Malondialdehyde concentrations in obsessive–compulsive disorder: a systematic review and meta-analysis

Background This meta-analysis aimed to investigate serum and plasma malondialdehyde (MDA) levels in patients with obsessive–compulsive disorder (OCD) in comparison to healthy controls. Methods Following the PRISMA protocol, we searched for the relevant studies through the databases of Scopus, PubMed, Google Scholar, and web of science until September 2019 with no time restriction. Overall, nine studies were included in the current meta-analysis. Data were pooled using a random-effects model; in addition, standard mean difference (SMD) and/or weight mean difference (WMD) was calculated. Cochran’s Q test and I-square (I2) statistics were used to evaluate between-study heterogeneity. The Newcastle–Ottawa scale (NOS) was used to evaluate the quality of the included studies. Statistical analyses were done using the STATA version 14. Results Our systematic review included nine case–control studies (including 367 cases and 337 controls). Pooling findings from these studies showed a significantly higher MDA level in OCD patient compared to control groups (SMD = 1.62; 95% CI [0.53, 2.72]; I2 = 96.9%; Pheterogeneity (Ph) < 0.001). This finding remained unchanged among studies which reported MDA in the same unit (WMD = 1.93; 95% CI [0.27, 3.59]; I2 = 99.2%; Ph < 0.001). Subgroup analysis by the study location and sample size revealed findings that were also significant. Conclusion We found that MDA levels are higher in OCD patients than healthy controls. This finding highlights the importance of inflammatory responses in OCD patients that should be considered for future investigations. Further studies are recommended to expand current knowledge on this issue.


Introduction
Obsessive-compulsive disorder (OCD) is characterized by intrusive, unwanted, and disturbing thoughts, images, and urges (obsessions) that elicit significant distress and repetitive behavioral or mental compulsions [1]. Such compulsions may lead to a temporary reduction in distress [2]. About 0.5-3% of people suffer from this disorder [3,4] which influences social relations, quality of life, and occupational functioning [5][6][7]. Moreover, family members of the patients who live with him/her are also affected by the disorder [8].
Oxidative stress (OXS), resulting from imbalances between the generation of oxidative free radicals in the body and neutralizing antioxidants, plays an important role in the pathophysiology of OCD [9]. OXS can cause several damages to the brain, including disruption of the membrane integrity, damages caused by the peroxidation of lipids, proteins, and nucleic acids, and changes in natural neuronal apoptosis [10]. Free oxygen radicals

Open Access
Annals of General Psychiatry *Correspondence: amkhv@yahoo.com † Ebrahim Balandeh Co-first author 2 Research Center for Biochemistry and Nutrition in Metabolic Diseases, Institute for Basic Sciences, Kashan University of Medical Sciences (KAUMS), Kashan, Iran Full list of author information is available at the end of the article have been recently acknowledged as great risk factors in the pathogenesis of several diseases, including neuropsychiatric disorders [11,12]. OXS also influences the genes responsible for various psychiatric disorders, including obsessive-compulsive disorder and other anxiety disorders [9,13]. Failure to regulate endogenous oxidative processes leads to brain insult, which may be implicated in various psychiatric disorders [14]. OXS may damage, modify, and denature the chemical structure of fatty acids [15]. Emerging evidence suggests that OXS and inflammation accelerate cellular aging and neuroprogression [16]. The brain is particularly vulnerable to the deleterious effects of OXS. Studies have linked OXS to bloodbrain barrier disruptions, altered patterns of neural growth, and changes in brain morphology [17,18]. This condition can cause psychological stress-related mental illnesses [16] and neurobehavioral disorders [19]. Therefore, the measurement of OXS in patients with OCD is important.
Malondialdehyde (MDA) is a natural product of peroxidation of unsaturated fatty acid with three or more double bonds. MDA concentration is usually measured by the thiobarbituric acid reacting substances test (TBARS), which is known as a simple and non-specific test [20]. MDA measurement is a reliable method and is known as one of the most popular methods for determining the status of OXS in clinical settings [20]. Moreover, MDA has been found in brain centers and its factors play crucial roles in the pathogenesis of neurodegenerative diseases [21]. Earlier studies have measured serum MDA levels in OCD patients compared to healthy controls [22,23]. Results suggested higher free radical metabolism in OCD and therefore tissue damage due to OXS [22]. However, few studies have summarized their findings to reach a firm conclusion. Although a meta-analysis in 2019 investigated levels of oxidative and nitrosative markers in patients with OCD [24], that study had several limitations [25] including missing some relevant publications [26,27]. In addition, the authors pooled data with different units in their meta-analysis, possibly leading to invalid results. A similar study by Oliveira et al. also had the same limitations [28]. Oliveira et al. suggested that "Ozdemir et al. was in fact selected for systematic review, but excluded from quantitative analyses, as stated in the PRISMA flow diagram"; however, they referenced Ozdemir et al. in included studies in Maia et al. [24]. Due to these mistakes, reaching a firm conclusion seems impossible. By contrast to the limitations of the prior study, ours had some strengths, namely in our systematic review, we included more studies with controversial findings and also including a sufficient number of studies allowed us to do subgroup analysis, all of which contribute to a more reliable conclusion.
As mentioned above, OXS greatly influences brain function and might influence disease prognosis in patients with OCD. Therefore, the measurement of OXS is crucial for these patients. MDA is a known independent marker of OXS and inflammation that has been frequently measured in such patients. Although several earlier studies have been published about serum levels of MDA in patients with OCD, there is still the need to summarize their findings. Measurement of MDA in patients with OCD compared to healthy controls gives us important information about OXS in these patients and the accuracy of MDA measurement for detection of OXS and for expectation of disease prognosis. Therefore, we aimed to conduct current systematic review and metaanalysis to summarize findings from earlier studies in which MDA was measured in OCD patients compared to healthy controls. It should be mentioned that among different markers of OXS, sufficient data were available only for MDA levels.

Search strategy
The present study was performed based on the PRISMA protocol for reporting systematic reviews and metaanalyses [29]. A systematic search was done by two independent investigators (AM and EB) through databases of PubMed, Web of Science, Science Direct, Scopus, and web of science on relevant studies published up to December 2019. Keywords used in our search strategy were: (("Obsessive-compulsive disorder" OR "Obsessive-compulsive disorder*" OR "OCD" OR "Obsessive-compulsive" OR "Obsession" OR "Compulsion" OR "Obsess" OR "Compuls" OR "Obsessive*-compulsive" OR "Obsessive-compulsive*" OR "Obsessive*-compulsive*" OR "Anankastic"[MESH]) AND ("MDA" OR "Malondialdehyde" OR "Malonyldialdehyde" OR "Propanedial")). No restriction was made in terms of publication time or language. Also, the reference lists of relevant papers were reviewed to avoid missing any publication. Data from unpublished studies were not included. If two studies were published on the same set of data, we included only the most complete one [30].

Inclusion and exclusion criteria
In this meta-analysis, publications with the following criteria were considered to be eligible: (1) studies in which OCD cases were diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-III or later edition), or its equivalent in the International Classification of Diseases (ICD); (2) studies that evaluated MDA levels in patients with OCD; (3) studies that recruited healthy subjects as controls; (4) those that reported mean ± standard deviation (SD) MDA level  [27] along with 95% confidence intervals. Letters, comments, short communication, reviews, meta-analyses, ecological studies, and animal studies were also excluded.

Data extraction
Necessary data were extracted from included studies by two independent reviewers. All reported mean ± SD for the MDA level compared with the healthy control group were extracted. We obtained the following information from each study: first author's name, year of publication, country of origin, the age range at study baseline, gender, sample size, study design, number of participants, methods used for assessing MDA levels and OCD, and any adjustments for confounding variables (Table 1).

Quality assessment
The Newcastle Ottawa scale (NOS), designed for observational studies, was used to assess the quality of selected studies [31]. The NOS gives a maximum of ten-point to each study: five for selection, two for comparability, and three for outcome assessment (ten represented the highest quality). Any discrepancies were resolved by discussion. In the current study, those studies with a NOS score of 5 or more were considered high-quality publications ( Table 1).

Statistical analysis
We pooled mean ± SD of MDA levels reported in persons with OCD compared to control groups using a randomeffects model. To assess between-study heterogeneity, the Cochrane Q test and I 2 statistics were used. Betweenstudy heterogeneity was considered significant when I 2 values were 50% or more. Besides, we conducted subgroup analyses by the predefined criteria to find probable sources of heterogeneity. Statistical analyses were done using STATA version 14 (StataCorp).

Results from meta-analysis
The results of overall and stratified analysis are summarized in Table 2. Pooling data from 9 studies, we found that MDA level was significantly higher in patients with OCD comparing to control groups (SMD =  (Fig. 2). Between-study heterogeneity was high among the included studies. We performed subgroup meta-analysis based on participants' country (Turkey and other countries) and study sample size (up to 60 and more than 60) to find possible sources of heterogeneity. Similar findings were also reached in all subgroups (Table 2).

Discussion
The current meta-analysis showed that MDA level was higher in OCD patients compared to control groups. This finding is in line with Maia et al. meta-analysis, where the same result was reached [24]. However, that study had numerous methodological limitations [25]. To solve those limitations, we added the missed studies. In addition, we have added more studies. Also, we have evaluated the MDA concentration alone. Jiménez-Fernández et al. (2015) and Liu et al. (2015) reported that the MDA level was higher in patients with depression disorder than controls [37,38]. Moreover, Flatow et al. (2013) reported higher MDA levels in patients with schizophrenia than controls [39]. Other studies also showed significantly higher lipid peroxidation and nitric oxide in patients with psychological disorders than controls [40,41]. In contrast with our study, Ranjekar et al. (2003) found no significant differences in serum levels of lipid peroxidation markers between patients with schizophrenic and healthy control [42]. Furthermore, Talarowska et al. (2011) did not find any significant association between plasma MDA levels and depressive disorder [43]. Oxidative markers are unspecific and can be influenced by numerous factors, such as body mass index, age, or smoking status [44][45][46]. Our subgroup analysis showed no overall change in our findings. However, due to the limited number of included studies in each subgroup, further studies are needed to reach a firm conclusion.
Although the exact mechanism of the association between OXS and OCD has not been clearly understood, some suggestions are made. ATP production in mitochondria by electron transport chain or through the respiratory burst of macronutrients in macrophages and neutrophils produce OXS [47]. Increased production of reactive oxygen species (ROS) overcomes  antioxidant systems and causes injuries to the lipids, proteins, including some neurotransmitters (NT), and DNA [41]. Several NT containing dopamine, serotonin, gamma-aminobutyric acid, and glutamate have been shown to have regulatory roles in OCD [34]. Changes in these NT pathways may then increase oxidative reactions in the central nervous system [48]. Previous studies have shown that MDA reduces levels of serotonin in the brain [49]. OCD symptoms also result from some damages in the dopamine receptor-rich basal ganglia [50]. Also, OXS could at first be adaptive through enhancement of NT. However, that prolonged ROS generation could lead to exaggerated neurophysiological responses, with disruption of physiological NT [17] and increase of blood-brain barrier permeability, ultimately promoting neuroinflammation and neuronal death [51]. Furthermore, genetic factors might contribute to pathogenesis and symptoms of OCD [52,53]. OCD is prevalent among those with mitochondrial dysfunctions induced by genetic disorders [34]. OXS is shown to be high among these subjects. This meta-analysis had several limitations. Firstly, the number of the included studies was insufficient to do subgroup analyses. Secondly, in the studies that MDA concentrations were reported as µm per mg of protein or hemoglobin, we needed mean protein or hemoglobin concentrations to convert these values to µm. We asked the first authors or study corresponding authors to send us the necessary data, but no response was received. Thirdly, differences in study location and limited number of participants were other limitations. Fourthly, findings were not adjusted for several important confounding factors, such as dietary intakes, body mass index, and smoking status. Fifthly, studies used different laboratory methods to measure MDA serum levels and diagnose OCD, so this should be concerned as a source of bias. Finally, most included studies were done among a certain population (Turkish).
In conclusion, our study showed that MDA levels in patients with OCD were higher than in healthy controls. Studies have shown that antidepressant agents, which are commonly used in OCD patients, reduce oxidative stress markers in the body [54,55]. The current study showed that oxidative stress was high in these patients. Therefore, it might be suggested that measurement of oxidative stress markers in these patients will be a treatment biomarker for the future. Further investigations (among different nations) are required to shed light on the role of OXS in OCD initiation and pathogenesis.