The Season in the Blood

The money, and what it does not prove

In 2025, the pharmaceutical and health-products industry reported 457.3 million dollars in United States federal lobbying, the largest industry total in the OpenSecrets tally of Senate disclosures. The Pharmaceutical Research and Manufacturers of America alone reported 38.2 million dollars that year. What the record shows is an imbalance any reader can audit. A hormone the skin makes from daylight, and a capsule that costs less than a copay, have no trade association that can spend like that. The companies that sell the drugs for the same symptoms can, and do, and they spend it on price, patent, coverage, and the rules of the market.

The bend that is documented is in the journals. Turner and colleagues, in the New England Journal of Medicine in 2008, set FDA-registered antidepressant trials next to the papers that reached print. Studies the agency judged negative or questionable were often left unpublished, or published with a conclusion that did not match the agency’s reading. The public argument was shaped by the trials that flattered the drugs. That is not a theory about vitamin D. It is a finding about how the first explanation, the prescription, was allowed to look cleaner than its own data.

Blame has no shortage of other addresses. Serotonin. Screens. Genes. Trauma. Character. Trauma is real. Genes are real. A phone at midnight will move a clock. None of those explanations draws 25-hydroxyvitamin D, and none of them requires the clinic to ask when the person last stood in bright morning light. The blood test is not a rival religion. It is a number.

Two routes, and only one of them is billed

Daylight reaches the brain by 2 routes, and they are not the same fact. Ultraviolet B on bare skin makes vitamin D3. The liver turns that into the storage form, 25-hydroxyvitamin D. The kidney, and some immune cells, make the active hormone after that. The second route never touches that chemistry. Morning light, even under cloud, is far brighter than a room, and it enters at the retina. A small set of cells there uses melanopsin, not the rods and cones that build a picture, and those cells tell the suprachiasmatic nucleus what time it is. Kenneth Wright and colleagues, in Current Biology in 2013, took people camping in the Rockies with no electric light. Within a week, melatonin and sleep had shifted back toward sunset and sunrise. The sun restored the night. No one in that camp needed a vitamin D result to feel the change.

Blame has no shortage of other addresses.

Psychiatry already uses the retinal route on purpose. Bright light, classically 10,000 lux in the morning, is an accepted treatment for winter depression. Golden and colleagues, in the American Journal of Psychiatry in 2005, pooled the trials and found a clear benefit for seasonal illness and a smaller one outside winter. Lam and colleagues, in JAMA Psychiatry in 2016, set 10,000 lux against fluoxetine in nonseasonal major depression. On the depression scale, light alone beat placebo. Fluoxetine alone did not. Light plus fluoxetine also beat placebo. That result belongs to the light. The sky, the street, and the horizon carry the dose. The vitamin is the other door, and it has a literature of its own.

The hormone is already at home in the brain. Eyles, Smith, Kinobe, Hewison, and McGrath, in the Journal of Chemical Neuroanatomy in 2005, mapped, in human brain tissue, the vitamin D receptor and the enzyme that activates the hormone. The receptor is present in neurons, in glia, and in regions that handle planning and mood. Patrick and Ames, in the FASEB Journal in 2014, argued that vitamin D helps govern the enzyme that makes serotonin in the brain. A hypothesis is not a trial. It is why the trials were worth the trouble, and why a low level is more than a cosmetic flag on a lab slip.

The same latitudes, counted twice

The map drawn for the body is the map the mind is standing on. In NHANES 2005 to 2006, deficiency meant a serum 25-hydroxyvitamin D at or below 20 ng/mL. 30.9% of non-Hispanic White adults were at or under that line. 69.2% of Hispanic adults were. 82.1% of non-Hispanic Black adults were. Kumar and colleagues, in Pediatrics in 2009, showed the split was already there in childhood. After adjustment, a non-Hispanic Black child had an odds ratio of 21.9 for a level under 15 ng/mL, compared with a non-Hispanic White child (95% CI 13.4 to 35.7). Among girls aged 13 to 21, about 59% of non-Hispanic Black girls, but only 5% of non-Hispanic White girls were under 15 ng/mL.

Anglin and colleagues put a figure on the adult pattern in the British Journal of Psychiatry in 2013. They pooled 31,424 people. Those with depression had lower vitamin D than controls, a standardized difference of 0.60. In the cross-sectional studies, the lowest vitamin D category had an odds ratio of 1.31 for depression against the highest (95% CI 1.0 to 1.71). The interval reaches 1. In the cohorts that followed people forward in time, the hazard ratio for later depression was 2.21 (95% CI 1.40 to 3.49). Both numbers are associations, with a direction and a size. Neither one, alone, proves that a capsule will close the gap. The paper said as much. The trials that could test a cause had not yet been run at the scale the claim needed.

What other countries measured

Denmark measured first, from a heel-prick. McGrath, Eyles, and colleagues, in Archives of General Psychiatry in 2010, used neonatal dried blood from 424 people later diagnosed with schizophrenia and 424 controls matched for sex and birth date. Compared with newborns in the 4th quintile, about 16 to 20 ng/mL, those in each of the 3 lower quintiles had about twice the risk of schizophrenia. The lowest quintile carried a relative risk of 2.1. The highest quintile, above about 20 ng/mL, carried a relative risk of 1.71. The lower bands and the top of that newborn sample both marked extra risk. The top band is not a toxic adult dose. The authors estimated a population-attributable fraction of 44% inside that Danish set if every newborn had sat in the 4th quintile. The percentage belongs to one study. It does not say vitamin D accounts for nearly half of schizophrenia on earth. It remains the most unsettling figure in this file, because the blood was drawn before any psychiatrist had written a word.

Norway studied adults who were overweight or obese, and who had not been chosen for depression or for a low level. Jorde and colleagues, in the Journal of Internal Medicine in 2008, randomized them to 20,000 or 40,000 IU of vitamin D a week, or to placebo, for a year. Beck depression scores improved in the vitamin D arms and did not improve on placebo. Iran asked the question of people who already had a depressive illness and a low level. A randomized trial there by Mozaffari-Khosravi and colleagues, in 2013, found symptom scores improved more when the deficiency was treated than when it was left alone. Sweden looked at adolescents. Högberg and colleagues, in Acta Paediatrica in 2012, gave vitamin D to depressed teenagers who were deficient, and mood scores improved. The study was open-label. The teenagers knew they were being treated. That is a direction, not a verdict. Australia still produces deficiency in people who live indoors, cover the skin, or have darker pigment. The Brisbane group that mapped the receptor has spent 2 decades on vitamin D and the developing brain.

The United States then ran the trial that should humble a slogan. Okereke and colleagues, in JAMA in 2020, followed 18,353 men and women aged 50 and older inside the VITAL trial. None of them had clinical depression at the start. Half took 2,000 IU of vitamin D3 a day. Over a median of 5.3 years, the hazard ratio for depression was 0.97 compared with placebo. Mood scores did not differ. A pill given to older adults who were not chosen for deficiency did not prevent depression. The study is large, and it is not a reason to discard Anglin. It bounds prevention in people who were not selected for an empty level. It does not prove that filling an empty level fails, and it does not prove that filling one works. Smaller trials in deficient patients look more promising than VITAL-DEP. They are not a result of the same size, and the open Swedish series is not a blinded verdict.

Anxiety has a thinner file than depression. Studies that measure both at one sitting often report a lower level in anxious patients. Those papers do not show that raising the level treats anxiety. The randomized evidence is not yet strong enough to call vitamin D an anxiety medicine. Light is ahead of the capsule where sleep is concerned, and panic moves with sleep. A child or an adult who never gets a bright morning is running that clock with no result in hand.

The same gradient, a different illness

The racial pattern in the blood is blunt, and it matches the pattern this app has already laid out for asthma and for lupus. Darker skin makes less vitamin D from the same ultraviolet dose. At American latitudes, Black children and adults carry the deficiency more often. Mental-health surveys do not copy that ladder one for one, and this piece will not pretend they do. Past-year major depression in national surveys is not, in any simple way, highest in the group with the lowest vitamin D. Clinic reports from psychiatric wards in the United States and in Europe often find deficiency as a common result rather than a rare one. This article does not assign that ward pattern a national rate of its own. A skin tone does not cause the illness. The storage form, in the NHANES figures above, is already lower where the skin is darker, before the first night on a ward.

Children are not an afterthought to the adult trials. The Kumar figures are pediatric. The Danish heel-pricks are neonatal. The Swedish series is adolescent. A child at a northern latitude, indoors from dawn-dark to dusk-dark, with a deep pigment, can finish a winter without making a useful amount of the hormone, while the adults in the room argue about behavior. The test is a blood draw. The argument can wait.

What the medicines cost while the level stays low

None of this is a reason to throw a medicine out. Some of these drugs hold a life in place. It is a reason to know what they cost, and to notice that a low vitamin D is often still sitting there after the prescription is filled.

Antidepressants of the SSRI and SNRI kind are not, as a class, kidney poisons or liver poisons. Serious liver injury is uncommon. The ordinary costs sit elsewhere: sexual effects, a withdrawal that can be mistaken for a relapse, a drop in sodium in older adults, a small rise in bleeding risk, and, in children, adolescents, and young adults, the suicidal thinking and behavior that forced the boxed warning. They are often the right first drug. They do not measure vitamin D, and they do not open a curtain. About 13.2% of American adults had used an antidepressant in the past month in 2015 to 2018, in the CDC’s National Center for Health Statistics brief. The share is higher in women. A medicine that common should not be the only item in the workup.

Antipsychotics press harder on metabolism. Weight, glucose, and lipids are an endocrine account whether or not the chart uses that word. In the CATIE trial, Lieberman and colleagues, in the New England Journal of Medicine in 2005, put those costs in public. Olanzapine was the clearest weight-gain drug in that set. Risperidone and paliperidone raise prolactin, a pituitary signal. Over years that rise can thin bone, break up menstrual cycles, and lower sex hormones, the last thing a person short on vitamin D needs if fracture and mood are being judged together. Some of these drugs also sedate. The level can fall because a sedated person gets outside less. A chemical loss of vitamin D is shown for enzyme-inducing anticonvulsants, not for antipsychotics as a class.

Lithium carries the best-documented kidney harms in psychiatry, which is not the same as the mildest. It can stop the kidney from concentrating urine. That failure is nephrogenic diabetes insipidus, not a sugar diabetes. Over years, the filtration rate can fall. McKnight and colleagues, in The Lancet in 2012, also placed hypothyroidism, hyperparathyroidism, and weight gain on the list of harms with a real signal. The parathyroid point belongs here. Vitamin D, calcium, and parathyroid hormone are one conversation. A person on lithium whose calcium is drifting needs that conversation, not another mood scale by itself.

Valproate carries a boxed warning for liver failure, highest in young children and in people taking more than one anticonvulsant. It also touches reproductive endocrinology, including features of polycystic ovary syndrome in some women, and it is a teratogen. Carbamazepine, phenytoin, and phenobarbital induce the enzymes that break vitamin D down. Holick’s map in the New England Journal of Medicine in 2007 lists enzyme-inducing seizure drugs among the reasons a level collapses. Those drugs can lower 25-hydroxyvitamin D. The illness remains the illness. That is not an instruction to stop them.

People with serious mental illness are often more deficient than other people in the same city. Indoor days, diet, smoking, obesity, dark skin, a northern winter, and these drugs stack. Treating the deficiency does not retire the diagnosis. Where the trials began with a low level, mood scores moved more often than they did in VITAL, which did not choose people for a low level. Lower stores mark higher depression risk in the observational work. Anglin’s hazard ratio of 2.21 is the risk for the lowest category against the highest, not a prize for driving a level that is already enough. The studies do not say a trophy level, past repletion, keeps adding benefit. They say the empty state is the one that marks trouble.

Ask for the number

The number worth asking for is 25-hydroxyvitamin D, the storage form. The active hormone, 1,25-dihydroxyvitamin D, can look ordinary while the stores are low, because the kidney defends it. The Institute of Medicine drew the line for bone at 20 ng/mL in the general population. Many clinicians who treat deficiency aim higher, often toward 30 ng/mL, when a person is already ill. The Endocrine Society’s 2024 guideline advises against routine screening of healthy people who have no particular risk. It does not list depression as a reason to test by habit. A low level is still worth drawing when the history gives a reason: a depression that will not lift, a psychotic illness, a life spent indoors, dark skin at a latitude with a weak winter, clothing that covers the skin, obesity, malabsorption, or a drug that speeds vitamin D breakdown.

Correcting a low level is not a promise that the illness will stand up and leave. It is the correction of a deficit the studies keep finding in the same patients. Lower stores, in the observational work, including the pooled 31,424, mark higher odds of depression. A level past repletion is not the aim. Denmark found a U-shape for later schizophrenia inside one neonatal sample, and the top quintile was not a high-dose level. A result that pushes calcium too high is harm, not wellness. A physician reads the number. The safe sun this app already times is how the skin writes the molecule without a sale. Morning light in the eyes, never a stare at the disk, is the other half, and it does not wait on a laboratory. Children and adults can both be measured. The season is already in the blood. Someone has to look.

Sources

1. Anglin RE, Samaan Z, Walter SD, McDonald SD. Vitamin D deficiency and depression in adults: systematic review and meta-analysis. Br J Psychiatry. 2013;202:100–107. 31,424 people. SMD 0.60 (95% CI 0.23–0.97). Odds ratio 1.31 (95% CI 1.0–1.71). Hazard ratio 2.21 (95% CI 1.40–3.49). PubMed 23377209
2. Forrest KY, Stuhldreher WL. Prevalence and correlates of vitamin D deficiency in US adults. Nutr Res. 2011;31:48–54. NHANES 2005–2006, at or under 20 ng/mL.
3. Kumar J, Muntner P, Kaskel FJ, Hailpern SM, Melamed ML. Prevalence and associations of 25-hydroxyvitamin D deficiency in US children: NHANES 2001–2004. Pediatrics. 2009;124:e362–e370. Adjusted odds ratio 21.9 (95% CI 13.4–35.7) for under 15 ng/mL, non-Hispanic Black versus non-Hispanic White. About 59% of non-Hispanic Black girls aged 13 to 21, and 5% of non-Hispanic White girls the same age.
4. McGrath JJ, Eyles DW, Pedersen CB, et al. Neonatal vitamin D status and risk of schizophrenia. Arch Gen Psychiatry. 2010;67:889–894. 424 cases, 424 controls. 4th quintile, 40.5 to 50.9 nmol/L, was the reference. Lowest quintile relative risk 2.1. Highest quintile, above 51 nmol/L, relative risk 1.71. Population-attributable fraction 44% inside that sample. PubMed 20819982
5. Okereke OI, Reynolds CF, Mischoulon D, et al. Effect of long-term vitamin D3 supplementation vs placebo on risk of depression. JAMA. 2020;324:471–480. VITAL-DEP. 18,353 adults. Hazard ratio 0.97. PubMed 32749491
6. Jorde R, Sneve M, Figenschau Y, Svartberg J, Waterloo K. Effects of vitamin D supplementation on symptoms of depression in overweight and obese subjects: randomized double blind trial. J Intern Med. 2008;264:599–609. Not selected for depression or deficiency. Beck scores improved on vitamin D, not on placebo.
7. Mozaffari-Khosravi H, Nabizade L, Yassini-Ardakani SM, Hadinedoushan H, Barzegar K. The effect of 2 different single injections of high dose of vitamin D on improving the depression in depressed patients with vitamin D deficiency. J Clin Psychopharmacol. 2013;33:378–385.
8. Högberg G, Gustafsson SA, Hällström T, Gustafsson T, Klawitter B, Petersson M. Depressed adolescents in a case-series were deficient in vitamin D and depression was ameliorated by vitamin D supplementation. Acta Paediatr. 2012;101:779–783. Open series, not a blinded trial.
9. Eyles DW, Smith S, Kinobe R, Hewison M, McGrath JJ. Distribution of the vitamin D receptor and 1α-hydroxylase in human brain. J Chem Neuroanat. 2005;29:21–30.
10. Patrick RP, Ames BN. Vitamin D hormone regulates serotonin synthesis. Part 1: relevance for autism. FASEB J. 2014;28:2398–2413. A mechanism paper, not a clinical trial.
11. Wright KP, McHill AW, Birks BR, Griffin BR, Rusterholz T, Chinoy ED. Entrainment of the human circadian clock to the natural light-dark cycle. Curr Biol. 2013;23:1554–1558.
12. Golden RN, Gaynes BN, Ekstrom RD, et al. The efficacy of light therapy in the treatment of mood disorders: a review and meta-analysis. Am J Psychiatry. 2005;162:656–662.
13. Lam RW, Levitt AJ, Levitan RD, et al. Efficacy of bright light treatment, fluoxetine, and the combination in patients with nonseasonal major depressive disorder. JAMA Psychiatry. 2016;73:56–63. Light alone and light plus fluoxetine beat placebo on the MADRS. Fluoxetine alone did not.
14. Holick MF. Vitamin D deficiency. N Engl J Med. 2007;357:266–281. Includes drug-induced loss of vitamin D.
15. Lieberman JA, Stroup TS, McEvoy JP, et al. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. N Engl J Med. 2005;353:1209–1223. The CATIE trial.
16. McKnight RF, Adida M, Budge K, Stockton S, Goodwin GM, Geddes JR. Lithium toxicity profile: a systematic review and meta-analysis. Lancet. 2012;379:721–728.
17. Turner EH, Matthews AM, Linardatos E, Tell RA, Rosenthal R. Selective publication of antidepressant trials and its influence on apparent efficacy. N Engl J Med. 2008;358:252–260.
18. Brody DJ, Gu Q. Antidepressant use among adults: United States, 2015–2018. NCHS Data Brief, no. 377. Hyattsville, MD: National Center for Health Statistics. 2020. 13.2% of adults.
19. OpenSecrets, from U.S. Senate lobbying disclosures. Pharmaceuticals/Health Products, 457.3 million dollars reported for 2025. PhRMA, 38.2 million dollars in 2025. OpenSecrets
20. Institute of Medicine. Dietary Reference Intakes for Calcium and Vitamin D. Washington, DC: National Academies Press; 2011. The 20 ng/mL line for bone.
21. Demay MB, Pittas AG, Bikle DD, et al. Vitamin D for the prevention of disease: an Endocrine Society guideline. J Clin Endocrinol Metab. 2024. Routine screening of healthy adults is not recommended.
22. FDA boxed warnings. Valproate and hepatic failure. Antidepressants and suicidality in children, adolescents, and young adults.