Statins (3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors) are considered first-line agents for treating hypercholesterolemia in the primary and secondary prevention of cardiovascular disease (CVD). Statins lower cholesterol by inhibiting a key enzyme that plays a central role in cholesterol production in the liver, in turn suppressing the inflammatory response to endotoxin and blunting lipopolysaccharide-induced monocyte tissue factor expression 1. Indeed, the benefits of statin use for individuals with existing coronary disease are well established; a number of past studies have demonstrated their efficacy. The Long-term Intervention with Pravastatin in Ischaemic Disease (LIPID) 2, Scandinavian 3 and Sacks et al. 4 studies have all demonstrated positive treatment effects of statins on CVD-related outcomes. However, despite their widespread use, conjecture remains regarding the risk-benefit profile of statins for the primary prevention of CVD, with emerging evidence of adverse effects of rapid lipid lowering on a range of psychological outcomes.

Data from prospective, case-control and longitudinal studies have identified low mood, aggression 5 and suicide 6 as possible side effects of cholesterol-lowering treatment. While the underlying physiology is not fully understood, authors of a review that found an association between low serum cholesterol and suicide suggest that lower serum cholesterol may be associated with reduced serotonin in the brain 7. Reductions in serotonin have been associated with an inability to suppress aggressive behavior, thus leading to poorer mental health outcomes. Such research has resulted in cholesterol-lowering agents, such as statins, being inextricably linked with adverse psychological outcomes. However, more recent data have challenged this notion.

Statin treatment has now been shown to be associated with improved psychological well-being in individuals with underlying coronary disease, independent of serum cholesterol levels. Young-Xu et al. compared 140 patients who had continuous use of statins, with 231 who did not. Statins were associated with a reduced risk of depression (OR 0.63, 95% CI 0.43 to 0.93) and anxiety (OR 0.69, 95% CI 0.47 to 0.99), after controlling for potential confounders 8. Data from several other studies conducted in hypercholesterolemic 9, cardiac 1011 and healthy 12 populations have provided further support for this association. A fourfold reduction in risk for depression among individuals taking statins has been observed 1011. Using observational data, Pasco et al. found that age-adjusted odds ratio for major depressive disorder (MDD) for statin users was 0.13 (95% CI 0.02 to 1.02) 12. This effect has been demonstrated in both the short term (9-month follow-up) 10 and long term (6-year follow-up) 11.

Using the existing evidence, a review has recently been conducted in an attempt to confirm a directional relationship between statin use and psychological outcomes. While and Keen 13 found conflicting evidence for a relationship between mood and statins, based on the findings of eight studies (prospective, observational, randomized controlled trials (RCTs)). A limitation of this review was that the authors did not meta-analyze the data in order to calculate effect sizes, therefore were unable to define the effects of statins on psychological outcomes. Other recent reviews conducted in the area of statin research that have performed a meta-analysis have failed to shed any light on the effects of statins on psychological outcomes; none have included mental health outcomes 51415. Thus, the true effect of statins on psychological wellbeing remained largely undetermined.

Using predetermined criteria, our aim was to conduct the first meta-analysis to confirm the impact of statins on psychological wellbeing, in individuals with and without hypercholesterolemia.

The literature search identified articles that measured the effects of statins on psychological outcomes (depression, anxiety, mood, psychological distress) in those with and without hypercholesterolemia. The following databases were utilized: Cochrane Central Register of Controlled Trials, PubMed, OVID, Medline, Proquest, CINAHL plus, SCOPUS, Web of Knowledge. Reference lists of relevant reviews and individual studies in this area were manually examined. Search engines (Google Scholar) were used to explore grey literature. No limits on year of publication were set. Studies were limited to those published in English. Bibliographies of extracted references were manually searched for relevant references.

The literature search was performed from February to March 2012. Using the search terms 'statins' (simvastatin, atorvastatin, fluvastatin, lovastatin, mevastatin, pitavastatin, pravastatin, rosuvastatin, cerivastatin) 'mood' 'distress' 'depress$' 'trial' in bibliographic databases (OVID, Cinahl, Scopus, Medline) and search engines revealed 1383 articles. After reviewing abstracts for relevance, 1370 did not meet inclusion criteria. Reasons for exclusion are displayed in Figure 1, according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines 21. Where uncertainty occurred, authors reached consensus by consulting with the third author. Examination of reference lists revealed one potentially relevant article, which was ultimately excluded because treatment and control groups remained deidentified due to the ongoing status of the trial at the time of publication 22. After completing this process, 13 articles remained. Further examination revealed that two studies did not report depression, anxiety, mood, or distress outcomes; one was the same study identified through previous search strategies; one evaluated a non-pharmaceutical, dietary lipid lowering intervention; one was a study without control condition; and one was the aforementioned paper in which treatment and placebo groups remained deidentified. Thus, these six articles were excluded. After completing this process, seven articles remained for inclusion.

The aim of the study was to determine the impact of statins on psychological wellbeing by conducting the first meta-analysis of its kind. Overall, pooled effects of seven RCTs demonstrated no adverse or beneficial effects of statins on psychological outcomes. These findings do not provide support for observational studies that have demonstrated either adverse 5 or protective effects of statins on psychological outcomes in healthy 12 and cardiac populations 1011. However, when a sensitivity analysis was conducted to include only those studies evaluating mood, statistically significant effects in favor of treatment were observed. Using Cohen's interpretation of effects, this effect size is considered of medium magnitude 29. While these findings should be considered with caution due to the limited number of studies included in this subanalysis, they provide some evidence to refute previous suggestions that lipid lowering medications may lead to adverse psychological outcomes. In fact, these data highlight a potential for statins to produce mood-related benefits. This finding is consistent with those RCTs conducted in hypercholesterolemia populations that have observed positive effects on mental health-related quality of life (QoL) 22.

There are several mechanisms by which these effects could be explained. There is now a body of evidence for an etiological role of inflammation and oxidative stress in the pathophysiology of depression 3031. As statins have anti-inflammatory and antioxidant properties, it has been suggested that these properties may target the inflammatory and oxidative pathways associated with the pathophysiology of the disorder 32. Alternatively, because individuals using statins are known to exhibit a decreased risk of cardiovascular events, it is possible that statin-induced mood improvements may be a function of improved QoL or a greater degree of health consciousness for individuals receiving long-term treatment 33.

Large-scale clinical trials with long-term outcomes are required to evaluate the effect of statins on psychological wellbeing of individuals with depression. When conducting our review, we were unable to identify any studies of this type. Because there is some support for the role of statins as a protective mechanism against the development of depression 10, a trial of this nature may offer a novel approach to the treatment and primary prevention of depression through exploring the role of inflammatory processes as a potential therapeutic target. Our findings provide some support for a favorable risk-benefit profile of statins in terms of mood outcomes for those with and without hypercholesterolemia. Statins were not linked to mood-related disturbances, but rather benefits, which may be due to their substantial anti-inflammatory and antioxidative properties. If statin-induced improvements in mood are observed in those with existing depression, the benefits of statins may go beyond that of reducing CVD-related outcomes to depression prevention or therapy. Not only are statins affordable and accessible, but additional benefits exist for those with comorbid conditions that are disproportionally more common in individuals with depression (for example, CVD) 34, because of their impact on shared pathophysiological processes (for example, inflammation) 31.

This meta-analysis has several limitations. Due to the fundamental nature of meta-analyses, where data from independent studies are synthesized, it is acknowledged that unexplained variances may exist, demonstrated by the high degree of heterogeneity observed between studies. For example, it is acknowledged that psychological wellbeing was assessed using different instruments that may measure different constructs. Further, only one study reported using a clinician-administered instrument; the majority of studies used self-report inventories to assess depression and/or anxiety. Therefore, it is recommended that future research pay greater attention to determining the efficacy of statins related to psychological outcomes, employing diagnostic psychiatric interviews to measure these constructs. Although time consuming, this is an accurate method for classifying mood disorders.

Additionally, it is acknowledged that the limited number of studies included in this review may have introduced bias. For example, it is likely that the file drawer effect influenced the findings of this study. This phenomenon sees published studies reporting inflated effect sizes and those yielding negative results remaining unpublished. This is of particular relevance to meta-analyses evaluating the effects of pharmacological interventions. While funnel plots could provide an indication of publication bias, they were not displayed in this paper, as it has been reported that more than ten estimates are required to reliably judge funnel plots 35. Small sample size further precluded extensive subanalyses of the data, due to limited power (for example, to determine specific effects between those with hypercholesterolemia versus established coronary disease versus healthy participants). Future research comprising a greater number of studies is required to confirm our findings.

Our findings refute previous evidence of the negative effects of statins on psychological outcomes, and provide some support for mood-related benefits. Further research comprising a greater number of studies is required to confirm the effects of this agent on psychological outcomes. RCTs could further examine the effects of statins in depressed populations.

FJ has received grant/research support from the Brain and Behaviour Research Institute, the National Health and Medical Research Council, Australian Rotary Health, the Geelong Medical Research Foundation, the Ian Potter Foundation, Eli Lilly and The University of Melbourne and has been a paid speaker for Sanofi-Synthelabo, Janssen Cilag and Eli Lilly. MB has received grant/research support from the NIH, Simons Foundation, CRC for Mental Health, Stanley Medical Research Institute, MBF, NHMRC, Beyond Blue, Geelong Medical Research Foundation, Bristol Myers Squibb, Eli Lilly, Glaxo SmithKline, Organon, Novartis, Mayne Pharma, Servier and Astra Zeneca. He has been a paid consultant for Astra Zeneca, Bristol Myers Squibb, Eli Lilly, Glaxo SmithKline, Janssen Cilag, Lundbeck and Pfizer and a paid speaker for Astra Zeneca, Bristol Myers Squibb, Eli Lilly, Glaxo SmithKline, Janssen Cilag, Lundbeck, Organon, Pfizer, Sanofi Synthelabo, Solvay and Wyeth. LJW has received grant/research support from Eli Lilly, Pfizer, The University of Melbourne, Deakin University and the National Health and Medical Research Council. JAP has received speaker fees from Amgen, Eli Lilly and Sanofi-Aventis and funding from the Geelong Region Medical Research Foundation, Barwon Health, Perpetual Trustees, the Dairy Research and Development Corporation, The University of Melbourne, the Ronald Geoffrey Arnott Foundation, ANZ Charitable Trust, the American Society for Bone and Mineral Research, Amgen (Europe) GmBH and the NHMRC.

AO'N conducted the literature search and quality assessment, performed data extraction and statistical analysis and drafted the original version and subsequent drafts of the manuscript. LS assisted with the inclusion/exclusion criteria, performed literature searching and contributed to the drafting of the manuscript. CR acted as an independent assessor and contributed to the drafting of the manuscript. KS consulted on statistical analysis and critically revised drafts of the manuscript. FJ, LJW and JAP critically revised drafts of the manuscript. MB conceptualized the paper and critically revised drafts of the manuscript. All authors read and approved the final manuscript.