Very low vitamin D was linked to more age-related disease in two huge health-record studies

Vitamin D is best known for keeping bones strong, and low levels have been linked to many chronic diseases. Whether low vitamin D causes poor health, or is simply a sign of people who are already ill or spend little time outside, has been argued about for years. A new study from Israel adds very large numbers to the debate: in routine medical records from two countries, severe deficiency went together with higher risks of death and of several diseases of aging. Pharmacy purchases of vitamin D went together with lower risks. But this is an observational study, so it shows links, not cause and effect.

The body makes vitamin D in the skin when sunlight hits it, and gets some from food and supplements. Doctors measure it in the blood as 25-hydroxyvitamin D. In this study, below 10 nanograms per millilitre counted as severe deficiency and above 30 as sufficient.

What they did

The team, from Tel Aviv University and the Israeli health organization Leumit Health Services, used anonymized electronic health records in four steps:

• How common is deficiency? They looked at more than 1.67 million vitamin D tests from over 468,500 Israeli adults between 2009 and 2020. Levels rise and fall with the seasons, lowest in mid-February, so each result was converted to its expected winter value.
• A matched comparison. 12,352 adults with severe deficiency were each paired with one adult with sufficient vitamin D of the same age group, sex, smoking status, weight category, blood-pressure category, community and income level. The pairs were followed for up to 10 years. To limit the chance that existing illness explained the results, people who already had a disease, or developed it within six months, were left out of that disease's analysis.
• A check in the US. The same comparison was repeated in 223,175 matched pairs from TriNetX, a network that pools records from over 70 US health organizations. More than half of these people were Black.
• Supplements. In 247,848 Israeli adults, the team tracked vitamin D products bought at the health organization's pharmacies over time, estimated how much each purchase would raise blood levels, and asked whether more of it went along with fewer deaths and diseases.

What they found

• Deficiency is common. About 20% of women and 10% of men in the Israeli records had severe deficiency in winter terms.
• Death. In the matched Israeli groups, about 14% of the severely deficient had died after 8 years, against about 8% of those with enough vitamin D: a risk about 2.6 times higher after adjustment. In the US records, the risk was about 1.6 times higher.
• Diseases. In both countries, severe deficiency was linked to higher risks of toe or foot amputation, heart attack, stroke, dementia, diabetic eye damage and starting dialysis. The links were smaller in the US data for most of these (for dementia, 1.14 times the risk against 1.35 in Israel).
• Skin cancer went the other way. People with very low vitamin D were less likely to get skin cancer, probably because they spent less time in the sun.
• Supplements. People whose pharmacy purchases were estimated to raise their blood level by more than 10 units had about 28% lower risk of death than people who bought little or none, with smaller reductions at lower amounts. Supplement purchases were also linked to fewer cases of diabetes and heart attack, but not to fewer amputations, which the authors put down to small numbers. A second, more elaborate statistical method gave similar results for death.

What to keep in mind

• Links, not proof. The authors themselves write that residual confounding "cannot be entirely excluded": people with very low vitamin D may differ in many ways the records do not capture, such as how much they go outdoors, how active they are, or illnesses not yet diagnosed. And people who buy supplements may have other healthy habits.
• Purchases are not pills taken. The supplement analysis counts what was bought at the health organization's own pharmacies. It cannot know whether people took it, and misses vitamin D bought elsewhere.
• Trials disagree. Randomized trials of vitamin D, which can test cause and effect, have given inconsistent results for chronic disease and death. The authors argue those trials mostly enrolled people who were not deficient; that is their explanation, not something this study tested.
• The numbers differ between countries. The higher risk of death was much larger in the Israeli data than in the US data, which shows how much such estimates depend on who is studied and how.
• Medical records are collected for care, not research: diagnoses can be missing or coded inconsistently.

Why it matters

Severe vitamin D deficiency was common in these records, and in two very different populations it went together with a broad range of age-related problems, from heart attacks to dementia. The authors see it as a possibly modifiable sign of vulnerability with age and estimate large benefits from correcting it, but those estimates assume that the links are causal, which this kind of study cannot establish. Whether correcting severe deficiency in older people prevents disease is a question for randomized trials that focus on people who really are deficient. Anyone wondering about their own vitamin D level should ask a doctor rather than draw conclusions from this study.