← Longevity & Supplement Guides
Vitamin D and Belly Fat: The Headline Had the Arrow Backwards
Somewhere in the last ten days, a study about a tape measure and a blood test turned into a story about a bottle. That is not the researchers' fault. It is simply what happens when a finding arrives with two variables in it and only one of them has something to sell.
The week the headline wrote itself
The paper is real and reasonably careful. Joyce da Silva Milliati and colleagues, writing in Diabetes, Obesity and Metabolism, went into the English Longitudinal Study of Ageing — a long-running cohort of older adults in England — and pulled 5,520 people aged 50 and over with both a waist measurement and a vitamin D blood test on file. Then they watched six years go by.
People with abdominal obesity alone had a reported 47% higher risk of dying in that window. People with vitamin D deficiency alone, up to 91% higher. People with both, 123% higher — more than double. That last number went around the world in about a day, and by the time it reached the general-interest press the practical lesson had quietly become: go and check your vitamin D.
It is a good study and a bad instruction. The reason is worth twelve hundred words, because it is the single most useful habit anyone can develop when reading health news: before you act on a correlation, ask which way the arrow is pointing.
What the paper actually measured
Abdominal obesity here means a waist over 102 cm in men and over 88 cm in women — the long-standing clinical thresholds. Vitamin D deficiency means a serum 25-hydroxyvitamin D below 30 nmol/L, about 12 ng/mL, which is genuinely low rather than merely unfashionable.
So the finding is this: among older English adults, those two markers travel together, and the pair marks a steeper mortality curve over six years than either marker alone. That is a real observation about how risk clusters in a population. It tells you something about who is in trouble.
What it cannot tell you is what happens if you change either number. Nobody was assigned anything. Nobody's waist or blood level was moved by the researchers and then followed to see what happened next. A cohort study describes the traffic; it does not test the steering.
The arrow probably runs the other way
Here is the part the headlines skipped. In the relationship between body size and vitamin D, the causal direction has already been investigated with a method designed for exactly this problem — and it does not favour the intuitive reading.
Karani Vimaleswaran and colleagues published a bi-directional Mendelian randomisation analysis in PLOS Medicine in 2013, pooling up to 42,024 adults across 21 cohorts. Using genetic variants as stand-ins for lifelong exposure, they found that a higher body mass index drives blood vitamin D down, while genetically lower vitamin D showed no meaningful effect on body mass index. One direction had evidence; the other did not.
The mechanism is almost boringly physical. Andjela Drincic and colleagues had argued in Obesity a year earlier that the low readings of larger bodies are best explained by volumetric dilution: the same quantity of a fat-soluble vitamin distributed through a bigger total volume of tissue simply reads lower in the blood. Not trapped, not destroyed. Diluted.
Dosing behaves accordingly. A 2023 analysis of the VITAL trial in JAMA Network Open found that an identical 2,000 IU daily dose raised circulating levels less in participants with higher body weight. The pill is the same size; the body it has to fill is not.
Put those three together and a low reading in a person with a large waist is, in part, a measurement of the waist. Two of the study's variables are not fully independent. They never were.
Figure 1
One low reading, three explanations that all fit
The straightforward reading
Little sun exposure, a northern winter, dark skin at high latitude, very little in the diet. Intake and synthesis genuinely fall short, and the blood level reports that honestly.
This is the case public-health guidance is written for: UK advice suggests 10 micrograms (400 IU) daily through autumn and winter.
The dilution reading
The same amount of a fat-soluble vitamin, spread through a larger total body volume, produces a lower concentration in a blood sample. Genetic evidence points from body size to blood level, not the other way round.
Drincic, Obesity 2012 (volumetric dilution); Vimaleswaran, PLOS Medicine 2013 (bi-directional Mendelian randomisation, up to 42,024 adults).
The bystander reading
Chronic illness, inflammation, and the housebound months that come with both push the number down. Here the low reading is a symptom of the situation rather than its origin.
Autier and colleagues, The Lancet Diabetes & Endocrinology 2014, reviewing the observational and trial literature together.
A map of explanations, not a diagnosis. All three can be true at once in the same person, which is precisely why a single blood value cannot tell you which lever to pull. Interpretation of a lab result belongs with your doctor.
Low vitamin D is also an excellent marker of a life going badly
In 2014, Philippe Autier and colleagues reviewed the vitamin D literature in The Lancet Diabetes & Endocrinology and reached a conclusion that has aged well and been quietly resented ever since: across a wide range of conditions, low 25-hydroxyvitamin D looked more like a consequence of poor health than a cause of it. The inflammation that accompanies chronic illness depresses the number. So does spending your months indoors.
Now consider the shape of the ELSA analysis again. Six years of follow-up, in adults over 50, with a baseline blood draw. That is the exact window in which illness already present but not yet named at baseline surfaces as a death at the end. Statistical adjustment helps with the confounders you can name and measure. It cannot fix an arrow.
None of this makes the finding useless. A person with a large waist and a low vitamin D level is, empirically, in a higher-risk group — that is genuinely worth knowing. It just isn't the same fact as "the deficiency is what is doing the damage."
What happened when they randomised the pill
We are not guessing about this. Vitamin D is one of the most heavily trialled supplements in existence, and two of the largest trials tested precisely the outcome the headlines gestured at.
VITAL randomised 25,871 US adults to 2,000 IU of vitamin D3 daily or placebo and followed them for a median of 5.3 years. Invasive cancer: hazard ratio 0.96 (95% CI 0.88–1.06). Major cardiovascular events: 0.97 (0.85–1.12). Both intervals sit across the line of no difference.
D-Health randomised 21,315 older Australians to 60,000 IU monthly or placebo for five years and looked directly at dying. All-cause mortality: hazard ratio 1.04 (95% CI 0.93–1.18). Again, nothing.
Figure 2
The same question, asked with randomisation instead of observation
Open markers with whiskers spanning 1.0: three large randomised trials that found no difference. Positions and interval widths are the published values. Both trials enrolled populations that were largely replete to begin with — which is the honest limit of what they settle. Sources: Manson, NEJM 2019 (VITAL, n=25,871); Neale, The Lancet Diabetes & Endocrinology 2022 (D-Health, n=21,315).
The honest caveat matters more than the null results. Neither trial screened for deficiency at entry; both largely enrolled people who already had adequate levels. So they answer "does adding vitamin D to a replete population change these outcomes" — no — and they leave "does correcting genuine deficiency change them" much less settled. That question is still open, and anyone who tells you otherwise in either direction is overselling.
One VITAL sub-analysis, published in JAMA Network Open in 2020, is worth holding onto: the signal for advanced cancer was concentrated in participants of normal body mass index and absent at higher weights. Whatever you make of a secondary finding — and secondary findings deserve suspicion — it points in the same direction as the dosing data. Body size keeps showing up as the thing that changes the answer.
Which of the two numbers you can actually move
Both variables in the original study are associations. Only one of them is attached to something you can do on a Tuesday.
Ahmad Jayedi and colleagues pooled 72 prospective cohorts in The BMJ in 2020 and found that each 10 cm of waist circumference carried a hazard ratio of 1.11 (95% CI 1.08–1.13) for all-cause mortality — and that this held independently of overall body mass index. Central measurements carry information that weight alone does not. That is still observational; it is not a promise that shrinking a tape measurement rewrites a hazard ratio. But it is the variable with behaviour on the other end of it, rather than a lab value that partly reports your body's volume back to you.
So here is the one thing to take from a week of headlines. Measure your waist properly, once, and write it down. Tape at the midpoint between the lowest rib and the top of the hip bone, against the skin, snug without compressing, at the end of a normal breath out, at the same time of day. The crude screen most guidance uses is waist divided by height, with 0.5 as the line.
Then leave it alone for a month. A number that moves slowly deserves a slow cadence — measuring it daily produces noise and anxiety, and nothing else. This is the same discipline that makes a walking-speed check or an omega-3 index useful rather than merely interesting: pick the interval that matches how fast the thing actually changes. Use numbers to correct fantasy, not to replace experience.
What vitamin D is genuinely for
None of this is an argument against the nutrient. It is an argument against asking it to do a job that belongs to a tape measure.
Vitamin D contributes to the normal absorption and utilisation of calcium and phosphorus, to the maintenance of normal bones, teeth and muscle function, and to the normal function of the immune system. Those are the authorised roles in the EU, and they are not small ones — muscle and bone are the tissues that decide whether the last decade of a life is spent upright. We wrote about the loading side of that story in what actually builds bone, and about the pairing question in vitamin D3 and K2 together.
Northern winters are a genuine gap: at high latitude, skin synthesis effectively stops for months, and public-health guidance in the UK suggests 10 micrograms — 400 IU — a day through autumn and winter for adults. That is modest, cheap, and well-founded. It is also a completely different claim from the one the headlines implied.
The affluent world would rather discuss a molecule than a waistband, and the supplement industry is delighted to oblige. A nutrient that does four ordinary jobs reliably does not need to be sold as an antidote to a body.
The bottom line
A large English cohort found that a big waist and a low vitamin D level together mark a steeper six-year mortality curve than either alone. Believe it: as a description of who is at risk, it is sound. Just don't read it as an instruction, because the evidence on direction points the other way — bigger bodies produce lower readings, illness lowers them further, and the large randomised trials of supplementing a replete population found nothing on cancer, cardiovascular events, or death.
Of the two numbers in that headline, one is largely a report on the other. Measure the waist. Take vitamin D for what it is actually good at, especially in winter. And regard any study that arrives with a bottle attached to it as a study that has been read too quickly.
