 (30.1%) 29.9%
   E+P-control 2182 143 (26.4%) 2039 (31.2%) 30.8%
Smoking status 105 <0.001
   Non-smoker 3709 266 (50.1%) 3443 (53.4%) 53.1%
   Former smoker 2781 200 (37.7%) 2581 (40.0%) 39.8%
   Current smoker 491 65 (12.2%) 426 (6.6%) 7.1%
Alcohol (servings/wk) 16 7070 1.77 2.43 2.38 0.002
Physical Activity (MET hr/wk) 16 7070 9.53 11.52 11.37 <0.001
RBC total n-3 (%) 0 7086 7.80 7.98 7.96 0.002
total n-3 dietary intake (g) 8 7078 1.48 1.37 1.38 0.015

At baseline, in models adjusted only for age and WHI hormone treatment assignment, RBC DHA+EPA, as well as total omega-3 and DHA were associated with depressive symptoms as measured by Burnam Score ≥ 0.06 or antidepressant use (Table 3). However, after adjusting for demographic and health behavior characteristics, no significant association was found between any RBC omega-3 fatty acid metric (DHA+EPA, total omega-3, DHA-only, or EPA-only) and depression by any measure (Burnam Score ≥ 0.06, Burnam Score ≥ 0.06 or antidepressant use, Burnam Score ≥ 0.009).

Table 3.

Baseline Depression by RBC and Dietary Omega-3

RBC (per 1% increase) Dietary Intake (per 100mg increase)

Depression Measure1 Crude OR (95% CI)2 p-value Adjusted OR (95% CI)3 p-value Crude OR (95% CI)2 p-value Adjusted OR (95% CI)3 p-value
Total Omega3
Burnam Score 0.06 0.952 (0.904, 1.003) 0.0662 0.985 (0.922, 1.052) 0.6565 1.018 (1.006, 1.025) 0.0021 1.017 (1.006, 1.028) 0.0027
Burnam Score 0.06 or Antidepressant 0.938 (0.900, 0.977) 0.0023 0.978 (0.926, 1.033) 0.4189 1.008 (1.000, 1.016) 0.0532 1.010 (1.001, 1.019) 0.0310
DHA
Burnam Score 0.06 0.950 (0.888, 1.016) 0.1367 0.975 (0.894, 1.063) 0.5591 1.059 (0.953, 1.176) 0.2861 1.128 (1.011, 1.257) 0.0306
Burnam Score 0.06 or Antidepressant 0.930 (0.879, 0.983) 0.0102 0.970 (0.901, 1.043) 0.4040 1.022 (0.938, 1.113) 0.6221 1.103 (1.005, 1.211) 0.0398
EPA
Burnam Score 0.06 0.845 (0.624, 1.145) 0.2743 1.011 (0.713, 1.435) 0.9486 1.095 (0.889, 1.349) 0.3931 1.226 (0.982, 1.531) 0.0723
Burnam Score 0.06 or Antidepressant 0.805 (0.627, 1.035) 0.0892 0.998 (0.761, 1.308) 0.9891 1.027 (0.867, 1.217) 0.7547 1.192 (0.987, 1.439) 0.0682
DHA+EPA
Burnam Score 0.06 0.954 (0.899, 1.012) 0.1180 0.982 (0.910, 1.059) 0.6338 1.037 (0.966, 1.114) 0.3120 1.082 (1.004, 1.166) 0.0378
Burnam Score 0.06 or Antidepressant 0.936 (0.893, 0.982) 0.0069 0.977 (0.918, 1.040) 0.4607 1.013 (0.956, 1.073) 0.6607 1.068 (1.002, 1.138) 0.0440
1

depression (yes/no) n = 542/6544 (Burnam Score >0.06), 967/6119 (Antidepressant or Burnam Score >0.06)

2

crude model adjusted for age and WHI hormone treatment assignment

3

multivariable-adjusted model adjusted for U.S. region, marital status, race, education, income, obesity, high blood pressure, high cholesterol, smoking status, bilateral oophorectomy, arthritis, cardiovascular disease, diabetes, prescription narcotic use, physical activity, and alcohol consumption

In adjusted models, dietary intakes of total omega-3, DHA, and DHA+EPA were associated with a higher prevalence of depressive symptoms at baseline using any of the 3 definitions for depression. Dietary intake of EPA was marginally associated with a higher prevalence of depressive symptoms at baseline when depressive symptoms were defined as Burnam Score ≥ 0.06 (p=0.072) and as Burnam Score ≥ 0.06 or antidepressant use (p=0.068) (Table 3). In the sensitivity analysis using depression defined as a Burnam Score Score ≥ 0.009, the adjusted OR estimates did not substantively differ although dietary intake of EPA became significantly associated with depression in adjusted models (aOR 1.232, 95% C.I. 1.048–1.448, p=0.011).

New Depression at Follow-up

After exclusion of prevalent cases at baseline, no associations were found for any of the RBC omega-3 fatty acid levels and the development of depressive symptoms by any of the three definitions during the follow-up period (Table 4), nor were associations between most measures of dietary omega-3 fatty acid intake and new depression identified, with the exception of an association between greater dietary intake of total omega-3 and an increasing risk of depressive symptoms as defined by the secondary measure of Burnam score ≥0.06 or starting an antidepressant. In the sensitivity analysis, the results using a Burnam score threshold of ≥0.009 did not substantively differ from the a priori threshold of ≥0.06.

Table 4.

New Depression at Follow-up by RBC and Dietary Omega-3 Intake

RBC (per 1% increase) Dietary Intake (per 100mg increase)

Depression Measure1 Crude OR (95% CI)2 p-value Adjusted OR (95% CI)3 p-value Crude OR (95% CI)2 p-value Adjusted OR (95% CI)3 p-value
Total omega3
Burnam Score 0.06 0.976 (0.910, 1.052) 0.5273 0.958 (0.872, 1.052) 0.3661 1.004 (0.989, 1.018) 0.6199 1.005 (0.988, 1.021) 0.5852
Burnam Score 0.06 or Antidepressant 0.960 (0.920, 1.008) 0.1005 1.007 (0.950, 1.072) 0.8282 1.009 (1.000, 1.018) 0.0543 1.015 (1.003, 1.026) 0.0118
DHA
Burnam Score 0.06 0.965 (0.880, 1.065) 0.4788 0.930 (0.828, 1.045) 0.2212 0.935 (0.791, 1.105) 0.4306 0.947 (0.786, 1.142) 0.5716
Burnam Score 0.06 or Antidepressant 0.950 (0.890, 1.008) 0.0854 0.993 (0.920, 1.076) 0.8555 0.943 (0.847, 1.051) 0.2888 1.027 (0.905, 1.165) 0.6823
EPA
Burnam Score 0.06 0.982 (0.690, 1.388) 0.9156 1.013 (0.676, 1.519) 0.9503 0.864 (0.625, 1.196) 0.3783 0.869 (0.598, 1.263) 0.4613
Burnam Score 0.06 or Antidepressant 0.830 (0.660, 1.053) 0.1252 1.053 (0.770, 1.436) 0.7408 0.908 (0.737, 1.119) 0.3645 1.078 (0.837, 1.387) 0.5616
DHA+EPA
Burnam Score 0.06 0.973 (0.900, 1.057) 0.5162 0.947 (0.856, 1.048) 0.2918 0.953 (0.852, 1.067) 0.4043 0.960 (0.845, 1.090) 0.5260
Burnam Score 0.06 or Antidepressant 0.950 (0.900, 1.003) 0.0630 0.998 (0.930, 1.070) 0.9446 0.963 (0.896, 1.035) 0.3046 1.021 (0.937, 1.112) 0.6349
1

depression (yes/no) n = 319/6225 (Burnam Score >0.06), 800/5319 (Antidepressant or Burnam Score >0.06)

2

crude model adjusted for age, WHI hormone treatment assignment, and baseline Burnam Score

3

multivariable-adjusted model adjusted for U.S. region, marital status, race, education, income, obesity, high blood pressure, high cholesterol, smoking status, bilateral oophorectomy, arthritis, cardiovascular disease, diabetes, prescription narcotic use, physical activity, alcohol consumption, days to follow-up

Omega-3 Tertile Analysis

No associations were found between RBC or dietary omega-3 fatty acid measures and baseline depressive symptoms (data not shown), nor was an association found for any measure of RBC omega-3 and new-onset depression over the follow-up period (Table 5). Subjects in the third tertile of dietary DHA+EPA intake (0.12g – 2.03g) were at a 0.707 (95% C.I. 0.503– 0.993, p= 0.04) odds of developing depressive symptoms during the follow-up period, relative to subjects in the first tertile (0g – 0.05g). RBC DHA+EPA percent concentration in the third tertile relative to the first tertile (OR 0.690, C.I. 0.460–1.036) was similarly suggestive of an inverse association with new depression at follow-up; however the results did not cross the threshold of significance.

Table 5.

New Depression at Follow-up by RBC Membrane and Dietary Omega-3 Tertiles

RBC % Dietary Intake (g/d)

Tertile range Depression1 (yes/no) Crude OR (95% CI) Adjusted OR (95% CI) Tertile range Depression1 (yes/no) Crude OR (95% CI) Adjusted OR (95% CI)
Total omega3 3.9077–7.1081 119/2012 --ref-- --ref-- 0.0890–0.9508 91/2086 --ref-- --ref--
7.1081–8.3968 111/2118 0.884 (0.560, 1.393) 0.835 (0.478, 1.461) 0.9508–1.5178 115/2127 1.181 (0.885, 1.576) 1.239 (0.891, 1.722)
8.3968–19.8454 89/2095 0.814 (0.560, 1.181) 0.734 (0.481, 1.120) 1.5178–8.0049 112/2005 1.183 (0.883, 1.584) 1.235 (0.882, 1.729)
p-value 0.2733 0.1528 0.4408 0.3683
DHA 1.7514–3.9037 118/2012 --ref-- --ref-- 0.0000–0.0345 113/2018 --ref-- --ref--
3.9037–4.9424 109/2116 0.883 (0.600, 1.292) 0.804 (0.488, 1.325) 0.0345–0.0781 104/2123 0.858 (0.649, 1.134) 0.824 (0.601, 1.129)
4.9424–11.6267 92/2097 0.787 (0.530, 1.179) 0.667 (0.432, 1.029) 0.0781–1.3119 101/2077 0.759 (0.570, 1.012) 0.718 (0.512, 1.006)
p-value 0.2481 0.0661 0.1681 0.1495
EPA 0.0428–0.5386 110/2026 --ref-- --ref-- 0.0000–0.0172 111/2024 --ref-- --ref--
0.5386–0.7359 109/2120 0.994 (0.640, 1.554) 0.920 (0.543, 1.557) 0.0172–0.0411 107/2113 0.892 (0.675, 1.179) 0.826 (0.602, 1.134)
0.7359–5.8678 100/2079 0.944 (0.660, 1.351) 0.909 (0.596, 1.396) 0.0411–0.7924 100/2081 0.786 (0.590, 1.048) 0.740 (0.530, 1.032)
p-value 0.7539 0.6484 0.2594 0.1937
DHA+EPA 2.0446–4.4680 118/2009 --ref-- --ref-- 0.0000–0.0527 111/2020 --ref-- --ref--
4.4680–5.6498 107/2122 0.863 (0.590, 1.256) 0.803 (0.535, 1.205) 0.0527–0.1193 110/2116 0.917 (0.695, 1.210) 0.861 (0.629, 1.178)
5.6498–15.6179 94/2093 0.798 (0.560, 1.130) 0.690 (0.460, 1.036) 0.1193–2.0328 97/2082 0.753 (0.563, 1.007) 0.707 (0.503, 0.993)
p-value 0.2188 0.0715 0.1511 0.1354
1

Burnam Score > 0.06

Discussion

Our primary analysis failed to demonstrate an association between RBC DHA+ EPA and new depressive symptoms, assessed after an average of 7.5 years of follow-up; nor was evidence of a relationship between RBC total omega-3, DHA, or EPA found for baseline or new depressive symptoms. Dietary intakes of DHA+EPA, DHA, and total omega-3at baseline were associated with a modestly increased risk of depressive symptoms by any measure, with similar trends for EPA. This positive relationship between dietary omega-3 intake and a greater prevalence of depressive symptoms at baseline is likely attributable to reverse causality, considering no similar trend is seen in follow-up analysis. In a secondary categorical analysis of new depression at follow-up, the highest tertile of dietary, but not RBC, DHA+EPA was associated with a lower risk of depression, suggesting that dietary consumption is only one factor in determining body omega-3 concentration and that there may be other biological variables influencing the use and conversion of dietary omega-3 to tissue omega-3.

Our findings may help begin to explain inconsistencies in the existing literature on omega-3 fatty acids and depression, particularly among post-menopausal women. While RBC omega-3 levels were not associated with depressive symptoms, the reduced odds of new depression at follow-up in those with dietary DHA+EPA in the highest tertile of intake relative to the lowest tertile would be consistent with that of previous studies in a variety of populations (Ali et al., 2009, Amin et al., 2008, Baghai et al., 2011, Freeman et al., 2011, Park et al., 2012, Pottala et al., 2012b). The Freeman et al. trial supplemented with EPA and DHA at 2 g/day, which although somewhat higher, corresponds to the third tertile of dietary DHA+EPA intake noted in our study (0.12g – 2.03g). The complex relationship seen between dietary measures and our depression measures may also explain some of the variability in the literature. Although a growing body of literature suggests a strong correlation between dietary measures and biological concentrations of omega-3 fatty acids (Block et al., 2008, Harris et al., 2012, Harris et al., 2007, Sala-Vila et al., 2011, Salisbury et al., 2011, Sands et al., 2005), this relationship may be confounded by a variety of factors. Among our study participants, dietary DHA+EPA was only moderately correlated with RBC DHA+EPA (r = 0.44); however this value only takes into consideration the linear association between RBC and dietary DHA+EPA – the true relationship between dietary and body levels of omega-3 fatty acids may be much more complex.

Our biological measures experienced some degradation due to inconsistency in storage, which represents a limitation that we attempted to mitigate through multiple imputation. For the analysis of dietary omega-3 fatty acids, supplement use was not captured, which may obscure its true association with depressive symptoms, as well as the correlation between RBC and dietary omega-3 fatty acid levels.

Measurement of depressive symptoms, obtained via the 8-item Burnam scale for depressive symptoms, was assessed at only two time points and we were thus unable to assess onset of depressive symptoms as a time-to-event variable in survival analysis. Further, the Burnam scale is not sufficient for identification of treated depression or for the diagnosis of Major Depressive Disorder, which is ideally made by a structured clinical interview; however, the tool has relatively high sensitivity and specificity to detect Major Depression and can identify the presence of clinically significant depressive symptoms.

Our study is the largest prospective study of its kind to date (Conklin et al., 2010, Hakkarainen et al., 2004, Hoffmire et al., 2012, Huan et al., 2004, Lucas et al., 2011, Sanchez-Villegas et al., 2007, Sublette et al., 2006, Suominen-Taipale et al., 2010, Tanskanen et al., 2001, Tiemeier et al., 2003, Timonen et al., 2004). Many previous studies assessed dietary indicators of omega-3 fatty acids, although percent concentration within the RBC membrane more accurately reflects body omega-3 levels - few studies integrate both and fewer still utilize a longitudinal design. The baseline prevalence of depressive symptoms among study participants was 7.65% (n=542), which corresponds well to what might be expected from data in representative populations (Kessler et al., 2003). Further, among our study sample, crude mean RBC membrane concentration of DHA+EPA was 5.3%, also approximating that expected (Harris et al., 2012). We attempted to address confounding through use of multivariable models, controlling for a variety of clinical and sociodemographic variables; however, exposure is not randomized in this observational study and potential for residual confounding persists. To control for type 1 error, we opted for a single a priori primary model for analysis - secondary analyses did not control for multiple comparisons and should be considered exploratory. The presence of biological assessments of omega-3 concentrations in a large, well-characterized sample represents a notable strength of this study. Another major strength of the Women's Health Initiative is its prospective design, which allowed us to identify incident cases and reasonably account for temporality.

Overall, our findings do not support a linear association between depressive symptoms and omega-3 fatty acid levels in the RBC membrane. Additionally, conflicting relationships between dietary omega-3 and depressive symptoms were found, although tertile analysis does suggest that higher intake of DHA+EPA may be associated with a lower risk of depression. Future studies are needed to further evaluate for any potential protective effects of omega-3 fatty acids on depression in this and other populations.

Acknowkedgements

The manuscript has been approved by the Women's Health Initiative (WHI) Publications and Presentations Committee. The following is a short list of WHI investigators:

Program Office: (National Heart, Lung, and Blood Institute, Bethesda, Maryland) Jacques Rossouw, Shari Ludlam, Dale Burwen, Joan McGowan, Leslie Ford, and Nancy Geller

Clinical Coordinating Center: Clinical Coordinating Center: (Fred Hutchinson Cancer Research Center, Seattle, WA) Garnet Anderson, Ross Prentice, Andrea LaCroix, and Charles Kooperberg

Investigators and Academic Centers: (Brigham and Women's Hospital, Harvard Medical School, Boston, MA) JoAnn E. Manson; (MedStar Health Research Institute/Howard University, Washington, DC) Barbara V. Howard; (Stanford Prevention Research Center, Stanford, CA) Marcia L. Stefanick; (The Ohio State University, Columbus, OH) Rebecca Jackson; (University of Arizona, Tucson/Phoenix, AZ) Cynthia A. Thomson; (University at Buffalo, Buffalo, NY) Jean Wactawski-Wende; (University of Florida, Gainesville/Jacksonville, FL) Marian Limacher; (University of Iowa, Iowa City/Davenport, IA) Robert Wallace; (University of Pittsburgh, Pittsburgh, PA) Lewis Kuller; (Wake Forest University School of Medicine, Winston-Salem, NC) Sally Shumaker

Women's Health Initiative Memory Study: (Wake Forest University School of Medicine, Winston-Salem, NC) Sally Shumaker

Dr. Robinson has received Research grants from Amarin, Amgen, Astra-Zeneca, Daiichi-Sankyo, Esperion, Genetech/Hoffman La Roche, Glaxo-Smith Kline, Merck, Regeneron/Sanofi. Dr. Harris owns OmegaQuant Analytics, LLC (Sioux Falls, SD) and is a Senior Research Scientist at Health Diagnostic Laboratory, Inc. (Richmond, VA), both of which offer the red blood cell fatty acid test used in this study. He is also a scientific advisor to Omthera Pharmaceuticals and Aker Biomarine Antarctic. Dr. Manson and colleagues at Brigham and Women's Hospital are conducting the Vitamin D and Omega-3 Trial, which is supported by the National Institutes of Health with study pills, matching placebos, and packaging donated by Pronovo BioPharma of Norway (Omacor fish oil) and Pharmavite of Northridge California (vitamin D3).

Financial Support: The WHI program is funded by the National Heart, Lung, and Blood Institute, National Institutes of Health, U.S. Department of Health and Human Services through contracts HHSN268201100046C, HHSN268201100001C, HHSN268201100002C, HHSN268201100003C, HHSN268201100004C, and HHSN271201100004C. Dr. Fiedorowicz is supported by the National Institutes of Health (1K23MH083695-01A210).

Footnotes

For a list of all the investigators who have contributed to WHI science, please visit: https://cleo.whi.org/researchers/SitePages/Write%20a%20Paper.aspx

Conflict of Interest All other authors (JEP, EMA, WHC, MAE, JGF) have no conflicts to disclose.

References

  1. Ali S, Garg SK, Cohen BE, Bhave P, Harris WS, Whooley MA. Association between omega-3 fatty acids and depressive symptoms among patients with established coronary artery disease: data from the Heart and Soul Study. Psychother Psychosom. 2009;78:125–7. doi: 10.1159/000203118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Amin AA, Menon RA, Reid KJ, Harris WS, Spertus JA. Acute coronary syndrome patients with depression have low blood cell membrane omega-3 fatty acid levels. Psychosom Med. 2008;70:856–62. doi: 10.1097/PSY.0b013e318188a01e. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Baghai TC, Varallo-Bedarida G, Born C, Hafner S, Schule C, Eser D, Rupprecht R, Bondy B, von Schacky C. Major depressive disorder is associated with cardiovascular risk factors and low Omega-3 Index. J Clin Psychiatry. 2011;72:1242–7. doi: 10.4088/JCP.09m05895blu. [DOI] [PubMed] [Google Scholar]
  4. Bertone-Johnson ER, Powers SI, Spangler L, Brunner RL, Michael YL, Larson JC, Millen AE, Bueche MN, Salmoirago-Blotcher E, Liu S, Wassertheil-Smoller S, Ockene JK, Ockene I, Manson JE. Vitamin D intake from foods and supplements and depressive symptoms in a diverse population of older women. Am J Clin Nutr. 2011;94:1104–12. doi: 10.3945/ajcn.111.017384. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Block RC, Harris WS, Pottala JV. Determinants of Blood Cell Omega-3 Fatty Acid Content. Open Biomark J. 2008;1:1–6. doi: 10.2174/1875318300801010001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Burdge GC, Wootton SA. Conversion of alpha-linolenic acid to eicosapentaenoic, docosapentaenoic and docosahexaenoic acids in young women. Br J Nutr. 2002;88:411–20. doi: 10.1079/BJN2002689. [DOI] [PubMed] [Google Scholar]
  7. Burnam MA, Wells KB, Leake B, Landsverk J. Development of a brief screening instrument for detecting depressive disorders. Med Care. 1988;26:775–89. doi: 10.1097/00005650-198808000-00004. [DOI] [PubMed] [Google Scholar]
  8. Conklin SM, Runyan CA, Leonard S, Reddy RD, Muldoon MF, Yao JK. Age-related changes of n-3 and n-6 polyunsaturated fatty acids in the anterior cingulate cortex of individuals with major depressive disorder. Prostaglandins Leukot Essent Fatty Acids. 2010;82:111–9. doi: 10.1016/j.plefa.2009.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Connor WE, Neuringer M, Lin DS. Dietary effects on brain fatty acid composition: the reversibility of n-3 fatty acid deficiency and turnover of docosahexaenoic acid in the brain, erythrocytes, and plasma of rhesus monkeys. J Lipid Res. 1990;31:237–47. [PubMed] [Google Scholar]
  10. Freeman MP, Hibbeln JR, Silver M, Hirschberg AM, Wang B, Yule AM, Petrillo LF, Pascuillo E, Economou NI, Joffe H, Cohen LS. Omega-3 fatty acids for major depressive disorder associated with the menopausal transition: a preliminary open trial. Menopause. 2011;18:279–84. doi: 10.1097/gme.0b013e3181f2ea2e. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hakkarainen R, Partonen T, Haukka J, Virtamo J, Albanes D, Lonnqvist J. Is low dietary intake of omega-3 fatty acids associated with depression? Am J Psychiatry. 2004;161:567–9. doi: 10.1176/appi.ajp.161.3.567. [DOI] [PubMed] [Google Scholar]
  12. Harris WS, Pottala JV, Lacey SM, Vasan RS, Larson MG, Robins SJ. Clinical correlates and heritability of erythrocyte eicosapentaenoic and docosahexaenoic acid content in the Framingham Heart Study. Atherosclerosis. 2012;225:425–31. doi: 10.1016/j.atherosclerosis.2012.05.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Harris WS, Pottala JV, Sands SA, Jones PG. Comparison of the effects of fish and fish-oil capsules on the n 3 fatty acid content of blood cells and plasma phospholipids. Am J Clin Nutr. 2007;86:1621–5. doi: 10.1093/ajcn/86.5.1621. [DOI] [PubMed] [Google Scholar]
  14. Harris WS, Sands SA, Windsor SL, Ali HA, Stevens TL, Magalski A, Porter CB, Borkon AM. Omega-3 fatty acids in cardiac biopsies from heart transplantation patients: correlation with erythrocytes and response to supplementation. Circulation. 2004;110:1645–9. doi: 10.1161/01.CIR.0000142292.10048.B2. [DOI] [PubMed] [Google Scholar]
  15. Hoffmire CA, Block RC, Thevenet-Morrison K, van Wijngaarden E. Associations between omega-3 poly-unsaturated fatty acids from fish consumption and severity of depressive symptoms: an analysis of the 2005–2008 National Health and Nutrition Examination Survey. Prostaglandins Leukot Essent Fatty Acids. 2012;86:155–60. doi: 10.1016/j.plefa.2012.03.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Huan M, Hamazaki K, Sun Y, Itomura M, Liu H, Kang W, Watanabe S, Terasawa K, Hamazaki T. Suicide attempt and n-3 fatty acid levels in red blood cells: a case control study in China. Biol Psychiatry. 2004;56:490–6. doi: 10.1016/j.biopsych.2004.06.028. [DOI] [PubMed] [Google Scholar]
  17. Kesse-Guyot E, Touvier M, Andreeva VA, Jeandel C, Ferry M, Hercberg S, Galan P. Cross-sectional but not longitudinal association between n-3 fatty acid intake and depressive symptoms: results from the SU.VI.MAX 2 study. Am J Epidemiol. 2012;175:979–87. doi: 10.1093/aje/kwr472. [DOI] [PubMed] [Google Scholar]
  18. Kessler RC, Berglund P, Demler O, Jin R, Koretz D, Merikangas KR, Rush AJ, Walters EE, Wang PS. The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R) Jama. 2003;289:3095–105. doi: 10.1001/jama.289.23.3095. [DOI] [PubMed] [Google Scholar]
  19. Lin PY, Huang SY, Su KP. A meta-analytic review of polyunsaturated fatty acid compositions in patients with depression. Biol Psychiatry. 2010;68:140–7. doi: 10.1016/j.biopsych.2010.03.018. [DOI] [PubMed] [Google Scholar]
  20. Lucas M, Mirzaei F, O'Reilly EJ, Pan A, Willett WC, Kawachi I, Koenen K, Ascherio A. Dietary intake of n-3 and n-6 fatty acids and the risk of clinical depression in women: a 10-y prospective follow-up study. Am J Clin Nutr. 2011;93:1337–43. doi: 10.3945/ajcn.111.011817. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Park Y, Kim M, Baek D, Kim SH. Erythrocyte n-3 polyunsaturated fatty acid and seafood intake decrease the risk of depression: case-control study in Korea. Ann Nutr Metab. 2012;61:25–31. doi: 10.1159/000339264. [DOI] [PubMed] [Google Scholar]
  22. Patterson RE, Kristal AR, Tinker LF, Carter RA, Bolton MP, Agurs-Collins T. Measurement characteristics of the Women's Health Initiative food frequency questionnaire. Ann Epidemiol. 1999;9:178–87. doi: 10.1016/s1047-2797(98)00055-6. [DOI] [PubMed] [Google Scholar]
  23. Pottala JV, Espeland MA, Polreis J, Robinson J, Harris WS. Correcting the effects of −20 degrees C storage and aliquot size on erythrocyte fatty acid content in the Women's Health Initiative. Lipids. 2012a;47:835–46. doi: 10.1007/s11745-012-3693-y. [DOI] [PubMed] [Google Scholar]
  24. Pottala JV, Talley JA, Churchill SW, Lynch DA, von Schacky C, Harris WS. Red blood cell fatty acids are associated with depression in a case-control study of adolescents. Prostaglandins Leukot Essent Fatty Acids. 2012b;86:161–5. doi: 10.1016/j.plefa.2012.03.002. [DOI] [PubMed] [Google Scholar]
  25. Rondanelli M, Giacosa A, Opizzi A, Pelucchi C, La Vecchia C, Montorfano G, Negroni M, Berra B, Politi P, Rizzo AM. Long chain omega 3 polyunsaturated fatty acids supplementation in the treatment of elderly depression: effects on depressive symptoms, on phospholipids fatty acids profile and on health-related quality of life. J Nutr Health Aging. 2011;15:37–44. doi: 10.1007/s12603-011-0011-y. [DOI] [PubMed] [Google Scholar]
  26. Sala-Vila A, Harris WS, Cofan M, Perez-Heras AM, Pinto X, Lamuela-Raventos RM, Covas MI, Estruch R, Ros E. Determinants of the omega-3 index in a Mediterranean population at increased risk for CHD. Br J Nutr. 2011;106:425–31. doi: 10.1017/S0007114511000171. [DOI] [PubMed] [Google Scholar]
  27. Salisbury AC, Amin AP, Harris WS, Chan PS, Gosch KL, Rich MW, O'Keefe JH, Jr., Spertus JA. Predictors of omega-3 index in patients with acute myocardial infarction. Mayo Clin Proc. 2011;86:626–32. doi: 10.4065/mcp.2011.0005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Sanchez-Villegas A, Henriquez P, Figueiras A, Ortuno F, Lahortiga F, Martinez-Gonzalez MA. Long chain omega-3 fatty acids intake, fish consumption and mental disorders in the SUN cohort study. Eur J Nutr. 2007;46:337–46. doi: 10.1007/s00394-007-0671-x. [DOI] [PubMed] [Google Scholar]
  29. Sands SA, Reid KJ, Windsor SL, Harris WS. The impact of age, body mass index, and fish intake on the EPA and DHA content of human erythrocytes. Lipids. 2005;40:343–7. doi: 10.1007/s11745-006-1392-2. [DOI] [PubMed] [Google Scholar]
  30. Sinn N, Milte CM, Street SJ, Buckley JD, Coates AM, Petkov J, Howe PR. Effects of n-3 fatty acids, EPA v. DHA, on depressive symptoms, quality of life, memory and executive function in older adults with mild cognitive impairment: a 6-month randomised controlled trial. Br J Nutr. 2012;107:1682–93. doi: 10.1017/S0007114511004788. [DOI] [PubMed] [Google Scholar]
  31. Sublette ME, Ellis SP, Geant AL, Mann JJ. Meta-analysis of the effects of eicosapentaenoic acid (EPA) in clinical trials in depression. J Clin Psychiatry. 2011;72:1577–84. doi: 10.4088/JCP.10m06634. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Sublette ME, Hibbeln JR, Galfalvy H, Oquendo MA, Mann JJ. Omega-3 polyunsaturated essential fatty acid status as a predictor of future suicide risk. Am J Psychiatry. 2006;163:1100–2. doi: 10.1176/ajp.2006.163.6.1100. [DOI] [PubMed] [Google Scholar]
  33. Suominen-Taipale AL, Partonen T, Turunen AW, Mannisto S, Jula A, Verkasalo PK. Fish consumption and omega-3 polyunsaturated fatty acids in relation to depressive episodes: a cross-sectional analysis. PLoS One. 2010;5:e10530. doi: 10.1371/journal.pone.0010530. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Tajalizadekhoob Y, Sharifi F, Fakhrzadeh H, Mirarefin M, Ghaderpanahi M, Badamchizade Z, Azimipour S. The effect of low-dose omega 3 fatty acids on the treatment of mild to moderate depression in the elderly: a double-blind, randomized, placebo-controlled study. Eur Arch Psychiatry Clin Neurosci. 2011;261:539–49. doi: 10.1007/s00406-011-0191-9. [DOI] [PubMed] [Google Scholar]
  35. Tanskanen A, Hibbeln JR, Tuomilehto J, Uutela A, Haukkala A, Viinamaki H, Lehtonen J, Vartiainen E. Fish consumption and depressive symptoms in the general population in Finland. Psychiatr Serv. 2001;52:529–31. doi: 10.1176/appi.ps.52.4.529. [DOI] [PubMed] [Google Scholar]
  36. Tiemeier H, van Tuijl HR, Hofman A, Kiliaan AJ, Breteler MM. Plasma fatty acid composition and depression are associated in the elderly: the Rotterdam Study. Am J Clin Nutr. 2003;78:40–6. doi: 10.1093/ajcn/78.1.40. [DOI] [PubMed] [Google Scholar]
  37. Timonen M, Horrobin D, Jokelainen J, Laitinen J, Herva A, Rasanen P. Fish consumption and depression: the Northern Finland 1966 birth cohort study. J Affect Disord. 2004;82:447–52. doi: 10.1016/j.jad.2004.02.002. [DOI] [PubMed] [Google Scholar]
  38. Tuunainen A, Langer RD, Klauber MR, Kripke DF. Short version of the CES-D (Burnam screen) for depression in reference to the structured psychiatric interview. Psychiatry Res. 2001;103:261–70. doi: 10.1016/s0165-1781(01)00278-5. [DOI] [PubMed] [Google Scholar]

RESOURCES