Silent Infections are not So Silent: The Emerging Role of Combined Infections, Inflammation, and Vitamin Levels in OCD

Abstract

Objective

Recent evidence highlights that different agents may trigger immune-mediated processes involved in the pathophysiology of different neuropsychiatric conditions. Given the limited information on obsessive-compulsive disorder (OCD), the present study aimed at assessing current/past infections and plasma levels of vitamin D, vitamin B12, folic acid, homocysteine and common peripheral inflammatory markers in a group of OCD outpatients.

Method

The sample included 217 adult outpatients with an OCD diagnosis according to the DSM-5 criteria. The Yale-Brown Obsessive-Compulsive Scale (Y-BOCS) was used to assess the clinical phenotype and symptom severity. Laboratory blood tests measured levels of vitamin D, vitamin B12, folic acid, homocysteine, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), blood count and antibodies titers for cytomegalovirus (CMV), Epstein Barr virus (EBV), Toxoplasma gondii and antistreptolysin titer.

Results

Sixty-one patients had a previous EBV infection, 46 were seropositive for CMV IgG, 24 showed positive antistreptolysin titer, 14 were seropositive for Toxoplasma gondii IgG, and four for CMV IgM. More than a half of patients showed vitamin D insufficiency. Compared to seronegative patients, patients with a past EBV infection displayed significantly higher scores on the Y-BOCS total score and compulsion subscale, and other symptoms. Vitamin D was negatively correlated with both the Y-BOCS total score and the subscales scores. Folic acid was negatively correlated with the Y-BOCS total and obsessions subscale score.

Conclusions

The findings of our study show an association between Epstein-Barr infection and hypovitaminosis D and the overall severity and specific symptom patterns of OCD. The laboratory measures used in this study are useful, cheap and easy parameters that should be routinely assessed in patients with OCD. Further studies are needed to clarify their role in OCD pathophysiology and outcomes, as well as the potential therapeutic impact of vitamins and antibiotics/immunomodulatory agents in OCD and other psychiatric conditions.

Introduction

Recently, much interest and research efforts are directed towards deepening the dynamic crosstalk between the immune system and the brain, as well as its role in the onset of neuropsychiatric disorders (Hornig, 2013; Najjar et al.; Pape et al., 2019). Further, the evident neurotropism of SARS-COV-2 virus that provoked the recent catastrophic COVID-19 pandemic, promoted a renewed focus on infective agents that might directly target neurons, microglia and astrocytes (Ludlow et al., 2016; Veleri, 2022; Moreira de Oliveira et al., 2023). As already reported during previous pandemics, the virus might impair not only the host's immune system, while eliciting a strong inflammation state, but it might also increase the inflammation through psychosocial stressors, and, as such, induce or worsen psychiatric disorders in children, adolescents and adults. Again, infections during pregnancy might lead to an increased risk of neurodevelopmental disorders for the fetus (schizophrenia and autism spectrum disorders), as already described in the aftermath of the Spanish flu (Loganovsky & Loganovskaja, 2021; Shuid et al., 2021). Inflammatory mediators derived from the activation of both innate (microglial activation with subsequent phagocytosis or cytokine production) and adaptive immunity (antigen-specific responsiveness and production of antibodies) (Kerr et al., 2005), could be responsible of a state of chronic neuroinflammation that may finally affect many biological pathways including synaptic plasticity, neuroendocrine activity, and neurotransmission (Hu et al., 1997; Hornig, 2013; Mousa & Bakhiet, 2013; Borsini et al., 2015; Levin & Godukhin, 2017; Abdoli et al., 2020). While considering the several microbes and infectious agents, as well as commensal microorganisms, which have been linked to neuropsychiatric syndromes and their outcomes, the activation of shared immune-signaling responses might contribute to non-specific disruption of immune-brain homeostasis, exposing the central nervous system (CNS) of susceptible individuals to detrimental consequences in circuits responsible for cognitive and behavioural manifestations (Pape et al., 2019). Dysfunctions in immune response, with eventual related disruption of the hypothalamic-pituitary-adrenal (HPA) axis particularly relevant during viral infections (Silberman et al, 2005), have been implicated in the pathophysiology of several neurological and psychiatric disorders (Kerr et al., 2005), including obsessive-compulsive disorder (OCD), a common and devastating psychiatric condition characterized by typical symptoms, namely obsessions and /or compulsions (Marazziti et al., 2018). The autoimmune hypothesis for the pathophysiology of OCD emerged from the early ‘90s, starting from the evidence that Sydenham’s chorea (SC), a manifestation of rheumatic fever due to A-haemolytic streptococcal infections, is often associated with obsessive–compulsive (OCD) symptoms and tics (Swedo et al., 1989; 1998). Several forms characterized by prepubertal acute onset, episodic course and concurrent neurological abnormalities, such as choreiform movements, occurring or exacerbating after exposure to infection, were initially identified as paediatric autoimmune neuropsychiatric disorder associated with group A beta-haemolytic streptococcus or with the acronym PANDAS (Swedo et al., 1998; Murphy & Pichichero, 2002). These conditions were subsequently renamed as paediatric acute neuropsychiatric syndrome or PANS, after the evidence that several agents other than streptococcus might be involved in the illness development (Swedo, 2012; Murphy et al., 2015). However, this pathophysiological model might be relevant even for adult OCD. Studies focused on inflammatory biomarkers in OCD patients revealed inflammation in the cortico-striato-thalamo-cortical (CSTC) circuit and the presence of greater plasma levels of interleukins (IL) IL-2, IL-4, IL-6, IL-10 and tumor necrosis factor-α (TNF-α) in adult OCD patients, compared to controls (Brambilla et al, 1997; Marazziti et al., 1999; 2018, Fontenelle et al., 2012), as well as higher rates of anti-basal ganglia antibodies (Khanna et al., 1997a; 1997b). Through the years other pathogens, including bacteria, viruses, and parasites, have been recognized to be possibly related to OCD in both children and adults (for review, see: Della Vecchia & Marazziti, 2022). A survey conducted on self-reports of individuals with a previous Lyme disease, an affection caused by the spirochete Borrelia burgdorferi, showed that a substantial proportion of people developed OCD symptoms, mainly gradually after the infection, and were responsive to psychological and pharmacological treatments, including antibiotics (Johnco et al., 2018). Toxoplasma gondii (T. gondii) is a common and widespread protozoan parasite that can infect any warm-blood animal; it requires an intermediate host, and a definitive one, represented by felines, in order to complete its life cycle. After the acute infection, a latent, chronic counterpart persists with the encystment of the parasite in muscle and brain cells determining host’s seropositivity throughout lifetime (Prandovszky et al., 2011). Seroprevalence for T. Gondii was found in schizophrenia (Prandovszky et al., 2011), depression (Henriquez et al., 2009), and even OCD (Miman et al., 2010; Flegr, 2015; Flegr & Horáček, 2017). Interestingly, a recent meta-analysis concluded that the seropositivity for T. Gondii is more frequent in OCD patients, as compared with control subjects (Nayeri Chegeni et al., 2019). In some cases, anti-protozoan treatment for toxoplasmosis demonstrated efficacy in reducing both infectious and OCD symptoms (Brynska et al., 2001). The explanation of behavioral changes due to T. Gondii infection could derive from the alterations of dopamine circuits through the enhancement of its metabolism in neuronal cells, but also of other neurotrasmitters, such as the gamma-aminobutyric acid (GABA), glutamate, serotonin and noradrenaline (Fuks et al., 2012; Xiao et al., 2013; Brooks et al., 2015; Fabiani et al., 2015; Kannan et al., 2016; McFarland et al., 2018), as well as HPA functioning (Henriquez et al., 2009).

Finally, herpes viruses, such as the Epstein-Barr Virus (EBV) which induces mononucleosis, and the cytomegalovirus (CMV), after an initial lytic phase, remain located in specific cells where they trigger a long-life latent infection which might be reactivated during occasional conditions. Encephalitis caused by acute infections are characterized by neuropsychiatric symptoms and alterations in the same brain structures involved in OCD, such as basal ganglia, frontal and temporal lobes (Damasio & Van Hoesen, 1985; Ludlow et al., 2016). There is some evidence of higher IgG against HSV-1 in serum and cerebrospinal fluid of OCD patients when compared to controls, but these 1997 studies have never been replicated since then (Khanna et al., 1997a; 1997b). Moreover, scattered reports described cases of tics and OCD that have been related to lesions induced by Varicella zoster virus in basal ganglia (Dale et al., 2003; Yaramiş et al., 2009). The mechanisms underlying the development of OCD symptoms are supposed to encompass several processes, such as stimulation of autoimmunity through molecular mimicry and impairment of monoamine neurotransmission, other than direct immune-mediated neuronal damage at the implantation site for neurotropic agents. However, all proposed mechanisms seem to converge to the induction of functional/structural alterations of CSTC circuits, which are known involved in OCD pathophysiology.

Given the limited information in OCD, the present study aimed at assessing some current/past infections, plasma levels of vitamin B12, D, folic acid, homocysteine, for their role in different brain and immune processes related to psychopathology and OCD, and common peripheral inflammatory markers, and their possible relationships, in a group of patients suffering from OCD, in order to better characterize its phenotypic presentations and possibly identify potentially affected symptom clusters or dimensions.

Materials and Methods

Sample recruitment

A total of 217 OCD adult outpatients of both sexes attending the Psychiatric Unit at Pisa University, recruited between October 2018 and September 2022, were included in the present study. All patients were diagnosed as suffering from OCD, according to the Diagnostic and Statistical Manual for mental disorders, fifth edition (DSM-5) (APA, 2013). All subjects were first assessed by a clinical evaluation with the ensuing diagnoses, subsequently to be supported by the structured clinical interview for DSM-5, research version, patient’s edition (First et al., 2015). The Yale-Brown Obsessive Compulsive Scale (Y-BOCS) was used to assess the severity of OCD (Goodman et al., 1989). Patients who were pregnant, with drug intoxication, with severe comorbid psychotic disorders or major medical illness or suffering from cognitive impairment or dementia, were excluded. All information regarding socio-demographic variables, psychiatric and medical comorbidities, and pharmacological treatment were derived from the medical history collected during the interview. After a complete description of the study, a written informed consent was obtained from each subject to participate in the study.

Psychopathological assessment Structured Clinical Interview for DSM-5, research version

The Structured Clinical Interview for DSM-5 (SCID-5) is a semi-structured interview guide for making the major DSM-5 diagnoses. The most comprehensive version of the SCID-5, the Research Version (SCID-5-RV) (First et al., 2015) contains more disorders than the Clinician Version or the Clinical Trials Version and includes all the relevant subtypes and severity and course specifiers. An important feature of the SCID-5-RV is its customizability, allowing the instrument to be tailored to meet the requirements of a particular study.

The Yale-Brown Obsessive-Compulsive Scale

The Yale-Brown Obsessive-Compulsive Scale (Y-BOCS) is a clinician-administered instrument to assess the presence and severity of obsessive-compulsive symptoms. It is divided into a symptom checklist and a severity scale. The former includes 54 dichotomous items assessing current or prior presence of specific obsessions (aggressive, contamination, sexual, hoarding/saving, symmetry, somatic, others) and compulsions (cleaning, control, repetitive rituals, count, reorganise, hoarding, others). The severity scale consists of 10 items that quantify the impact of obsessions and compulsions identified using the symptom checklist rated on 5-point Likert-type scales along the dimensions of time, interference, distress, resistance, and control, with a higher score corresponding to a more severe symptomatology for all items (0: no symptoms - 4: extreme symptoms). The total Y-BOCS score is defined by the sum of items 1 to 10 (range 0 to 40), with separate subtotals for the severity of obsessions (items 1 to 5) and compulsions (items 6 to 10). (Goodman et al., 1989).

Biological assessment

The antibody titers (IgM and IgG against CMV, EBV and Toxoplasma gondii) were measured by common clinical chemiluminescence methods, while antistreptolysin O titer (ASO) was evaluated with immune-turbimetry method. To assess vitamin B12, folate plasma, homocysteine, vitamin D levels, inflammatory markers and antibody titers, common clinical chemical methods were used. Standard serum B12 assays quantify both the inactive and active forms of serum cobalamin and are based on intrinsic factor binding and immune-chemiluminescence based techniques. Standards of reference of B12 plasma normal levels as well as its deficiency are not established. Although, it has been proposed that serum B12 of <148 pmol/L (200 ng/L) would have the sensitivity for diagnosing the 97% of true cobalamin deficiencies. Laboratories use different reference ranges and units of measurement (pmol/L or ng/L), in the absence of a standardized methodology (Devalia et al., 2014). Competitive folate binding protein assays using chemiluminescence or fluorescence detection systems are the techniques most frequently used by laboratories to measure folate plasma levels. Although there is no consensus on folate plasma levels indicative of folate deficiency, usually levels <7 nmol/l (3 μg/L) are taken as reference since they are associated with a strong increase in risk for megaloblastic anaemia. However, the significance of low levels, between 7 and 10 nmol/L (3 and 4-5 μg/L), is still undefined (Devalia et al., 2014). To conduct this analysis, we took as reference the levels suggested by the World Health Organization (WHO) as indicative of folate and B12 deficiencies: < 10 nmol/L (4 ng/mL) for serum folate and < 150 pmol/L (203 pg/mL) for plasma vitamin B12 (de Benoist, 2008). 25(OH)D concentrations were used as a marker of vitamin D status. It was considered optimal for levels >30 ng/mL, sufficient between 20 and 30 ng/mL, insufficient for 12-20 ng/mL, deficient for 6-12 ng/mL, and severely deficient for <6 ng/mL. As regards other biomarkers, we referred to common standard values of reference.

Vitamin D was evaluated in 142 patients, while folic acid and vitamin B12 in, respectively, 97 and 98 patients. EBV was evaluated in 99, T. gondii IgG and IgM in 98, CMV IgG and IgM in 96, and ASO in 93 patients. Homocysteine was available for 44 patients, blood counts for 59, Erythrocyte Sedimentation Rate (ESR) for 51, and finally, C-reactive protein (CRP) for 54 patients. (The numbers of patients undergoing the biomarkers is different because some individuals refused to do the blood test or to do all the proposed battery).

Statistical analyses

All demographic, clinical and laboratory data were presented for continuous variables in terms of mean ± standard deviation (SD), variation range (min and max values), or medians, when required. Categorical variables were expressed as frequencies (number) and percentages.

The Kolmogorov-Smirnov test was used to determine normality of distribution of the variables. Comparisons for continuous variables were performed with the independent-sample Student’s t-test. Comparisons for categorical variables were conducted by the use of the χ2 test. The Mann-Whitney test was used to explore the possible differences in Y-BOCS total-score and obsession/compulsion subscales between the subgroups. Furthermore, when required, the median test was performed. It is a non-parametric test that is used to test whether two (or more) independent groups differ in central tendency - specifically whether the groups have been drawn from a population with the same median. The null hypothesis is that the groups are drawn from populations with the same median.

The correlations between different parameters, and between characteristics of the subjects and biological markers were investigated by calculating the Pearson’s correlation coefficient. Cohen’s standard may be used to evaluate the correlation coefficient to determine the strength of the relationship, or the effect size. Correlation coefficients between .10 and .29 represent a small association, coefficients between .30 and .49 represent a medium association, and coefficients of .50 and above represent a large association or relationship. All p values lower than .05 were considered statistically significant.

Statistical analysis was performed using SPSS 25.0 software (IBM Corp., 2017, Armonk, NY).

Results

Socio-demographic data

The total sample of 217 OCD patients included 126 (58.1%) men and 91 (41.9%) women. The age of the sample (years, mean ± SD) was 34.06 ± 12.45, with men showing a statistically significant younger age than women (32.13 ± 11.62 vs 36.75 ± 13.12, t = 2.738, p = 0.007). The socio-demographic and clinical characteristics of the study group are summarized in table 1.

Table 1.

Clinical characteristics and treatments of outpatients with OCD

Age of onset(years)		19.08 ± 7.24	18.12 ± 6.13	20.42 ± 8.40
Course	Episodic	38 (17.5%)	15 (6.9%)	23 (10.6%)
	Chronic	179 (82.5%)	111 (51.2%)	68 (31.3%)
Onset	Acute	77 (35.8%)	45 (20.9%)	32 (14.9%)
	Progressive	138 (64.2%)	80 (37.2%)	58 (27%)
Onset related to stressful life events		83 (39%)	41 (19.3%)	42 (19.7%)
Onset related to the end of a lerelationship		53 (24.6%)	28 (13%)	25 (11.6%)
Perinatal traumas		56 (26%)	43 (20%)	13 (6%)
Obsessive-compulsive personality traits		129 (60%)	78 (36.3%)	51 (23.7%)
DRUGS:SSRIs			133 (61.6%)
Clomipramine			57 (26.4%)
Valproic Acid			42 (19.4%)
Gabapentin/Pregabalin			31 (14.4%)
SGAs			29 (13.4%)
Lithium			20 (9.3%)
Other mood stabilizers			18 (8.3%)
Benzodiazepines			16 (7.4%)
FGAs			12 (5.6%)
Other antidepressants			10 (4.6%)
Other tricyclic antidepressants			9 (4.2%)
SNRIs			4 (1.9%)
CBT			33 (15.3%)

Note: Values are reported as mean ± standard deviation.

SSRIs: selective serotonin reuptake inhibitors; SGAs: second-generation antipsychotics; FGAs: first generation antipsychotics;

SNRIs: serotonin-noradrenaline reuptake inhibitors; CBT: cognitive-behavioral therapy.

One hundred and sixty-six (76.5%) patients were unmarried and about a quarter (n = 51, 23.5%) were married/cohabitant. One hundred and twenty-nine (59.4%) patients had a job, while 88 (40.6%) were unemployed. Half of patients (n = 115, 53.0%) had high school diploma, followed by patients (n = 80, 36.9%) who achieved university graduation. Twenty patients (9.2%) had a primary school diploma and only two (0.9%) patients had a PhD graduation.

The majority (n = 192, 88.5%) of patients did not suffer from any significant medical comorbidity. The most frequent (n = 8, 3.7%) medical condition reported was autoimmune thyroiditis.

Psychopathological assessment

The age of onset (years, mean ± SD) in the total sample was 19.08 ± 7.24 sample, with men showing an earlier onset (18.12 ± 6.13 years) than women (20.42 ± 8.40 years) (t = 2.191, p = 0.030).

About a quarter (n = 56, 26%) of patients had a history of perinatal traumas. Most patients showed a chronic course (n = 179, 82.5%) and a progressive onset (n = 138, 64.2%) of OCD. On the contrary, 38 (17.5) reported an episodic course and 77 (35.8%) an acute onset. Particularly, 83 (39.0%) patients reported that the onset of OCD was subsequent to a stressful life event, while 53 (24.6%) reported a link with the end of a love relationship. One hundred and twenty-nine (60%) patients had OC personality traits.

The majority of patients suffered from psychiatric comorbidity (n = 128). The most frequent comorbid conditions were bipolar disorder (n = 69), generalized anxiety disorder (n = 16) and panic attack disorder (n = 13). A minority (n = 22) of patients had a previous or current use of substances without reaching the diagnostic threshold for substance use disorder and only seven patients reported at least one past suicide attempt (see table 1).

The Y-BOCS total score (mean ± SD) of the sample was 24.67 ± 7.87 that corresponds to severe OCD. The obsession and compulsion subscales scored, respectively, 12.60 ± 3.90 and 12.07 ± 4.26. No statistically significant gender-related differences emerged.

The most frequent current obsessions were aggressive (64%), followed by symmetry/exactness (45.2%), contamination (42.4%), somatic (29.5%), sexual thoughts and images (16.6%), hoarding (15.7%), and religious obsessions (12.4%). Then, the most frequent current compulsions were checking rituals (72.8%), cleaning/washing (45.2%), repeating (37.8%), ordering/arranging (19.8%), rituals of counting (17.1%), and hoarding/collecting compulsions (9.7%) (table 2). Sexual thoughts and images were significantly more frequent in men, as compared with women (lifetime: χ2 = 6.391, p = 0.011; current: χ2 = 8.966, p = 0.003).

Table 2.

Y-BOCS total scores and obsession/compulsion subscales scores (mean ± standard deviation) in outpatients with OCD

	Total Sample	Men N=126 (58.1%)	Women N=91 (41.9%)
Y-BOCS total scores	24.67 ± 7.87	24.63 ± 7.86	24.71 ± 7.93
Obsession subscale scores	12.60 ± 3.90	12.67 ± 3.86	12.49 ± 3.98
Compulsion subscale scores	12.07 ± 4.26	11.96 ± 4.37	12.22 ± 4.13
Obsessions subtypes	Total Sample	Men	Women
Aggressive	139 (64%)	76 (35%)	63 (29%)
Symmetry/exactness	98 (45.2%)	58 (26.7%)	40 (18.4%)
Contamination	92 (42.4%)	53 (24.4%)	39 (18%)
Somatic	64 (29.5%)	43 (19.8%)	21 (9.7%)
Sexual thoughts/images	36 (16.6%)	29 (13.4%)	7 (3.2%)
Hoarding	34 (15.7%)	19 (8.8%)	15 (6.9%)
Religious	27 (12.4%)	15 (6.9%)	12 (5.5%)
Miscellaneous	159 (73.3%)	97 (44.7%)	62 (28.6%)
Compulsions subtypes
Checking rituals	158 (72.8%)	92 (42.4%)	66 (30.4%)
Cleaning/washing	98 (45.2%)	57 (26.3%)	41 (18.9%)
Repeating	82 (37.8%)	51 (23.5%)	31 (14.3%)
Ordering/arranging	43 (19.8%)	23 (10.6%)	20 (9.2%)
Rituals of counting	37 (17.1%)	24 (11.1%)	13 (6%)
Hoarding/collecting	21 (9.7%)	12 (5.5%)	9 (4.2%)
Other types	129 (59.5%)	72 (33.2%)	57 (26.3%)

At the time of the evaluation, more than a half (n = 133), of patients were taking SSRIs, followed by clomipramine (n = 57), valproic acid (n = 42), gabapentin/pregabalin (n = 31) second generation antipsychotics (n = 29), lithium (n = 20), other mood stabilizers (n = 18), first-generation antipsychotics (n = 12), other antidepressants (n = 10), other tricyclic antidepressants (n = 9), and serotonin-noradrenaline reuptake inhibitors (SNRIs) (n = 4, 1.9%). Sixteen patients referred the occasional use of benzodiazepines for sleep problems. Thirty-three patients were attending cognitive-behavioral therapy (CBT) sessions (table 1). Forty-two patients only were drug-free.

Biological assessments

Antibody titers

Sixty-one patients (61.6%) presented signs of a previous EBV infection (seropositivity for VCA or EBNA IgG), 46 (48.4%) were seropositive for CMV IgG, 24 (25.8%) showed positive ASO titers, 14 (14.3%) were seropositive for T. gondii IgG, four (4.2%) for CMV IgM and nobody was seropositive for T. gondii IgM (table 3).

Table 3.

Infectious agents seropositivity in 100 out of the total 217 OCD patients and in the two sexes

Infectious agents seropositivity	Total Sample	Men	Women
Past EBV infection	61 (61.6%)	36 (36.4%)	25 (25.2%)
IgG CMV	46 (48.4%)	25 (26.3%)	21 (22.1%)
ASO	24 (25.8%)	14 (15.1%)	10 (10.7%)
IgG T. Gondii	14 (14.3%)	8 (8.2%)	6 (6.1%)
IgM CMV	4 (4.2%)	3 (3.1%)	1 (1.1%)
IgM T. Gondii	0 (0%)	0 (0%)	0 (0%)

CMV: cytomegalovirus; EBV: Epstein-Barr virus; ASO: antistreptolysin O; T. Gondii: Toxoplasma Gondii.

Thirty-six patients were seropositive for two, 33 for only one and 12 for three or more antibody titers.

Vitamins and related parameters

Vitamin D levels (mean ± SD, ng/mL) were 19.07 ± 10.27 (men 19.10 ± 9.06, women 19.03 ± 11.90). Only 17 (12%) patients showed levels of vitamin D >30 ng/mL, while 37 (26.1%) showed levels between 20-30 ng/mL. Vitamin D insufficiency (12-20 ng/mL) was detected in 55 (38.7%), deficiency (6-12 ng/mL) in 26 (18.3%) and severe deficiency (<6 ng/mL) in seven (4.9%) patients (table 4).

Table 4.

Vitamins levels (mean + SD) distribution in subsamples of outpatients with OCD

Vitamins		Total Sample		Men		Women
		N (%)		N (%)	mean ± SD	N (%)	mean ± SD
Folate ng/mL	Normal	71 (73.2%)		33 (34%)	4.47 ± 5.65	38 (39.2%)	5.29 ± 4.78
	Deficiency	26 (26.8%)		19 (19.6%)		7 (7.2%)
Vitamin B12 pg/mL	Normal	84 (85.7%)	310.38 ± 262.17	44 (44.9%)	320.82 ± 265.71	40 (40.8%)	298.41 ± 261.51
	Deficiency	14 (14.3%)		10 (10.2%)		4 (4.1%)
Vitamin D ng/ mL	Optimal	17 (12%)		10 (7.0%)		7 (4.9%)
	Sufficiency	37 (26.1%)		23 (16.2%)		14 (9.9%)
	Insufficiency	55 (38.7%)		33 (23.2%)	19.10 ± 9.06	22 (15.5%)	19.03 ± 11.90
	Deficiency	26 (18.3%)		16 (11.3%)		10 (7.0%)
	Severe deficiency	7 (4.9%)		2 (1.4%)		5 (3.5%)
Homocysteine μmol/L			11.88 ± 8.90		13.25 ± 10.63		9.90 ± 5.20

Serum folate levels (mean ± SD, ng/mL) were 4.86 ± 5.24 (men 4.47 ± 5.65, women 5.29 ± 4.78), with only 26 (26.8%) patients showing levels <4 ng/mL, corresponding to folate deficiency. Serum vitamin B12 levels (mean ± SD, pg/mL) were 310.38 ± 262.17 (men 320.82 ± 265.71, women 298.41 ± 261.51), and 14 (14.3%) patients showed levels under sufficiency threshold (<203 pg/mL). Serum homocysteine levels (mean ± SD, μmol/L) were 11.88 ± 8.90 (men 13.25 ± 10.63, women 9.90 ± 5.20), within normal levels (table 4, figure 1).

Figure 1.

Vitamins

No statistically significant differences emerged between the two sexes.

Vitamins levels and Y-BOCS items scores

Vitamin D values were negatively and significantly correlated with Y-BOCS total (r = -0.254, p = 0.002) and both obsessions (r = -0.287, p = 0.001) and compulsions subscales (r = -0.210, p = 0.012) scores, while folate showed a significant and negative correlation with total Y-BOCS score (r = -0.230, p = 0.046) and obsessions subscale score (r = -0.226, p = 0.049) (table 5). In addition, folate showed a negative correlation with “resistance to obsessive thoughts” (r = -0.270, p = 0.019) and “resistance to compulsive behaviour” (r = -0.247, p = 0.031) items. Specifically, patients with sufficient or optimal levels of vitamin D showed lower scores in all Y-BOCS obsessions/compulsions items and with the additional item of “pathological doubting” (for details see table 6).

Table 5.

Correlations between age of onset, vitamins and Y-BOCS total and subscale scores

	Y-BOCS total score		Obsession subscale score		Compulsion subscale score
	r	p	r	p	r	p
Age of onset	-0.197*	0.004	-0.201*	0.003	-0.181*	0.008
Vitamin D	-0.254*	0.002	-0.287*	0.001	-0.210*	0.012
Folic Acid	-0.230*	0.046	-0.226*	0.049	-0.224	0.052
Vitamin B12	0.006	0.960	-0.014	0.905	0.024	0.839

*statistically significant

Table 6.

Positive impact of sufficient or optimal levels of vitamin D on Y-BOCS total and subscales scores

	Mean rank optimal or sufficient vitamin D values (N= 54)	Mean rank not optimal or sufficient vitamin D values (N= 88)	U	z	p
Y-BOCS total	55.98	81.02	1538.000	-3.527	<0.001
Y-BOCS obsessions	55.75	81.16	1525.500	-3.594	<0.001
Y-BOCS compulsions	57.47	80.11	1618.500	-3.198	0.001
Y-BOCS obsessions
Time	55.87	81.09	1532.000	-3.752	<0.001
Interference	58.31	79.60	1663.500	-3.221	0.001
Distress	52.64	83.07	1357.500	-4.550	<0.001
Resistance	61.87	77.41	1856.000	-2.263	0.024
Control	60.37	78.33	1775.000	-2.636	0.008
Y-BOCS compulsions
Time	61.27	77.78	1823.500	-2.445	0.014
Interference	58.87	79.25	1694.000	-3.006	0.003
Distress	58.30	79.60	1663.000	-3.202	0.001
Resistance	59.61	78.80	1734.000	-2.779	0.005
Control	59.21	79.04	1712.500	-2.913	0.004
Pathological doubting	60.60	78.19	1787.500	-2.571	0.010

Interestingly, after dividing the sample into two groups, based on the vitamin D value, i.e. the first group with insufficient values (< 20 ng/mL) and the second group with sufficient or optimal values (> 20 ng/mL), a statistically significant difference emerged in the Y-BOCS total score values. Patients with vitamin D deficiency have higher values (Z = 3.527, p <0.001).

No significant correlations between vitamin B12, homocysteine, ESR, PCR values and Y-BOCS main or subscales scores emerged.

Finally, a positive correlation emerged between folate and vitamin D and B12 levels (respectively r = 0.282, p = 0.018 and r = 0.533, p <0.001). As expected, homocysteine levels positively correlated with ESR values (r = 0.482, p = 0.004).

Forty-one patients presented only one vitamin deficit, 16 presented two hypovitaminosis and only four presented deficits in all vitamins considered. No significant intergroup differences were detected in Y-BOCS total and subscales scores. Finally, we reported the Y-BOCS total scores for different sample subgroups (figure 2 and table 7) and subdivided by the Vitamin D level groups (figure 2).

Figure 2.

Y-BOCS scores divided by subgroups

Legend of figure 2 . Subclinical symptoms = 0-7; Mild = 8-15; Moderate = 16-23; Severe = 24-31; Extreme = 32-40.

Table 7.

Median values of different groups and p value of the median test for independent samples

	Yes (Median and range)	No (Median and range)	p*
Deficit Vitamin D	28.50 (13 - 40)	21.00 (10 - 40)	0.010
Deficit Vitamin B12	27.00 (13 - 40)	27.50 (10 - 40)	0.837
Deficit Folate	27.50 (15 - 40)	27.00 (10 - 40)	0.867
Past EBV infection	28.00 (10 - 40)	20.50 (10 - 40)	0.075
IgM CMV	26.50 (15 - 40)	26.50 (10 - 40)	0.610
IgG CMV	25.00 (13 - 40)	22.00 (15 - 34)	0.538
IgG T. Gondii	21.50 (13 - 40)	27.00 (10 - 40)	0.773
Aso titers	26.50 (12 - 40)	28.00 (10 - 40)	0.777

*Yates’ correction

Figure 3.

Y-BOCS Score subdivided by Vitamin D groups

Legend. 1 = Severely Deficient (<6 ng/ mL); 2 = Deficient (6-12 ng/mL); 3 = Insufficient (12-20 ng/mL); 4 = Sufficient (20-30 ng/mL); 5 = Optimal ((>30 ng/ mL).

Subclinical symptoms = 0-7; Mild = 8-15; Moderate = 16-23; Severe = 24-31; Extreme = 32-40.

Inflammatory markers

Leukocyte count, CRP and ESR mean values were all within normal range of reference for the total sample and in the two sexes (data not shown). Patients taking or not taking medications and with or without comorbid conditions showed no differences on any parameter.

Infections and inflammatory markers and Y-BOCS items scores

Patients with signs of a past EBV infections displayed higher scores in Y-BOCS total (z = 2.135, p = 0.033) and compulsions subscales (z = 2.253, p = 0.024), “distress related to obsessions” (z = 2.139, p = 0.032), “resistance over compulsive behaviours” (z = 2.058, p = 0.040) and “control over compulsive behaviours” (z = 2.563, p = 0.010), “avoidance” (z = 2.428, p = 0.015), “indecisiveness” (z = 2.104, p = 0.035), “overvalued sense of responsibility” (z = 2.181, p = 0.029) and “pervasive slowness” (z = 2.235, p = 0.025).

No significant diferences emerged when having one or more seropositivity to viral infections was considered.

Only ESR values showed a positive correlation with patients’ age (r = 0.319, p = 0.023).

Discussion

The present research investigated a set of biological and peripheral immunological markers and their connections with clinical features to better define clinical phenotypes and to clarify the potential role of so-called silent infections, hypovitaminosis and inflammatory markers in a large cohort of 217 adult OCD outpatients. The ensuing results led to diferent findings. Our sample was well balanced in terms of sex composition, with only a slight prevalence of men. Again, the two genders did not significantly difer in terms of socio-demographic features, Y-BOCS total and subscales scores. However, men were younger at the index evaluation, showed a statistically significant earlier onset and experienced a higher lifetime and current presence of sexual obsessional thoughts. Although the literature on the topic is mainly mixed, it is supposed that gender could play a role on OCD onset, course and presentation (Torresan et al., 2013; Mathes et al., 2019). Indeed, OCD seems to be more prevalent amongst males when considering children and adolescents samples, while in adults’ cohorts there seems to be an equal distribution between sexes or a higher prevalence of females (Dell’Osso et al., 2016; Mathes et al., 2019). Therefore, there is a general consensus about the earlier OCD onset (Lensi et al., 1996; Bogetto et al., 1999; Fontenelle et al., 2003; de Mathis et al., 2011) or symptoms worsening in male patients (Torresan et al., 2013; Mathes et al., 2019). In line with our results, men are generally more prone to sexual and/or religious obsessions (Lensi et al., 1996; Bogetto et al., 1999; de Mathis et al., 2011; Torresan et al., 2013; Mathes et al., 2019), while the higher prevalence of contamination obsessions and cleaning/ washing compulsions, which is generally described in women (Lensi et al., 1996; Bogetto et al., 1999; Torresan et al., 2013), was not detected in our sample. As a consequence of the earlier onset and thus the more pervasive impact on global functioning, men are sometimes described as more often single (Lensi et al., 1996; Bogetto et al., 1999), unemployed and living with the original family (Torresan et al., 2013; Benatti et al., 2022), a finding that was not replicated in our study. As described, these results are partially in line with what globally emerges from available literature; however, gender issues are complex and entail biological, psychological and socio-cultural interacting variables which could be responsible for the overall heterogeneity of findings (Torresan et al., 2013).

Individuals with an early onset, a story of perinatal traumas and /or obstetric complications (such as a dystocic or breech delivery, with application of forceps, prolonged hypoxia), or chronic course also showed higher Y-BOCS scores, reflecting more distressing and interfering OC global symptomatology. Early-onset OCD (<17 years old) has been described as characterized by a male gender predominance, greater number of clinically significant OC symptoms with higher Y-BOCS scores, especially the compulsions subscale (Rosario-Campos et al., 2001), higher frequency of repetitive rituals, apparent initial lower responsiveness to therapies, higher frequency of sensory phenomena preceding the compulsions, tic-like compulsions and comorbid tic disorders (Sobin et al., 2000; Rosario-Campos et al., 2001; Fontenelle et al., 2003), potentially outlining a subgroup of patients with phenotypic specificity (Fineberg et al., 2019). This hypothesis is further corroborated from genetic studies reporting specific gene variants of serotonergic, dopaminergic and glutamatergic receptors and synaptic brain related genes, neurotrophic and transcription factors associated with early-onset OCD (Fineberg et al., 2019). Some studies detected a more frequent history of gestation and delivery complications in OCD patients than control subjects (Vasconcelos et al., 2007; Geller et al., 2008). In particular, perinatal traumas, other than the potential damage caused to brain structures which can disrupt the development of CSTC, that is involved in OCD pathogenesis (Geller et al., 2008), could be responsible for a premature impairment of the HPA axis, which constitutes the main responsible system of stress-related response (Real et al., 2011). Obstetrical complications have been linked with impairment of HPA axis in children and adolescents (Taylor et al., 2000; Real et al., 2011) so that could constitute an early hit to the HPA axis determining an increased vulnerability to stressful life events to a later life (Real et al., 2011). These findings suggest that early insults could influence the later expression of psychopathology, as well as data from genetic studies suggests that non-genetic environmental factors could explain a considerable proportion of the variance in OCD or OC spectrum symptom occurrence (Geller et al., 2008). On the contrary, when the onset was related to a specific stressful life event (SLE), including the end of an affective relationship, the outcomes seemed slightly better in terms of overall symptomatology (lower Y-BOCS scores). Some authors recognized “affective relationships” as risk factors for OCD, especially when just started (Thompson et al., 2020). Therefore, OCD and being in love have been hypothesized to share neurobiological mechanisms involving the serotonin (5-HT) system, besides others (Marazziti et al., 1999). Indeed, strongly emotional situations, such as onset of romantic love, intimate relationships, or a forthcoming marriage can, indeed, provoke a critical reduction in 5-HT levels that may become pathological in individuals with a pre-existing vulnerability, triggering onset or worsening of OCD symptomatology (Marazziti, 2020). Previous studies outlined a potential distinct clinical pattern when the onset of OCD is related to a SLE, defined as a “stress-life related subtype” of OCD, characterized by later onset, history of complicated birth, less family history of OCD and prevalence of contamination/cleaning symptoms (Real et al., 2011) and more presence of sensory phenomena (Kracker Imthon et al., 2020). Unfortunately, we did not apply a standardized method to investigate SLE types, but only its presence or absence. This could have made the data recollection susceptible to recall biases and overlooked some particular life events, like pregnancy and postpartum, which are not always considered as SLEs, but should be taken into specific account for their high affective and biological impact. On the other hand, from our data emerged that individuals with a SLE before OCD onset showed lower Y-BOCS total and subscales scores, a finding not reported in previous studies.

Married and employed patients both showed better prognostic characteristics as shown by later age of onset and lower severity of symptomatology, so that these could be assumed as features associated with better levels of social functioning.

It should be noticed that individuals displaying OC personality traits, such as cognitive and behavioral inflexibility, perfectionism, conscientiousness, dutifulness, hoarding, and orderliness, seemed to have a marked overvalued sense of responsibility, demonstrated a lower insight level and appeared less reliable to clinicians. Obsessive-compulsive personality disorder (OCPD) is the most frequent personality disorder found in patients suffering from OCD (Pozza et al., 2021) and entails greater overall OCD severity, functional impairment, poorer insight and treatment outcomes (Coles et al., 2008; Lochner et al., 2011; Wetterneck et al., 2011). However, these results suggest that patients with more pronounced OC personality traits also show characteristics in insight and reliability which could potentially impact on outcomes influencing access to health services, adherence to therapy and treatment targets.

Patients attending CBT sessions adjunctive to the pharmacological therapy displayed benefit in some elements implying the compulsive and avoidant symptomatology. These results further assess the importance of psychotherapies for peculiar aspects of OCD manifestations, mainly stimulating the development of more adaptive behavioral responses.

The biological assessments led to intriguing data. Antibodies titers and inflammatory markers were evaluated in 100 patients of whom more than 60 % resulted positive to at least one infection, with men showing slightly higher seropositivity rates than women for all pathogens considered. Interestingly, no patients recalled a previous infection. Sixty-one patients were positive for EBV infection, 46 for CMV IgG, 24 for ASO titers, 14 for T. gondii IgG, and four for CMV IgM. No difference was noted in clinical features between patients with or without infections. However, surprisingly, the EBV seropositivity was significantly related to higher scores of Y-BOCS total and compulsions subscale, and in some specific aspects of obsessive/compulsive symptomatology, that is “distress related to obsessions”, “resistance” and “control over compulsive behaviours”, “avoidance”, “indecisiveness”, “overvalued sense of responsibility” and “pervasive slowness”. Although no significant differences emerged when one or more seropositivity was considered, a trend towards a greater severity was noted in those in patients with three infections and hypovitaminosis (Watson et al., 2013; Hamdani et al., 2017; Abdoli et al., 2020). All the infective agents considered herein have been associated with detrimental consequences towards CNS tissues. The IGs produced against Group A beta-hemolytic streptococcal (GABHS) components can cross-react with endogenous proteins of the basal ganglia (Kirvan et al., 2006), and CMV, EBV, and Toxoplasma gondii, thanks to their neurotropic features, cause latent CNS infection after the acute phase (Virus et al., 2021; Della Vecchia & Marazziti, 2022). Again, latent toxoplasmosis was associated with alterations in dopaminergic (Gaskell et al., 2009; Prandovszky et al., 2011), GABAergic, glutamatergic, serotonergic, and noradrenergic (Henriquez et al., 2009; Fabiani et al., 2015; Kannan et al., 2016; McFarland et al., 2018) transmissions, as well as disruption of the HPA axis (Henriquez et al., 2009). Infections from HSV-1, Borna disease virus (BDV), GABHS, Borrelia burgdorferi, and Toxoplasma gondii are described in patients with OCD (Della Vecchia & Marazziti, 2022). In any case, available data are scarce and jeopardized due to the heterogeneity of methods and samples considered. More information is available for EBV, that is the pathogen primary responsible of infectious mononucleosis, but that in the context of both acute/ reactivation and chronic infections may also cause encephalitis, meningitis, cerebellitis, cranial or peripheral neuropathies and polyradiculomyelitis, that is a post-infectious immune-mediated syndrome (Ludlow et al., 2016) and perhaps multiple sclerosis (Bjornevik et al., 2022). Basal ganglia belonging to the so-called OCD circuit are equally involved together with the cerebral hemisphere and cerebellum during the EBV infection, but patients with thalamic involvement are more prone to develop long-term sequelae (Ludlow et al., 2016). Therefore, there seems to be a particular engagement of brain structures related to OCD by EBV, that infects 90% of humans and persists long-life (Cohen, 2000). Some peculiar aspects which seemed particularly affected in this population (avoidance, indecisiveness, pervasive slowness) resemble “sickness behavior”, a complex of symptoms observed during acute infections, characterized by changes in mood, social behavior and cognitive abilities. It is an immune-mediated process thought to be a model for cytokine-dependent behavioral changes, due to a state of inflammation, independently of the specificity of the causative pathogen. It is supposed to have an adaptive function increasing the individual’s chances of survival (McCusker & Kelley, 2013) during a low-grade transient inflammation, but, when it becomes chronic, it could lead to neurotoxicity and neurodegeneration determining impairment of cognitive performances (Pape et al., 2019). Herpes viruses’ latent infection have been linked to impaired cognitive performance in animal models as well as in humans, in both psychiatric patients and healthy controls (Watson et al., 2013; Pape et al., 2019), but no studies have been conducted on EBV infections and OCD so far. In line with a previous study carried out by our research team on a smaller sample of OCD outpatients (Marazziti et al., 2023), almost a half of individuals showed serum levels of vitamin D below the sufficiency threshold. Unlike previous results, however, lower vitamin D levels were significantly associated with all Y-BOCS total and specific subscales scores and with every item considered. Again, the most severe patients showed the lowest vitamin D levels. The larger sample size might have helped in emphasizing the potential role of vitamin D in OCD global symptomatology, particularly in terms of overall severity. This finding is consistent with recent data (Soyak & Karakükcü, 2022). Available literature does not allow to clearly determine a higher susceptibility of OCD patients to vitamin D deficiency than the general population (Balandeh et al., 2021), it still suggests that both young and adult OCD patients are highly exposed to vitamin D insufficiency (Celik et al., 2016; Esnafoğlu & Yaman, 2017; Soyak & Karakükcü, 2022) that could contribute, at least, to worsen global symptomatology. Additionally, when patients with sufficient or optimal vitamin D values were specifically considered, significant lower scores in all main Y-BOCS domains were detected. Although current reference values are defined exclusively on the bone outcomes irrespectively to wider systemic implications (Bivona et al., 2019), this result could be an incentive to further clarifying studies. Indeed, vitamin D, through its immune-modulatory action, exerts a neuroprotective function in CNS, and its deficiency is associated with enhanced pro-inflammatory , state, increased formation of Aβ oligomers and reduced amyloid clearance in the hippocampus, which may accelerate age-related cognitive decline and contribute to the pathophysiology of several neuropsychiatric diseases (Marazziti et al., 2021). Given its pleiotropic functions on neurotransmission modulation, especially monoamine transmission, neuroplasticity and neuroprotection, more attention should be paid to the evaluation of vitamin D status in the basal clinical examination of psychiatric, particularly OCD, patients. Although folate levels were mostly within the normal range, a significantly negative correlation was revealed between folate levels and Y-BOCS total and obsessions scores and with both obsessions and compulsions resistance items. Literature on the topic is scarce and characterized by few small studies showing contradictory results (Atmaca et al., 2005; Türksoy et al., 2014; Esnafoğlu & Yaman, 2017; Balandeh et al., 2021; Yan et al., 2022). Only one study demonstrated a negative correlation between serum folate levels and Y-BOCS scores, although only total scores were considered (Atmaca et al., 2005). In our opinion, our finding suggests the possible impact of folic acid on specific OCD symptoms of OCD and confirms the need for further studies on this topic. Both vitamin B12 and homocysteine were within the normal range in our sample, nor any relationship between the two or Y-BOCS domains was found. A few data, mainly consisting of retrospective studies, show a vitamin B12 deficiency in OCD (Hermesh et al., 1998; Türksoy et al., 2014; Balandeh et al., 2021; Yan et al., 2022). Only one study conducted on OCD children and adolescents demonstrated a negative correlation with Y-BOCS scores (Esnafoğlu & Yaman, 2017). Similarly, some studies detected higher levels of homocysteine in OCD patients and their positive correlation with Y-BOCS scores (Atmaca et al., 2005; Türksoy et al., 2014). This limited availability could have contributed to the lack of significant results. Hyperhomocysteinemia alters glutathione metabolism impairing DNA repairing processes and promoting damages induced by oxidative stress and accumulation of β-amyloid. Therefore, high levels of homocysteine and low levels of folate seem to accelerate and amplify CNS age-induced damages (Kruman et al., 2002; Reynolds, 2006). Folate and vitamin B12 are demonstrated to be essential for neuronal differentiation and damage-repair cellular mechanisms, acting on nucleotide synthesis and, therefore, DNA integrity, epigenetic processes and neurotransmission, and finally influencing brain growth, development, several mental functions and ageing (Reynolds, 2006). Considering that their evaluation requires routine, inexpensive and non-invasive methods, it may be worth taking it into account on routine assessment, to deepen their connections with clinical presentations and therapeutic implications.

The large sample size and the accurate characterization of the patients, well balanced in terms of sex, represent the main strengths of this study. The main limitations are the following. First, a group of healthy controls is missing, so that the seropositivity and vitamins or immune peripheral parameters were compared to normative values only. Furthermore, this implies that it was not possible to evaluate the possible impact of some factors such as lifestyle, diet, region of origin, etc., which could influence the values of the variables considered and consequently the psychopathological presentation. Second, the study had a cross-sectional design, therefore, it did not allow to infer causal relationships between the variables described above. Although deriving from biomarkers tested in different laboratories, these findings spark renewed interest on the topic with compelling need of further investigation.

To summarize, the present study highlighted some gender-related differences regarding OCD: men showed younger age at index evaluation, younger age of onset, and higher prevalence of lifetime and current sexual obsessional thought, although no significant differences in Y-BOCS scores were detected between the two sexes. An earlier disease onset, a history of perinatal traumas and a chronic course were associated with more severe symptoms, while acute onset and preceded by stressful life events was related to better clinical characteristics. More than a half of the tested patients showed seropositivity for a previous EBV infection, followed by those positive for CMV, ASO and T. gondii, with no past history of such infections. Although all infections should be considered, it should be stressed that only EBV seropositivity correlated with the severity and all OCD symptoms, as assessed by the Y-BOCS. These findings suggest that might be more specifically related to OCD pathophysiology and/or clinical picture than the other agents. The same seems true for vitamin D and perhaps folate serum.

Conclusions

There is still a compelling need for a novel pathophysiological model of OCD, aiming at improving an early detection and effective treatment, while considering the impact on patients’ quality of life and functioning, as well as its direct and indirect community costs. The identification of particular phenotypic patterns might help in clinical evaluation suggesting novel therapeutic approaches towards specific biological targets, such as vitamin integration or antibiotics. The evaluation of serum biological markers, such as seropositivity to common viral and bacterial agents, vitamins or inflammatory markers, in routine clinical practice might be a first step that could open unexpected perspectives in terms of prevention and novel treatment approaches. Given the impact on overall brain homeostasis, longitudinal and controlled investigations are needed to deepen the role of common bacterial and viral infections that, albeit silent, could be detrimental for the brain in the long-term. Interestingly, a recent paper highlighted a relationship between EBV and multiple sclerosis (Bjornevik et al, 2022) .The impact of hypovitaminosis on OCD pathophysiology, particularly of vitamin D, seems also useful. Taken together, the findings of this study suggests that new therapeutic strategies, namely, immunomodulatory compounds or vitamin supplementation, might be useful at least in a subgroup of OCD patients.