Mental health problems and low Vitamin D feed each other – 14 conditions reviewed

Bottom line

Low vitamin D is linked to many health problems, and many health problems lower vitamin D. Each feeds the other.
For depression, genetic (Mendelian randomization) studies now point in both directions.
Six common ways a health problem lowers vitamin D: less sun, poorer diet, inflammation, obesity, medications, and the disease itself.
Three consequences: observational studies overstate what vitamin D alone does; sick people usually need more vitamin D to reach the same blood level; trials in people who already have enough vitamin D tend to find nothing.

The vicious cycle

Low vitamin D → higher risk of the problem → less time outdoors, poorer appetite, inflammation, weight gain, medications → even lower vitamin D → ...

6 ways a health problem lowers vitamin D

Less sun – people who are depressed or withdrawn get less sunlight [Konzok 2023]; dementia may keep people indoors [NutraIngredients 2023]
Poorer diet – loss of appetite, common in mental illness, reduces vitamin D intake and absorption [Konzok 2023]
Inflammation – after knee surgery, 25(OH)D fell about 40% within 2 days and was still about 20% lower 3 months later [[Reid 2011](https://eprints.gla.ac.uk/51689)] [Clinician summary]
Obesity – each 10% higher genetically predicted BMI gave 4.2% lower 25(OH)D, while vitamin D had little effect on BMI [Vimaleswaran 2013]; obese people need 2–3× the dose [Ekwaru 2014]
Medications – people on enzyme-inducing anti-seizure drugs needed about 1,590 vs 1,110 IU/day to reach the same level [AES 2018]
The disease itself – genetic risk of depression lowers 25(OH)D [Mulugeta 2020]; schizophrenia, autism and bipolar disorder may also lower it [Jiang 2023]

The "illness lowers vitamin D" view was argued most strongly by Autier, who concluded from 290 cohort studies and 172 RCTs that low 25(OH)D is mainly a marker of ill health [Autier 2014]. Newer genetic studies suggest that for dementia, depression and schizophrenia it is both a marker and a cause.

Mental health problems: evidence in each direction

Evidence tiers

Two-way, with genetic evidence: Depression [[Mulugeta 2020](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7823924/)] [Wang 2026], Schizophrenia [Jiang 2023]
Genetic evidence, one direction: Dementia (low D → dementia) [Navale 2022]; Autism and Bipolar (problem → low D) [Jiang 2023]
Observational only: the other 9 (a genetic study of anxiety found no link)

Problem	Low D linked to problem	Problem lowers D	Vitamin D trials	VitaminDWiki
Depression	Yes; 1 of 2 genetic studies agree Wang 2026	Yes, genetic evidence Mulugeta 2020	Small benefit, more if deficient Ghaemi 2024	Depression
Postpartum depression	OR 3.67 if under 20 ng/mL Wang 2018	Not studied	1,800–3,500 IU/day: small-to-medium benefit Tsai 2023	Pregnancy + Depression
Seasonal affective disorder	Suspected, not proven	Not studied	3 RCTs: no benefit EBM review	SAD
Anxiety	Genetic study: no link Wang 2026	Not studied	No effect Ghaemi 2024	Anxiety
Bipolar	Only 6% of 290 patients (mostly bipolar) had enough D Medical Dialogues	Genetic evidence Jiang 2023	Mixed Life 2023 review	Bipolar
Schizophrenia	OR 2.16 if deficient Valipour 2014	Genetic evidence Jiang 2023	3 small RCTs: some gains in cognition and negative symptoms Tsiglopoulos 2022	Schizophrenia
Suicide attempts	Highest risk under 15.5 ng/mL Military study 2013	Not studied	No RCTs; VA cohort: ~45% fewer attempts on D Lavigne 2023	Suicide
OCD	Lower levels in patients Othman 2024	Not studied	None found	OCD
ADHD	OR 1.97 if deficient MECP 2025	Not studied	4 small RCTs: small benefit added to methylphenidate Gan 2019	ADHD
Autism	Low maternal D linked to autism Prenatal review	Genetic evidence Jiang 2023	Stereotyped behavior improved; most core symptoms not CPN 2023	Autism + Vitamin D
Sleep problems	Lower D with sleep disorders Food Sci Nutr 2020	Not studied	Better sleep quality, moderate certainty Abboud 2022	Sleep
Delirium	2.5× risk if under 10 ng/mL Pilling 2020	Likely: acute illness lowers D Reid 2011	None found	Delirium
Cognitive decline	Modest link J Alzheimers Dis 2013	Not studied	Small benefit, larger if deficient Chen 2023	Cognitive
Dementia and Alzheimer's	Genetic: 54% more risk at 10 vs 20 ng/mL Navale 2022	Possibly: less time outdoors NutraIngredients 2023	No long RCTs; cohort: 40% less dementia on D Ghahremani 2023	Dementia

Why the two-way link matters

Observational studies overstate vitamin D's effect – some of the low vitamin D seen in patients is a result of the illness, not its cause [Autier 2014]
Trials in people who already have enough vitamin D miss the effect – VITAL-DEP participants averaged 31 ng/mL at the start [MDedge]; the genetic dementia effect appeared only in the deficient range [Navale 2022]
Sick people need more vitamin D – 2–3× the dose if obese [Ekwaru 2014], about 40% more on enzyme-inducing anti-seizure drugs [AES 2018]
A test during acute illness can mislead – 25(OH)D falls during inflammation, so a level measured then is not a reliable guide to status [Reid 2011]
Treat the people caught in the cycle – trial benefits show up mainly in people who are deficient or already depressed [Cheng 2020]

Caution for people on antipsychotics: higher vitamin D was linked to lower blood levels of aripiprazole and quetiapine (both cleared by CYP3A4); the authors advise checking drug levels when supplementing [Keskin 2022]

What this does NOT show

That vitamin D alone treats these problems – most trial effects are small, and the anxiety and SAD trials found none
That every problem in the table lowers vitamin D – that direction has been tested genetically for only 4 of the 14
That the genetic studies settle it – the 2 depression studies disagree on whether low D causes depression, and earlier genetic studies of depression and schizophrenia found no causal effect [PMC7107583]
That the cohort results (40% less dementia, ~45% fewer suicide attempts) are causal – people who take supplements may be healthier in other ways
What dose or blood level is best – most trials used modest doses for short periods