Folate levels in psychiatric outpatients


Correspondence address: AndrésHerrán MD, Clinical and Social Psychiatry Research Unit, Department of Psychiatry, University Hospital ‘Marqués de Valdecilla’, Avda. de Valdecilla 39008, Santander, Spain. Email: <>


This study examines folate in psychiatric outpatients. Fifty-three outpatients with schizophrenia and 24 outpatients with depressive disorder assessed with the Schedules for Clinical Assessment in Neuropsychiatry interview are included. Patients with schizophrenia had lower serum folate levels than age- and sex-matched controls, while red cell folate levels did not differ. Serum folate levels showed a negative correlation with the Clinical Global Impression, disorganized dimension, and total Positive and Negative Syndrome Scale score. Patients with depressive disorder had lower serum folate levels than healthy controls, but showed no differences in red cell folate levels. Only two patients with schizophrenia had red cell folate levels below the normal range.


Psychiatric illness has repeatedly been shown to be related to low levels of folate. In particular, folate deficiency has been related to depressive disorders, while results found in patients with psychotic disorders have been less conclusive. 1,2 A related area of interest has been the use of folate as an adjunct to therapy. 3 Despite the research in this field, results in different studies are inconclusive. Skerritt has recently examined a number of factors affecting these results; small samples, different exclusion criteria, unreliable measurement methods, and study of serum but not red cell folate levels. 4 Especially relevant is the lack of studies which include controls matched for age and sex. 1,2 This present study assesses the levels of serum and red cell folate in a representative sample of outpatients with schizophrenia or depressive dis- orders and compares them with age- and sex-matched healthy controls.


Schizophrenic patients

All schizophrenic patients receiving regular treatment from an outpatient department of a Mental Health Unit were included in the initial sample. This de- partment attends a population of 46 735 people. We selected patients fitting Diagnostic and Statistical Manual (DSM-IV) criteria of schizophrenia 5 by agreement of two senior psychiatrists. Also, in order to ensure the diagnosis, patients were evaluated with the Item Group Checklist (IGC) section of the Schedules for Clinical Assessment in Neuropsychiatry (SCAN). 6,7 From the 81 patients constituting the initial sample, 12 were excluded because they did not meet criteria for schizophrenia. A further seven were excluded for physically-related reasons (AIDS (1), diabetes mellitus (2), chronic diarrhoea (1), gal- lstones (1), hypochromic anaemia (1) and pregnancy (1)). Additional exclusions involved one patient on vitamin supplement treatment, two patients with a history of irregular contact, three uncooperative patients, and three with incomplete data on record. Consequently, 28 patients were excluded from the initial sample, and 53 (28 male, 25 female; mean age 37.5 years) patients were included in the study. All were receiving regular antipsychotic treatment. Five had not previously undergone pharmacological treatment before attending the mental health unit.

Patients with depressive disorder

Consecutive new patients referred to the mental health unit were included in the initial sample. Those with clinical features compatible with depressive disorder were assessed with the SCAN interview. Finally, 24 patients (three male, 21 female; mean age 43.6 years) with DSM-IV criteria of major depressive episode were included. They were not taking treatment before analytical procedures.

Healthy controls

We included healthy hospital staff volunteers. Fifty-five controls (29 male, 26 female; mean age 37.6 years) were used for comparison with schizophrenic patients. Thirty-one (4 male, 27 female; mean age 41.4 years) were compared with depressive patients. None of the patients or the controls were vegetarians. Except for the schizophrenic patients excluded from the study, none of the patients had conditions or were taking drugs known to be associated with vitamin deficiency. Alcohol consumption in patients with schizophrenia (48%) did not differ from consumption in the healthy controls (37%) (chi-squared test: x2 = 1.13, d.f. = 1, P = 0.2).


Clinical data

Demographic, social, and medical history variables were assessed with a specially designed questionnaire. Additional data regarding evolution and treatment were derived from patient’s clinical histories. For schizophrenic patients, clinical information included DSM-IV schizophrenia dimensions 5 (disorganized, psychotic and negative), the Clinical Global Impression (CGI), 8 and the Positive and Negative Syndrome Scale (PANSS). 9 Patients with major depression were assessed with the 17-item Hamilton depression scale. 10,11

Analytical procedures

Early morning fasting blood samples were taken for both patients and controls. Folate was determined with commercial specific radioimmunoassay (RIA) (ICN Pharmaceutical Division Diagnostics, Costa Mesa, CA, USA). The laboratory defines low serum folate levels as < 1.5 ng/mL, and low red cell folate levels as < 120 ng/mL. Data were analyzed with the statistical package SPSS (Statistical Package for the Social Sciences) Release 7.0. 12 Comparisons between groups (schizophrenia–control and major depression– control) were made using the Mann–Whitney U-test. Comparison among schizophrenia types was made with the Kruskal–Wallis test. For correlations of folate levels and clinical features, Spearman correlation was used.


Schizophrenic group

The most prevalent type was residual schizophrenia (21 patients). Other types were paranoid (15), undifferentiated (10), and disorganized (7). The mean time from the onset of the illness was 12.7 years (standard deviation (SD), 10.8).

Serum folate

Mean serum folate in schizophrenic patients was 4.7 ng/mL (C.I. 95%: 4.2–5.2). Corresponding figures for controls were 6.5 ng/mL (C.I. 5.8–7.2). the Mann–Whitney U-test showed significant differences (P = 0.00), with patients having lower serum folate levels than controls. Patients with a diagnosis of disorganized schizophrenia showed the lower serum folate values (mean 4.0 ng/mL), but the Kruskal–Wallis test did not demonstrate differences among different types (P = 0.7).

For the total schizophrenia group, serum folate levels showed a strong negative correlation with CGI (P = 0.00), disorganized dimension (P = 0.03), and total PANSS score (P = 0.02). Current antipsychotic doses (equivalents of chlorpromazine/day) did not correlate with serum folate levels (P = 0.1). Neither patients nor controls had serum folate levels outside the normal range.

Blood cell folate

Mean red cell folate in patients was 313.7 ng/mL (C.I. 95%: 275.1–351.8) and in controls was 358.3 ng/mL (C.I. 95%: 311.5–405.0). The differences did not reach a significant level (P = 0.1). Disorganized schizophrenics again showed the lower red cell folate levels (mean 266.2 ng/mL), but the different subgroups did not differ significantly (P = 0.7).

Current antipsychotic doses correlated negatively with blood cell folate levels (P = 0.02). Two patients had red cell folate levels slightly below the normal range (113 and 118 ng/mL).

Major depression group

The 24 depressive patients had a mean of 20.6 on the Hamilton scale and were compared with 31 age- and sex-matched controls.

Serum folate

Mean serum folate in depressive patients was 5.0 ng/mL (C.I. 95%: 4.3–5.8), and in healthy controls 6.7 (C.I.: 5.7–7.7), a significant difference (P = 0.01). Serum folate levels did not correlate with depression severity as measured with the Hamilton scale. All patients had values within normal range.

Blood cell folate

Depressive patients had a mean of 336.0 ng/mL red cell folate (C.I. 239.1–432.8). The mean for the controls was 381.1 ng/mL (C.I. 305.6–456.6). The difference was not significant (P = 0.2). One healthy control was found to have a serum folate level of 64 ng/mL, below the normal range.


As has been stated, the failure in recent studies to replicate earlier works demonstrating significant folate deficiency in psychiatric patients is due to a number of reasons. 4 Lack of stringent methodology could be the main reason. We excluded patients with conditions known to be associated with folate deficiency, and also those without strict diagnostic criteria, by applying the SCAN interview. Only two patients with schizophrenia had (red cell) folate below reference values.

We found that clinical features correlated with serum folate levels. Disorganized symptoms and ‘global severity’ (measured with CGI and total PANSS scores) correlated negatively with serum folate levels. The higher the degree of severity, the lower the serum folate levels. However, neuroleptic doses did not show a relationship with folate levels. The lower level of serum folate could be related to the pathogenic process of schizophrenia, as it has been suggested by some authors 13 or could be a reflection of changes in habits in those patients with worse clinical features (nutritional factors and so on).

The lack of statistical differences in red cell folate levels may be a reflection of a type II error or may represent a true lack of differences. The negative correlation found with antipsychotic doses is difficult to interpret, since it could be confounded by the effect of the clinical severity over antipsychotic doses.

For patients with depressive disorder, the results are comparable, with lower serum folate levels than controls, but similar red cell folate levels. These similarities are relevant, given the differences in the pathogenesis, treatment, and global functioning between patients with schizophrenia and patients with depression.

The main contribution of this study has been to assess serum and red cell folate levels in a representative sample of outpatients with either schizophrenia or depressive disorders, and to compare them with age- and sex-matched controls. This has circumvented the problem of potential confounding factors (non-representative samples, doubtful diagnostic criteria, hospitalization status, and lack of controls) of previous studies. We have found folate levels within the normal range, but lower serum folate levels, as a group, in psychiatric outpatients as compared to healthy controls.