Low vitamin D turned up far more often in people with OCD. As a treatment, the review could point to one case report

Blood tests for vitamin D have been compared in people with obsessive-compulsive disorder and people without it for years, and the small studies have not always agreed. A review in Frontiers in Nutrition has now pooled them. The authors searched the literature to March 2026 and kept ten observational studies, seven in adults and three in children and adolescents, all using DSM-IV or DSM-5 criteria to identify OCD.

The comparison of average levels covered eight studies, with 507 people with OCD and 506 controls. Levels were lower in the OCD groups by about two-thirds of a standard deviation (standardised mean difference −0.669, 95% CI −0.943 to −0.394). Six studies also counted people as deficient or not. The pooled odds ratio was 4.06, with a wide confidence interval of 1.98 to 8.31. Adding up the counts in the paper's table gives 233 of 395 people with OCD (59%) and 125 of 417 controls (30%). That sum is ours and is cruder than the review's weighted estimate, but it shows the size of the gap in plain numbers.

Lower vitamin D also went with worse symptoms on the Yale-Brown scale. The pooled correlation with total score was −0.361 across six studies and 511 people. Squared, that comes to roughly 13% of the variation in scores, so most of the differences between people were not accounted for by this blood test. The link looked stronger in adults (−0.598) than in children (−0.336). For compulsions alone it vanished in children (0.031, CI −0.174 to 0.236), though that estimate rests on two studies and the authors warn against reading much into it.

The closer you look at the studies, the less tidy the picture gets. Between 73% and 87% of the variation in each main result came from differences between studies rather than chance. Deficiency meant below 20 ng/mL in two studies, below 15 in one and below 10 in another, and two did not state a cut-off. One comparison group was made up of other psychiatric patients, not healthy people, and 58% of them were deficient, against 68% of the OCD group. One study of 344 people, the only one using LC-MS/MS, reported deficiency in 56% of cases against 11% of controls and a correlation of −0.705. It was flagged as unusually influential in the odds and correlation analyses. Two more cohorts may overlap, which could count some patients twice.

The bigger gap is direction. Low vitamin D could feed into OCD through its effects on the brain, as the authors suggest. OCD could also lower vitamin D, because avoiding outdoor activities and eating a restricted diet are both patterns the authors mention. Sun exposure, season, latitude, diet, gender and other illnesses could not be examined because the original studies did not supply the data. The authors themselves say causality cannot be established and call the evidence for vitamin D as an independent contributor inconclusive.

Their position on treatment is plainer than the headline numbers might suggest. Screening people with OCD for low vitamin D could be worth investigating, they write, but the same cannot be said for routine supplements as part of OCD treatment. Apart from a single case report, clinical evidence is largely missing, and they ask for proper trials. So a person with OCD who has other reasons to suspect low vitamin D, such as little time outdoors or a very narrow diet, has a sensible question for their doctor about a blood test. Nothing here supports swapping therapy or medication for a supplement.