If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Psychiatry & Mind-Body Interface Laboratory, China Medical University Hospital, Taichung, TaiwanSchool of Medicine, China Medical University, Taichung, TaiwanDepartment of Psychological Medicine, Institute of Psychiatry, King’s College London, London, United Kingdom
Address correspondence to Carmine M. Pariante, M.D., Ph.D., Institute of Psychiatry, King’s College London, Department of Psychological Medicine, Room 2-055, The James Black Centre, 125 Coldharbour Lane, London SE5 9NU, United Kingdom
Interferon (IFN)-α therapy for chronic hepatitis C virus infection is frequently associated with depression. The routine prophylaxis with antidepressants might expose patients to adverse effects, hence, the need for alternative preventive interventions. Omega-3 polyunsaturated fatty acids are safe and effective essential nutritional compounds used for the treatment of depression, putatively through an anti-inflammatory action. In addition, lower erythrocyte levels of omega-3 polyunsaturated fatty acids have been associated with an increased risk of IFN-induced depression.
We conducted a 2-week, double-blind, placebo-controlled trial comparing eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), and placebo for the prevention of IFN-α-induced depression. A total of 162 patients consented to participate and were randomized to the study. All of the patients completed the 2-week trial; 152 participants were followed throughout the 24 weeks of IFN-α treatment and were included in the analysis.
Compared with placebo, the incident rates of IFN-α-induced depression were significantly lower in EPA-treated but not in DHA-treated patients (10% and 28%, respectively, versus 30% for placebo, p = .037). Both EPA and DHA significantly delayed the onset of IFN-induced depression (week of onset: 12.0 and 11.7, respectively, versus 5.3 for placebo, p = .002). EPA and DHA were both well tolerated in this population. EPA treatment increased both EPA and DHA erythrocyte levels, but DHA only increased DHA erythrocyte levels.
EPA is effective in the prevention of depression in hepatitis C virus patients received IFN-α therapy. Our study confirms the notion that anti-inflammatory strategies are effective antidepressants in the context of depression associated with inflammation.
Finding the best strategy to prevent neuropsychiatric adverse effects induced by interferon (IFN)-α will improve clinical outcome and shed light on the pathogenesis of inflammation-induced depression, but previous studies have had mixed results (see below), especially in patients who receive IFN-α for chronic hepatitis C virus (HCV) infection. Chronic HCV infection is a major public health issue and has a high rate of progression to liver cirrhosis and hepatocellular carcinoma (
). However, the clinical impact of IFN-α is reduced by the common and severe neuropsychiatric adverse effects. For example, almost all the patients experience acute sickness behavior, including symptoms of fatigue, malaise, myalgia, arthralgia, anorexia, apathy, and cognitive impairment (
). The use of antidepressants is supported by the frequently cited trial, in patients with malignant melanoma, demonstrating a significant preventive effect of paroxetine, a selective serotonin reuptake inhibitor (SSRI) (
). As most patients receiving IFN-α do not develop clinically significant depression, the routine pretreatment with antidepressant drugs might expose patients to unnecessary medications; it is thus important to find alternative strategies for the prevention of IFN-α-induced depression.
Omega-3 polyunsaturated fatty acids (ω-3 or n-3 PUFAs), such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), are essential nutritional compounds with potential preventive and therapeutic effects against depression. Patients with major depressive disorder have lower levels of omega-3 PUFAs (
), have shown that omega-3 PUFAs have antidepressant effects. Taken together also with the evidence discussed below, these studies support the use of omega-3 PUFAs as an effective depression prophylactic strategy in at-risk groups, such as indeed patients taking IFN-α.
One of the hypothesized mechanisms underlying PUFAs’ antidepressant effects is their neuroprotective and anti-inflammatory action (
). Of particular relevance for the present study, we have recently demonstrated that lower DHA levels in the peripheral blood are associated with an increased risk of developing IFN-α-induced depression over the following weeks (
). Based on this and the other evidence discussed above, we have conducted this 2-week, double-blind, placebo-controlled trial to test the differential effects of the omega-3 PUFAs, EPA, and DHA against placebo in the prevention of IFN-α-induced depression.
We have specifically prescribed a short (2 weeks) intervention before IFN-α therapy, to potentially correct the lower omega-3 fatty acid levels that we had previously identified as a risk factor for the development of IFN-α -induced depression (
). Indeed, we also have measured the levels of PUFAs in the erythrocytes before and after the trial and correlated these with treatment response. In addition, we have chosen to test a prophylactic intervention that would be acceptable to most patients because of its brevity and because it would precede, and not overlap with, the IFN-α (and ribavirin) therapy. According to most studies, the active antidepressant component from omega-3 PUFAs is EPA (
Since 2005, a psychiatric team has been working together with the hepatologists to provide an integrated care package for HCV patients referred for IFN-α therapy at the Liver Centre of China Medical University Hospital, Taichung, Taiwan, where the Institutional Review Board approved the study. In the period between July 2009 and June 2012, the hepatologists identified eligible HCV patients before they started the combination therapy with peginterferon α-2b (1.5 µg per kilogram of body weight once weekly) and ribavirin (1000–1200 mg daily). Patients were excluded from this study if they had a major depressive episode at the initial assessment; a lifetime history of psychotic disorders (e.g., schizophrenia or bipolar disorder); a history of alcohol or drug dependence within 1 year before entry into the study; and evidence of any unstable chronic medical conditions (e.g., cardiovascular, endocrine, hematological, renal, or neurological diseases). The diagnoses of psychiatric disorders were based on the structured Mini-International Neuropsychiatric Interview (
). All patients who agreed to participate in this study provided their signed written informed consent before enrollment.
Study Design and Recruitment
Two hundred seven patients with HCV were screened, 162 of them consented to participate and were randomized to the study, and all of them completed the 2-week trial; 152 participants were followed throughout the 24 weeks of IFN-α treatment and were included in the analysis. For allocation of the participants following simple double-blind randomization procedures, a computer-generated list of random numbers was used. The identical capsules were prepacked in bottles and consecutively numbered according to the randomization schedule by an independent nutritionist.
Figure S1 in Supplement 1 provides a flow chart summarizing study recruitment. Ten subjects discontinued the IFN-α treatment; they did not differ from the completers in any demographic features, including gender, age, married status, education years, and past history of depression. While the noncompleters did have significantly higher baseline scores than completers in depressive symptoms (Hamilton Rating Scale for Depression [HAMD]: 8.6 ± 3.65 versus 4.5 ± 4.49; p = .018) and neurovegetative symptoms (Neurotoxicity Rating Scale [NTRS]: 52.9 ± 41.74 versus 26.9 ± 29.58; p = .010), they were equally distributed among the three groups (EPA, n = 4; DHA, n = 3; placebo, n = 3).
The subjects were randomly assigned in double-blind fashion to EPA, DHA, or placebo, administered for 2 weeks before starting IFN-α therapy. Specifically, 2 weeks before the initiation of IFN-α therapy (week –2), patients started receiving a daily treatment of five identical capsules of EPA (3.5 g/day), DHA (1.75 g/day), or placebo (high oleic oil) in single or divided administration. The experimental capsules contained concentrated EPA (700 mg), DHA (350 mg), or high oleic oil (800 mg); they weighed 1000 mg, were deodorized with orange flavor, and were supplemented with tertiary-butyl hydroquinone (.2 mg/g) and tocopherols (2 mg/g) as antioxidants. The sources of EPA, DHA, and oleic acids were, respectively, anchovy fish body oil (purchased from AK BioTech, Ulsan, Korea), algal vegetable (purchased from DSM Nutritional Products, Basel, Switzerland), and safflower oil (purchased from Aarhus Karlshamn, Hull, England).
The recruited participants were evaluated at weeks –2 (when omega-3 fatty acid prophylactic intervention started) and 0 (when the prophylactic intervention stopped and IFN-α therapy started) and during weeks 2, 4, 6, 8, 12, 16, 20, and 24 of IFN-α therapy to assess the occurrence of major depressive episode with the structured Mini-International Neuropsychiatric Interview. Sociodemographic factors, including gender, age, education, and marital status, as well as the past psychiatric history, substance use history, and family psychiatric history, were recorded at the initial assessment. Severity of depressive symptoms and of neurovegetative symptoms were measured using, respectively, the 21-item HAMD (
), both administered at weeks –2, 0, 2, 4, 6, 8, 12, 16, 20, and 24. The NTRS is a checklist questionnaire that has been frequently used for the evaluation of neuropsychiatric symptoms related to cytokine therapy; the items are categorized into general symptoms, nonpainful somatic symptoms, and painful somatic symptoms, with each item rated from 0 to 10 on a visual analog scale, and the final score ranging 0 to 390 (
). During IFN-α therapy, allowable concomitant medications included acetaminophen and other nonsteroidal anti-inflammatory agents for pain symptoms and fever; granisetron or ondansetron for nausea; lorazepam for severe anxiety; and zolpidem for insomnia. The results of routine biochemical laboratory examinations, the occurrence of adverse effects, and any reason for IFN-α discontinuation were recorded.
Fatty acid composition of erythrocyte membranes was analyzed by thin-layer chromatography, and the level of individual fatty acid was measured with gas chromatography of methyl esters (Lipid Standards, FAMEs, Sigma Co., St. Louis, Missouri). Fatty acid profiles were identified by comparing the retention times with those of appropriate standard fatty acid methyl esters. The detailed step-by-step procedures have been published and described elsewhere (
). The levels of each fatty acid were expressed as a percentage of total fatty acids. Laboratory measures were conducted on coded samples by workers who were blind to subjects’ information.
Data Analyses and Statistics
All participants who were recruited into the study (n = 162) completed the 2-week, double-blind, randomized-controlled trial; 152 participants were followed throughout the 24 weeks of IFN-α treatment and were included in the analysis. The primary outcome measurement of efficacy was the incidence of major depressive episode in the three groups at any time point during 24 weeks of IFN-α therapy, while the HAMD and NTRS were secondary outcome measurements. The power was calculated a priori: a sample size of 50 in each group would have had 80% of power (at p < .05) to detect a reduction of 10% in the risk of depression. The categorical data (occurrence of major depressive episode or not) were analyzed using χ2 (chi-squared) test. Changes in HAMD and NTRS scores as a function of time were assessed using repeated measures analysis of variance using mixed linear modeling, and at each time point, the mean scores of HAMD and NTRS were compared among the three groups by analysis of variance and post hoc analyses when appropriate. Furthermore, Kaplan-Meier estimates, survival curves, and log-rank test were used to compare the cumulative time free from depression between groups. All probabilities were two-tailed, with p less than .05 considered statistically significant. We used the SPSS statistical software version 15 (SPSS Inc., Chicago, Illinois).
There were no statistical differences between the three groups (EPA, DHA, or placebo) in demographics (age, gender, education, and marriage status) (Table 1), psychiatric characteristics (past history of depression, baseline HAMD and NTRS scores) (Table 1), and HCV-relevant biological markers (alanine aminotransferase and HCV genotypes and HCV RNA titers) (Table 1, Table 2).
Table 1Demographics and Clinical Characteristics Among EPA, DHA, and Placebo Groups
As shown in Table 2, the incident rates of IFN-induced depression were significantly different among EPA, DHA, and placebo groups (10% vs. 28% vs. 30%, respectively; χ2 = 6.62, p = .037). The post hoc analyses showed that incident rate was significantly lower following EPA (χ2 = 5.99, p = .014) but not following DHA (χ2 = .05, p = .8), as compared with placebo. Both EPA and DHA significantly delayed the onset of IFN-induced depression as compared with placebo (mean 12.0 vs. 11.7 vs. 5.3 weeks, respectively; F = 7.80, p = .002).
Figure 1 shows the Kaplan-Meier survival curves of IFN-induced depression comparing EPA, DHA, and placebo groups. Log-rank tests showed a significant overall effect for treatment group (χ2 = 6.74, df = 2, p = .034). Specifically, subjects receiving EPA prophylactic treatment showed a lower risk of IFN-α-induced depression as compared with placebo treatment (χ2 = 6.52, df = 1, p = .011) or DHA treatment (χ2 = 3.23, df = 1, p = .072). However, there was no significant difference between DHA and placebo groups (χ2 = .95, df = 1, p = .3). The hazard ratios were 2.9 (confidence interval: 1.0–7.9, p = .044) for the comparison between EPA and DHA and 3.5 (confidence interval: 1.3–9.6, p = .015) for the comparison between EPA and placebo.
The HAMD and NTRS scores during the IFN-α therapy are shown in Figure 2A and 2B, respectively. The mixed model analyses showed significant effects on time of repeated measurements (F = 63.8, df = 198, p < .001 for HAMD and F = 19.3, df = 184, p < .001 for NTRS), confirming the IFN-α-induced changes over time, but there were no significant effects of group (F = 1.4, df = 189, p = .2 for HAMD and F = .4, df = 153, p = .6 for NTRS) and no group by time interaction (F = 1.2, df = 198, p = .3 for HAMD and F = .7, df = 184, p = .8 for NTRS).
As shown also in Table 2, participants in both omega-3 PUFA groups did not differ from those on placebo at weeks 0 and 24 during IFN-α therapy for either HAMD or NTRS. There were significantly lower HAMD scores in the EPA group only at week 4 (compared with placebo group; p = .013) and at week 8 (compared with both DHA and placebo; p = .022 and p = .002, respectively). However, there were no significant differences between the three groups in the HAMD at weeks 2, 12, 16, and 20 and in the NTRS at any week of assessment.
Fatty Acid Biomarkers
As shown in Table 2, there was no significant difference in DHA and EPA levels between the three groups before the trial intervention (at week −2). However, there were significant effects of EPA and DHA intervention on the PUFA levels at the end of the trial (week 0) (Table 2). Specifically, the EPA intervention significantly increased EPA (+32%, p = .009) and DHA levels (+17%, p = .016). The DHA intervention significantly increased DHA (+27%, p = .002) but not EPA levels (−5%, p = .6). When the three groups were compared at week 0, analyses of variance followed by post hoc tests found that the EPA intervention group had higher levels of EPA (p = .002 vs. DHA intervention and p = .03 vs. placebo groups), while the DHA intervention group had only higher levels of DHA (p = .035 vs. EPA intervention and p = .001 vs. placebo groups). The complete profile of erythrocyte composition in fatty acids is shown in Table S1 in Supplement 1.
In the overall sample, a total of 34 patients (22%) developed IFN-α-induced depression at some point during the 24-week therapy, while 118 (78%) patients did not develop IFN-α-induced depression. At week −2 (before the intervention), subjects who later developed IFN-α-induced depression had lower baseline EPA levels (2.00 ± .63) than those who did not (2.47 ± .92; p = .031) but not DHA levels (4.17 ± 1.02 for depressed vs. 4.53 ± 1.41 for nondepressed, p = .28). However, at week 0 (before IFN-α therapy, after the 2 weeks of omega-3 PUFA intervention), there were no significant differences in EPA or DHA levels between subjects who subsequently did or did not develop IFN-α-induced depression (EPA: 2.64 ± 1.16 for depressed vs. 2.50 ± .78 for nondepressed, p = .6; DHA: 5.22 ± 1.32 for depressed vs. 5.11 ± 1.64 for nondepressed, p = .8).
EPA and DHA were found well tolerated in this HCV population. No participant was withdrawn because of adverse events by investigators’ decision, and reported events were all very mild and self-limited. There was no effect found in any blood laboratory parameter, such as abnormal bleeding time or liver function. In addition, there were no significant differences in the proportions of high HCV RNA levels (≥200,000 IU/mL) before and after IFN therapy among the three groups, showing no effects of the PUFA treatments on viral load.
To our knowledge, this is the first study to demonstrate the beneficial effects of omega-3 fatty acids in the prevention of IFN-α-induced depression. The main finding is that EPA pretreatment significantly decreased the incidence of IFN-α-induced depression in HCV patients. In addition, both EPA and DHA pretreatment significantly delayed the onset of IFN-α-induced depression as compared with placebo pretreatment. Omega-3 fatty acids have been shown to have prophylactic effects in bipolar disorder (
). Therefore, our findings confirm and extend the notion that this nutritional intervention can have preventive effects in mental health and corroborate the evidence that anti-inflammatory strategies may have antidepressant effects, especially in the context of depression associated with inflammation.
The results of the current study support our previous findings showing that omega-3 PUFAs play a role in the risk of IFN-α-induced depression (
). In our previous study, we measured the erythrocyte levels of DHA and EPA and analyzed genetic variation in the phospholipase A2 (PLA2) and cyclooxygenase 2 (COX2) genes, the two key enzymes in the metabolism of omega-3 PUFAs. We found that participants with PLA2 BanI GG or COX2 rs4648308 AG genotypes have a higher risk of IFN-α-induced depression. In addition, the at-risk PLA2 polymorphism is associated with lower EPA levels and the at-risk COX2 polymorphism is associated with lower levels of both DHA and EPA during IFN-α therapy; furthermore, in the whole sample, having lower baseline DHA levels before IFN-α therapy is associated with a higher risk of developing depression during IFN-α therapy (
EPA but not DHA appears to be responsible for the efficacy of omega-3 long chain polyunsaturated fatty acid supplementation in depression: Evidence from a meta-analysis of randomized controlled trials.
) found a dose-response effect supporting 1 g/day as superior to 2 g/day or 4 g/day, though the latter study was limited by the lack of a placebo arm. A recent meta-analysis has suggested that both EPA and DHA contribute to antidepressant effects but that the effects of EPA are stronger (
). Our current study, showing that EPA reduces the incidence of depression while DHA only delays the onset of depression, further supports this notion. Moreover, EPA can be metabolized into DHA, and EPA intervention can increase blood and brain levels of DHA (
). Indeed, it is interesting to highlight that the EPA intervention in our study increased both EPA and DHA levels. Incidentally, in our previous animal study, we demonstrated that a PUFA dietary intervention is able to increase PUFA levels in both erythrocytes and the brain, thus supporting the notion that PUFA changes measureable in the periphery reflect changes in the brain (
). These results therefore indicate possible synergetic effects of EPA and DHA on depressive symptomatology.
The anti-inflammatory action of EPA is likely to be particularly important in this context, where depression is triggered by an immune challenge. The prevailing model for IFN-α-induced depression points to a pivotal role of increased pro-inflammatory cytokines both in the periphery and in the brain (cerebrospinal fluid) of patients, with subsequent activation of the indoleamine 2,3-dioxygenase (IDO) pathway and the production of potentially depressogenic tryptophan metabolites, such as quinolinic acid (
Our findings have direct clinical relevance for HCV patients receiving IFN-α. Although prophylactic effects of SSRI antidepressants have been studied in several randomized-controlled clinical studies (
), the use of such medications needs to be weighed against the risks of adverse effects and complications. Omega-3 fatty acid intervention, however, is a safe and well-accepted alternative antidepressant treatment (
). It has been suggested that the potential antithrombotic effect of omega-3 PUFAs may theoretically increase the risk for bleeding: clinical trials have, however, shown that high-dose omega-3 PUFAs consumption is safe, even when concurrently administered with other agents that may increase bleeding, including aspirin and warfarin (
), and adding antioxidant vitamin E to omega-3 PUFAs reduces oxidation.
In this study, we chose 3.5 grams of EPA and 1.7 grams of DHA per day, because our previous studies conducted in Taiwan using PUFAs as antidepressants showed the effective EPA dose to be between 2.2 g/day and 4.4 g/day and the effective DHA dose to be between 1.2 g/day and 2.2 g/day (
). We used an animal source for EPA and a vegetal source for DHA because of the availability; therefore, we cannot exclude that the reduced effectiveness of DHA might be linked to its vegetal source. The placebo contained 15% of linoleic acid, which may exert pro-inflammatory effects and theoretically worsen depression (
). Therefore, on balance, we believe that this was a truly inactive placebo.
The fact that EPA prevents the occurrence of depression without affecting the HAMD and NTRS scores of the whole group confirms the notion that EPA main effect is confined to subjects with clinically significant depression and does not extend to subjects with no or minimal depressive symptoms (
). Indeed, patients in this study developed clinically significant depression at any time points between week 2 and week 24 and the prophylactic effects are limited to around 20% of the sample and thus too diluted to statistically affect the scores of the whole group. It is worth highlighting, however, that Figure 2 clearly shows that the scores of both scales are numerically lower in the EPA groups than in the other two groups; moreover, HAMD scores are significantly lower in the EPA group compared with placebo at weeks 4 and 8 and compared with DHA at week 8. The effects of DHA are simply less strong than those of EPA, as shown by the fact that DHA delays the onset of depression: basically, it prevents depression up to week 10 (Figure 1), but the effects are not sustained. Obviously, it is possible that both EPA and DHA effects might have been more pronounced if the treatments had been continued beyond 2 weeks. Finally, this study is limited by its relatively small sample size, the lack of information on prescription of hypnotic agents, and the fact that noncompleters had higher depression scores at baseline, although they were equally distributed among the three groups.
In conclusion, EPA was beneficial for the prevention of IFN-α-induced depression, while DHA had only modest effects on delaying the onset. The data suggest that EPA or EPA/DHA combination is the best preventive strategy in this group of patients and potentially a suitable strategy for the wider pool of patients with depression associated with inflammation.
This work was supported by the following grants:102-2911-I-039-501,101-2628-B-039-001-MY3, and101-2320-B-038-020-MY2from theNational Science Council;NHRI-EX101-10144NIfrom theNational Health Research Institute; andCMR99-114andDMR-101-081from theChina Medical University in Taiwan; National Alliance for Research on Schizophrenia and Depression Young Investigator Award; and Royal Society Joint Research Projects for Taiwan and United Kingdom. Professor Pariante is also supported by the United Kingdom Medical Research Council (MR/J002739/1) and the South London and Maudsley National Health Service Trust and King’s College Hospital National Institute for Health Research Biomedical Research Centre for Mental Health.
Drs Su and Pariante created the concept. Drs Su, Lai, Yang designed the study, acquired the data, and take responsibility for the integrity and accuracy of the data. Drs Su and Pariante drafted and revised the manuscript. All the authors had full access to all the data in the study.
Professor Pariante has received research funding from companies interested in developing anti-inflammatory strategies for depression, such as Janssen Pharmaceuticals, but the current study is unrelated to this funding. All other authors report no biomedical financial interests or potential conflicts of interest.
EPA but not DHA appears to be responsible for the efficacy of omega-3 long chain polyunsaturated fatty acid supplementation in depression: Evidence from a meta-analysis of randomized controlled trials.