Serum Vitamins and Homocysteine Levels in Obsessive-Compulsive Disorder: A Systematic Review and Meta-Analysis

Studies have shown that vitamins and Hcy might have a role in OCD pathophysiology. The findings in the current meta-analysis elucidate that the levels of vitamins and Hcy differ among OCD patients compared to healthy control groups. The interaction among vitamins and Hcy, genetic factors, and neurotransmitter’s metabolism has significance in OCD symptomatology.

Obsessive-compulsive disorder (OCD) is characterized by unwanted, intrusive, and disturbing images and thoughts (obsessions), which provoke significant distress and repetitive mental or behavioral rituals (compulsions) [1]. These unwanted affairs lead to subsequent temporary reduction in distress [2]. This prevalent disorder affects approximately 2.5% of the population [3]. The lifetime prevalence of OCD has been reported to be 0.8–3.2% in the general population [4, 5]. These patients are faced with many problems especially in their social relationships and occupational function [6-8].

OCD may occur due to genetic factors [9], alterations in frontal striatal connections [10], perinatal injuries, previous psychosocial experiences, and environmental factors such as infections, especially streptococcal infections, toxic pathogens, stress, and traumatic events [11]. The exact etiology of OCD is still unclear, but various pieces of evidence hint to a prominent role of genetic risk factors in OCD [10, 12]. The familiarity in OCD has been reported to be largely due to genetic factors based on twin and familial studies [10]. There are some potential environmental risk factors for OCD such as perinatal complications, stressful or traumatic life events, and reproductive cycle events but without conclusively considered causal role [13].

It has been noted that psychiatric symptoms, mainly affective and psychotic ones, can be related to the levels of vitamin B₁₂, folate, and homocysteine (Hcy) [14]. Folic acid, vitamin B₁₂, and Hcy levels were associated with certain neuropsychiatric disorders. High levels of Hcy and deficiency in vitamin B₁₂ and folate level were observed to play an important role in brain functions and development of nonspecific psychiatric symptoms [15]. There is some evidence that demonstrates the association between vitamin B₁₂ deficiency and elevated Hcy and folic acid with OCD in adult patients [14, 16]. Hcy is a sensitive indicator of folate deficiency [17-19]. Antidepressant impacts of folate supplementation might indicate its importance in psychopathology [20, 21].

Vitamins E, C, and A are essential for functions of the central nervous system (CNS), and decreased level of these vitamins could be associated with structural and functional cellular damages. These vitamins could exert neuroprotective effects alone or even in combination with CoQ10 [22-24]. Their neuroprotective role in neurodegenerative disorders has been attributed to their potent antioxidant activity. Vitamin A is able to inhibit formation of fibrillar alpha-synuclein aggregates that subsequently destabilizes preformed alpha-synuclein aggregates. Vitamin A binds to hydrophobic domains of soluble and fibrillar alpha-synuclein using hydrophobic and antioxidant motifs. Thus, it inhibits aggregation and stabilization of alpha-synuclein [23]. Vitamin A can also be used in control and prevention of synucleinopathies [23, 24]. Besides having antioxidant roles during cell injury [25, 26], vitamin C has specific roles in activation of 2 enzyme classes, the iron-containing hydroxylases and the copper-containing hydroxylases (e.g., dopamine-hydroxylase) [27]. Vitamin E, localized in the cell membrane, is targeted for its relation to vascular function and atherosclerosis. Decreased level of antioxidants such as vitamin E accompanying stimuli such as bacterial colonization, exposure to various toxins, infection, and/or metabolic alterations such as enhanced Hcy concentration would be associated with increased free radical level and altered brain vascular function [27].

Vitamin D is another supplement with suggested involvement in the etiology of psychiatric disorders. Its deficiency has been observed in pathogenesis of schizophrenia, autism, and OCD [28-33]. It has been suggested that vitamin D deficiency could specifically impair inhibition of perseverative responses which could explain its relationship with stereotypical/repetitive behaviors in psychological disorders such as autism and OCD [34]. Vitamin D regulates both tryptophan hydroxylase 2 and tyrosine hydroxylase, the rate-limiting enzymes essential for synthesis of dopamine, serotonin, adrenaline, and noradrenaline. Thus, vitamin D exerts an essential role in neuroprotection, neurotransmission, and synaptic plasticity. Vitamin D deficiency may result in a wide range of behavioral and emotional problems [15, 35, 36].

Despite the number of studies investigating the association between vitamins B, D, E, C, and A as well as Hcy levels with OCD, there are still some contradictory results. To the best of our knowledge, there is paucity of meta-analysis or systematic review studies investigating the association between serum vitamins and Hcy levels in OCD. Thus, we aimed to evaluate the association between serum vitamins and Hcy levels with OCD patients through a meta-analysis approach.

This study was done according to the PRISMA protocol to report systematic reviews and meta-analyses [37]. A systematic search was performed by 2 independent researchers (A.M. and E.B.) from the online database of ISI Web of Science, PubMed, Scopus, Google Scholar, and Science Direct to find relevant publications until September 2019. The 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 Term])) AND (“Vitamin” OR “Vitamin A” OR “Retinol” OR “Retinoic acid” OR “Vitamin D” OR “Colecalciferol” OR “Ergocalciferol” OR “Calcitriol” OR “Vitamin B₆” OR “Pyridoxamine” OR “Pyridoxal” OR “Pyridoxine” OR “Folate” OR “Vitamin B₉” OR “Folic acid” OR “Vitamin M” OR “Pteroylglutamic Acid” OR “Folic Acid, (DL)-Isomer” OR “Cyanocobalamin” OR “Cobalamin” OR “Vitamin B₁₂” OR “B-vitamins” OR “Ascorbic Acid” OR “L-Ascorbic Acid” OR “Vitamin C” OR “Vitamin E” OR “Tocopherols” OR “Hcy”)). No restriction was considered for the time and language of publications. In addition, the reference list of the relevant studies was reviewed to avoid missing any publication, and the duplicate studies were also removed afterward.

In this part, studies with the following criteria were eligible for inclusion: (1) cases defined according to OCD diagnosis, as defined by the Diagnostic and Statistical Manual of Mental Disorders (DSM-III or later edition), or its equivalent in the International Classification of Diseases (ICD); (2) all studies evaluating serum vitamins and Hcy levels in patients with OCD; (3) studies that reported serum vitamins and Hcy levels at baseline through clinical trial on OCD patients versus healthy controls; and (4) those that reported mean ± standard deviation (SD) for serum vitamins and Hcy levels. If the same dataset had been reported in more than one publication, only the study with more complete findings was included in our systematic review and analysis. In this meta-analysis, letters, comments, short communications, case reports, clinical trials without the healthy control group, reviews, meta-analyses, and animal studies were excluded from the analysis.

Data extraction and study selection were independently performed by 2 researchers, and then any disagreements were resolved through discussion. Any reported mean ± SD for serum vitamins and Hcy levels in patients compared with the control group was extracted. We also extracted the following information from each study: first author, year of publication, country of origin, age range at study baseline, gender, sample size, design study, number of participants, methods used for assessing serum vitamins and Hcy levels, and OCD and statistical match for confounding variables (Table 1).

A form of the Newcastle-Ottawa Scale (NOS) was designed for observational studies and was used to assess quality of selected studies. The NOS considers a maximum of 10 points to each study: 5 for selection, 2 for comparability, and 3 for assessment of outcomes (10 represented the highest quality). In the current study, those that had an NOS score of 5 or more were considered as high-quality publications [38]. Indeed, we applied the JADAD checklist for evaluation of the quality of interventional studies. This checklist consists of 3 major items including randomization, blinding, and description of dropouts. High-quality studies were those clinical trials with scores of 3 or more [39]. Two independent reviewers filled out the scores for each eligible study, and any discrepancies were also resolved by discussion.

In the current meta-analysis, serum vitamins and Hcy levels were reported using mean and SD and 95% confidence intervals (CI) for both OCD and control groups. The overall mean ± SD was calculated using a random-effects model and/or fixed-effects model. The standard mean difference (SMD) and CI were considered as the overall. If there was a true heterogeneity between the included studies, we employed a random-effects model, otherwise a fixed-effects model was used. To assess between-study heterogeneity, Cochran’s Q test and I² were used. A fixed-effects model was used if p_{heterogeneity} (p_h) > 0.1, while a random-effects model was used if p_h < 0.1 [40]. In addition, we conducted subgroup analysis according to the age and country. Furthermore, we performed a sensitivity analysis to find the effects of a single study on the overall estimated mean ± SD by excluding that study. Potential publication bias was assessed by visual inspection of Begg’s funnel plots and also using Egger tests. Statistical analyses were done using review manager software (Revman 5.3; Cochrane Collaboration, Oxford, England).

We identified a total of 1,288 titles, in which 372 publications were duplicates and another 848 publications were excluded according to the title and abstract. Terminally, 68 full-text articles were screened for inclusion in our meta-analysis, in which 58 publications were excluded due to nonhuman samples (animal studies), absence of measurement of vitamins and Hcy in OCD, and lack of evaluation of OCD outcomes as shown in Figure 1. The last remaining 10 articles [14-16, 41-47] were included in this study, and the characteristics of these studies are shown in Table 1. Eligible studies provided the data regarding the serum level of 7 different vitamins and Hcy in enrolled subjects. The sample size of these studies varied from 46 to 119 participants (377 patients [48.9% women vs. 51.1%] and 321 controls [47.4% women vs. 52.6% men]). One study did not report the number of participants by gender [42]. In the study, mean age of patients was estimated to be 22.49 vs. 21.92 years in the control group. It should be noted that one study has not reported the mean age of participants [42]. The included studies had been published between 1988 and 2017. Six studies were conducted in Turkey [14-16, 44, 46, 47], one in India [48], one in Bangladesh [41], one in USA [43], and one in Israel [42]. In 3 studies [14, 41, 47], the mean duration of disease ranged from 7.3 to >120 months, and residual studies did not evaluate this parameter [15, 16, 42-44, 46, 48]. In addition, the score of Y-BOCS ranged from 19.9 to 28.2 which was reported in 4 studies [15, 16, 47, 48]. The diagnosis manual in 7 studies was established according to DSM-IV [14, 16, 41, 44, 46-48]. However, 2 studies conducted their diagnosis based on DSM-5 [15] and DSM-3 [43] protocols. One study did not report the applied protocol [42]. Three included studies investigated vitamins D [15, 46, 47], and/or C [41, 44, 48], and/or B₁₂ [14, 15, 42], and/or Hcy [14-16], one study evaluated vitamin A [41] and/or B₆ [43], 2 studies were for vitamin E [41, 44], and 4 studies investigated vitamin B₉ [14-16, 42] (Table 1). The number of participants included in each subgroup was as follows: vitamin B₁₂ (117 cases/82 controls), vitamin B₉ (140 cases/105 controls), vitamin A (48 cases/48 controls), vitamin D (145 cases/109 controls), vitamin E (78 cases/78 controls), Hcy (110 cases/75 controls), vitamin B₆ (27 cases/26 controls), and vitamin C (117 cases/111 controls). The mean ± SD of serum levels of the vitamins and Hcy per study is depicted in Tables 2and3.

The results of overall and stratified analyses are summarized in Table 4 and Figure 2. Regarding vitamin B₉, vitamin B₁₂, and Hcy, there was a statistically significant association between low level of B₁₂ and OCD (SMD = −0.58, 95% CI = [−1.08, −0.08], p = 0.02). Besides, we observed a significant association between higher level of Hcy and OCD (SMD = 1.11, 95% CI = [0.48, 1.75], p = 0.0006). However, there was no significant difference between case and control groups regarding the level of vitamin B₉ (SMD = −0.23, 95% CI = [−1.01, 0.55], p = 0.56).

True heterogeneities were observed across studies for the 3 abovementioned parameters (B₉: I² = 88%, p_h < 0.0001; B₁₂: I² = 65%, p_h = 0.06; and Hcy: I² = 73%, p_h = 0.02). Egger’s test did not show any evidence of publication bias. p_Egger’s for B₉, B₁₂, and Hcy was 0.875, 0.78, and 0.513, respectively. The subgroup meta-analysis based on the age category showed that the serum level of B₁₂ and Hcy was not associated with OCD after excluding the study with teenager subjects (B₁₂: SMD = −0.34, 95% CI = [−0.71, 0.03], p = 0.07, I² = 0%, p_h = 0. 32; Hcy: SMD = 1.19, 95% CI = [−0.02, 2.40], p = 0.05, I² = 87%, p_h = 0.006). Subgroup meta-analysis based on country revealed no changes compared to overall findings (Table 4). The sensitivity analysis showed that our results were reliable in all analyses except when we excluded the study by Esnafoğlu and Yaman [15] for Hcy and B₁₂ and also by Türksoy et al. [14] for B₁₂.

Vitamin D levels were also slightly lower but not statistically significant in OCD patients compared with the control group; however, this difference was not statistically significant (SMD = −0.63, 95% CI = [−1.41, 0.15], p = 0.11, I² = 88%, p_h = 0.0002). Begg funnel plot and Egger test (p_Egger’s = 0.703) provided no evidence for the presence of publication bias (Fig. 3).

A significant association was found between both lower vitamin E (SMD = −0.89, 95% CI = [−1.23, −0.56], p < 0.00001) and vitamin C (SMD = −1.40, 95% CI = [−2.44, −0.36], p = 0.008) level and OCD. While evidence of heterogeneity was not observed with vitamin E (I² = 23%; p_h = 0.26), vitamin C showed evidence of heterogeneity (I² = 92%; p_h < 0.0001). There was no possible evidence of publication bias for vitamin E and vitamin C (p_Egger’s = 0.97). The sensitivity analysis also showed that our results were reliable in all analyses except when we have excluded the Shohag et al. [41] and Chakraborty et al. [45] studies for vitamin C.

Vitamins are essential for various brain processes (Fig. 4) and neuroplasticity, which implies to their possible role in psychiatric disorders [49]. However, this concept is poorly characterized in OCD. In our meta-analysis, we found a statistically significant higher Hcy level and lower concentration of C, E, and B₁₂ vitamins. Indeed, there was not a statistically significant association between deficiency in vitamins D and B₉ and OCD.

In detail, meta-analyses revealed that Hcy was significantly higher in OCD cases. Also, our results are supportive for a lower vitamin B₁₂ level in patients with OCD. However, we found a nonsignificantly lower level of folate. In line with our findings, Salagre et al. [50] reported that serum and plasma levels of Hcy were higher in subjects with bipolar disorder when compared to healthy controls. The similar investigations were also performed in other neurological disorders. Petridou et al. [51] demonstrated a significantly lower level of B₁₂ in depressed cases. Cao et al. [52] in their meta-analysis found an insignificantly higher vitamin B₁₂ level in schizophrenia cases. Wang et al. [53] found higher Hcy concentration in cases with Alzheimer’s disease compared to normal subjects. In other meta-analyses, Xie et al.’s [54] results suggested that patients affected with Parkinson disease and cognitive dysfunction were more likely to have higher Hcy, lower vitamin B₁₂ levels, and higher folate level. We could not find studies with opposite data regarding Hcy. The reason of such differences between the results of studies may be basically related to the different impacts of the diseases on the diet, and vice versa. Dietary problems along with trace storage of B₁₂ in the brain can justify the increase in Hcy level [49]. Although many literatures indicate an association between elevated Hcy concentration and psychiatric disorders, the nature of this association remained unknown. Since Hcy contributes to many physiological processes, its role in the CNS is very complicated [50]. It has been suggested that changes in glutamatergic neurotransmission might be a chief mechanism linking Hcy with psychiatric disorders [55, 56]. Also, Hcy could play the main role as an agonist for N-methyl-D-aspartate receptors. This may lead to an elevated influx of calcium ions, accumulation of reactive oxygen species, and increased intracellular second messenger calcium with subsequent induction of cell injury and stimulation of both necrotic and apoptotic cell death pathways [55, 57, 58]. Higher levels of produced Hcy due to methylenetetrahydrofolate reductase deficiency alter GABAergic and glutamatergic levels that contribute to CNS neurotoxicity [59]. The capacity of Hcy metabolism is dependent on sufficient supplies of B₉ and B₁₂ [60]. Deficiency of vitamin B₁₂ results in decreased rate of Hcy methylation to methionine which could result in a functional deficiency of 5,10-methylenetetrahydrofolate. Reduced remethylation of Hcy to methionine and S-adenosylmethionine due to absence of folate or vitamin B₁₂ could lead to increased concentration of Hcy. Reduced synthesis of S-adenosylmethionine could result in hypomethylation state, for instance leading to impaired synthesis of proteins and neurotransmitters necessary for the brain structural integrity [61, 62]. Folate metabolism contributes to regulation of monoamine, and then changes in Hcy metabolism may result in insufficient production of monoamine and consequent dysregulation of dopamine, norepinephrine, and serotonin [63, 64]. Vitamin B₁₂ role in one-carbon metabolism helps in the methylation procedures of proteins, neurotransmitters, and phospholipids of the neural membrane and is essential for DNA synthesis [15]. These metabolic pathways could play a main role in the incidence of neuropsychiatric symptoms [65] and development of OCD [62]. Hyperhomocysteinemia might participate in cholinergic metabolism via decreasing activity of choline acetyltransferase in some neurons which may possibly result in cognitive impacts [66]. Cognitive damage is an important factor involved in psychosocial functioning [67]. There are numerous conflicting data regarding age-related Hcy changes. And, Hcy results were insignificant in our age category subgroup. Since dietary habits might influence folate and vitamin B₁₂ levels, the results of surveys could vary in data derived from different societies [62]. Further studies are needed to investigate the association between Hcy and OCD.

The present meta-analysis shows that OCD patients had an insignificantly lower level of vitamin D compared to the healthy controls. The similar investigations were performed in other neurological disorders. Anglin et al. [68], Kotsi et al. [69], and Wang et al. [32] showed lower levels of vitamin D in patients with depression, attention deficit hyperactive disorder, and autism spectrum disorder, respectively. Although our meta-analysis supports the lack of meaningful results, the results are of low validity because the data are related to a special race and age range. Also, the control groups of included studies in our meta-analysis had vitamin D deficiency. Due to the abovementioned limitations, further studies are required in different ethnicities and ages to obtain valid results regarding vitamin D and OCD.

There are several probable associations between vitamin D and OCD such as the relationship between tyrosine/tryptophan hydroxylase and active form of vitamin D3 (1,25-dihydroxyvitamin D3). Tyrosine hydroxylase is the rate-limiting enzyme in synthesis of epinephrine, norepinephrine, and dopamine. The rate-limiting enzyme in synthesis of serotonin is tryptophan hydroxylase. The level of these 2 enzymes is regulated by 1,25-dihydroxyvitamin D3 [65, 66]. Thus, vitamin D deficiency might contribute to OCD pathogenesis through its impacts on the pathway of catecholamines and serotonin synthesis. Another link between vitamin D and OCD goes back to the neuroprotective effects of vitamin D. The role of free radicals, mainly enhanced nitric oxide level, has been demonstrated in OCD [67]. Vitamin D inhibits inducible nitric oxide synthase, an essential enzyme for NO production, through its antioxidant properties [68]. Therefore, vitamin D deficiency could be involved in OCD pathogenesis through impairment of neuroprotection [62].

Our meta-analysis shows that OCD patients had lower level of vitamins C and E compared to the healthy control. Flatow et al. [70] showed higher plasma levels of vitamins E and C in chronic schizophrenia patients. A meta-analysis by Dong et al. [71] showed lower vitamin E level in the plasma of Alzheimer’s disease patients compared to the healthy control group; however, no differences were detected in the vitamin C level of their plasma which is varying to results in OCD. The simultaneous reduction of vitamins C and E in OCD patients is justified by vitamin E as an antioxidant which protects the body against free radicals and converts them into vitamin E radicals. In turn, the resulting vitamin E radicals can be recycled by the action of vitamin C [72]. In addition to the role of vitamin E in the primary antioxidant process, it can also affect the system of endogenous antioxidant defense. This system consists of a series of enzymes involved in antioxidation, detoxification, and transportations that are controlled by the nuclear factor erythroid 2-related factor 2 (Nrf2) transcription factor. Cellular levels of Nrf2 are regulated by Kelch-like ECH-associated protein 1, its repressor protein, by controlling the level of Nrf2 degradation. Expression of Nrf2 in a steady-state status is normally very low. On the other hand, in stress conditions, the half-life of Nrf2 as well as the expression of genes with the promoter region containing elements responsive to Nrf2 increases. These mentioned items are known as antioxidant response elements [73]. The effect of vitamin E on the activity of Nrf2 greatly depends on the administered type of the vitamin, but there is much evidence that vitamin E influences the expression of antioxidant enzymes. It should be noted that vitamin E might at least partially influence the expression of antioxidant enzymes via more general mechanisms such as mRNA and protein degradation induced by oxidative stress. In addition, this vitamin has been revealed to exert influence on the mRNA stability, and also it can increase the activity of glutathione peroxidase-1, but not the superoxide dismutase-1 or catalase activities. This process may be due to elevated mRNA stabilization rather than transcriptional stimulation [74]. Thus, the rate of tissue injury in neuropsychiatry may be lower by optimizing vitamin E status through consumption of recommended amounts of food intake or supplementation.

Although B₆ is essential for the Hcy metabolic pathway, as is possibly involved in the pathogenesis of OCD, only one study was performed regarding the association between B₆ and OCD with insignificant results. Also, there was no powerful study regarding the role of vitamin A on development of OCD. Therefore, further study is needed to investigate the association between OCD and vitamins A and B₆.

This meta-analysis included several limitations. The main limitation is the insufficient number of eligible studies to perform subgroup analysis. Second, serum vitamins and Hcy levels are nonspecific, and they can be affected by various factors, including diet, age, body mass index, and gender [75]. Although some of these factors such as age and gender were considered in included studies, not all the included studies have considered all of these factors. For better understanding of the association between serum vitamins and Hcy levels and OCD, it is advisable to consider all of the abovementioned factors. Additionally, the included studies, derived from different sample sources and using different methodologies, would naturally increase the data heterogeneity. Lastly, the low number of eligible articles, the moderate quality (mean of 5 in the NOS), and the high heterogeneity of effect sizes among the included studies in each independent meta-analysis are factors which affect reliability and are better to be avoided. Some other confounding factors such as disease phase (acute or remission phase), the age of disease onset, and the presence of concomitant psychopharmacological treatments of OCD patients were not also considered in the meta-analysis.

In conclusion, we found evidence supporting elevated Hcy with lowered B₁₂, E, and C vitamins in OCD cases, whereas there was an insignificantly lower level of D and B₉ vitamins in this regard. These findings indicated that vitamin supplementation or treatment might have some other beneficial impacts on OCD cases. Further research is needed to confirm this association.

We thank everyone who consulted in the article.

Since no human or animal subjects were involved in the current study, an ethical approval was not required.

The authors had no conflicts of interest.