A new vitamin D study has completely changed my view and the advice I give to my patients in the clinic. Because it takes a novel approach, it uncovered something we may have been missing to this point.
Table of Contents
The Background
Not so long ago, researchers were alarmed about vitamin D. Specifically, they were worried there was an epidemic of deficiency that may be contributing to a host of serious diseases.
Early observational studies uncovered concerning linkages everywhere. When we tested blood levels of vitamin D and compared these to health outcomes, it didn’t look good.
The results showed an association between low vitamin D and cancer, infectious diseases, autoimmune conditions, diabetes, and heart disease [1].
So suddenly, there was an explosion of interest in vitamin D. Everyone was getting tested, and supplement use went through the roof. Some influential guidelines in 2012 from the Endocrine Society helped to fuel the trend.
But then the big clinical trials came out, like the VITAL trial. Twenty-five thousand adults were given vitamin D or a placebo and followed for 5 years. And the dramatic benefits people hoped for simply were not there. There was no clear reduction in heart attacks or strokes. Vitamin D didn’t seem to help much with cancer, either [2].
The VITAL trial was a nationwide, randomized, placebo-controlled, 2x2 factorial trial of daily vitamin D3 (2000 IU) and marine omega-3 fatty acids (1 g) in the primary prevention of cancer and cardiovascular disease among 25,871 U.S. men aged ≥50 and women aged ≥55, including 5,106 African Americans. Median treatment duration was 5.3 years. Vitamin D did not significantly reduce the primary endpoint of total invasive cancer incidence. Nor did it significantly reduce the co-primary endpoint of major cardiovascular events.
Guidelines soon caught up with the data. The latest Endocrine Society guidelines on vitamin D looked at all these trials and brought things back down to earth. For the general adult population, there was no clear evidence that supplementing with vitamin D improved health outcomes. So they concluded we should just stick with the standard recommended daily allowance of 800 IU. No need for aggressive dosing [3].
And since studies haven’t given us a clear target for an optimal vitamin D level in the blood, they also recommended against routine testing to check our levels [3].
In other words, stop chasing the number if there isn’t a clear reason to be measuring it.
As an evidence-based clinician, I took that seriously. I haven’t ordered vitamin D tests on my otherwise healthy patients. If someone is older, housebound, or clearly at high risk of deficiency, I have simply recommended a moderate daily dose of vitamin D rather than send them for a blood test that they would have to pay for themselves, and up until this point has not been shown to improve anything.
The New Study
But now I’m not so sure.
Last week, I was scrolling through the abstracts from a cardiology conference when a title caught my eye. It was for a randomized clinical trial — called the TARGET-D trial — involving vitamin D and serious heart-related problems like heart attacks and strokes [4].
It involved a very high-risk group.
I clicked, half expecting to add another “negative vitamin D trial” to the long list.
Instead, I found something that threatened to upset the simple story.
Take one of my patients for example, who let me share his story. I’ve switched his name. Let’s call him John. John is sixty-two, and he hasn’t paid much attention to his health so far. He ignores his chest discomfort for a day because work is busy and he does not want to make a fuss. When he finally goes to the hospital, his EKG is not subtle. He has had a significant heart attack.
The cardiology team performs an angiogram and places a stent in the blocked artery to open it. He starts the usual medications, and after a few days, he’s stable enough to go home. He is grateful, and a little shell-shocked, promising everyone that he will eat better, exercise, and take every tablet exactly as prescribed.
In the TARGET-D trial, people like John had one more decision point before they left the hospital. Everyone received usual cardiac care. But half of them were also randomized to a vitamin D management strategy. They had their blood level of vitamin D measured. Most had what we would call “low” levels, with an average around twenty-seven ng/mL [4].
In the vitamin D management group, the clinicians then used an algorithm to prescribe a vitamin D supplement and adjust the dose over time. Many of these patients started on five thousand units per day. The goal was to get that blood level above forty and keep it there [4].
Patients returned for repeat measurements and dose adjustments.
The other half of the trial, the usual care group, did not have that structured protocol. They just received normal care for their circumstances.
Researchers then followed everyone for an average of just over four years and watched for serious cardiovascular events. These include death, another heart attack, hospitalization for heart failure, or stroke. That combined outcome is what we call MACE — major adverse cardiovascular events [4].
So what happened to people like our imaginary John?
On the surface, not a lot.
When researchers looked at that combined MACE outcome, there was no statistically significant difference between the groups. Roughly eighteen out of one hundred people in the usual care group had one of those events during follow-up. This compares with about sixteen out of one hundred in the vitamin D management group [4].
The hazard ratio was about 0.85, which indicates a 15% lower risk. But the confidence interval was wide and comfortably crossed one. In plain language, the difference here might just be due to chance [4].
If this were the only information I saw, I would shrug and file TARGET-D next to VITAL in the “probably negative” drawer.
But when we look at the components of that composite outcome, something stands out. The risk of having another heart attack was lower in the vitamin D group. In the usual care group, about eight percent of patients had another heart attack during the study. In the vitamin D group, it was closer to four percent [4].
That translates to an impressive risk reduction of 52% [4].
Put differently, for every hundred people like John, there were about four fewer heart attacks over those four years in the vitamin D management group.
That is not a guarantee. It is still a small absolute difference. The confidence intervals are wide, which means we have a lower certainty about the actual effect size. But this is not nothing.
The researchers also did what’s called a per-protocol analysis. Instead of counting everyone included in the trial, they asked a different question. Among those in the vitamin D arm who actually got their levels above forty and kept them there, how did outcomes compare with those in the usual care arm?
In that analysis, the numbers look even more impressive, at least for the composite MACE outcome and deaths. The risk reductions are greater here in the “achieved the target” group.
And this is where it would be very easy to get carried away.
There’s a potential story here that feels compelling. Because there’s something novel about the study design. It’s the first time we’ve checked for the effects of vitamin D supplements by using customized dosing to raise people to a target level, and followed them up over such a long period of time to see what happens to MACE [4].
Past studies have always just given the same supplement to everyone involved. And they’ve also often included lots of people whose vitamin D levels weren’t particularly low.
So maybe the reason the clinical trials haven’t found benefits is because we haven’t been looking at what happens when we go from low vitamin D levels to adequate ones.
So here’s how the plausible-sounding story goes:
“We finally did the right trial. We treated vitamin D properly, to a target level rather than with a random blanket dose, and the heart attacks melted away. The sceptics were wrong. Everyone after a heart attack should have their vitamin D level tested and treated aggressively.”
I won’t be surprised if I see some health influencer saying this.
Cautions
But we need to pause here for a moment because there are some big caveats.
TARGET-D is randomized, which is good. But it does not appear to be blinded. Patients and clinicians knew who was in the vitamin D management group and who was not. And this means there is a serious danger of bias in the results.
For one thing, there’s the potential for a placebo effect with those taking vitamin D. Let me explain. When they know they are on an “active” protocol, that might nudge them to be more engaged with their care. Doctors may be more diligent with other risk factor management. Nurses may call them back more promptly. No one means to be acting in a way that would skew results. It just happens. And that’s why we use blinding in studies.
A more robust study design may have been for both groups to get regular blood tests, and a third party manages the vitamin D dosing. This way, the patients and the study nurses wouldn’t know who is in which group, what the vitamin D levels are, or what dose of vitamin D is being taken. That design would have significantly reduced potential bias.
It’s also worth noticing that the primary outcome of the trial, the composite MACE metric, didn’t show a statistically significant result. The reduction in heart attacks, where we do see a benefit, is a component of that composite. Once you start looking at multiple outcomes, though, the chance that one of them reaches statistical significance by luck alone goes up.
When it comes to the per-protocol analysis, where only people who actually reached the target level are counted as “treated,” we need to be even more cautious. Even though there was a trend toward improvement in some areas, as we saw, the results here didn’t reach statistical significance. So treatment with vitamin D essentially failed the sensitivity analysis. In other words, when researchers tried to narrow their analysis to just those cases where we’d expect to see the strongest results if vitamin D is actually effective, the picture was totally unclear. And the numbers actually looked a bit worse when it comes to heart attacks.
We also have to notice the uncomfortable parts of the forest plot. The hazard ratios for heart failure hospitalization and for stroke drift higher in the vitamin D group [4].
Those findings are not statistically significant, and the confidence intervals are very wide. They may well be noise. But they are a reminder that biology is messy. Pushing people to higher blood levels of a hormone is not guaranteed to be free of trade-offs.
Takeaways
So what are the implications here?
TARGET-D does not give us the clean, definitive answers we might like. There’s nothing here that is going to overthrow the current recommendations from the Endocrine Society. It gives us some tantalizing evidence that we could see heart-related benefits in at least some populations by trying to reach target vitamin D levels. But we need better-designed follow-up studies that can address the limitations we’ve looked at. And it would be interesting to see the same treat-to-target approach taken with a general population.
But the trial results do hint that the reality might be more complicated than the simple story from the current Endocrine Society guidelines about vitamin D that I explained earlier.
So here’s how my approach is changing as a result.
Let us go back to that hypothetical patient, John. With the TARGET-D trial in mind, I’m going to be having a slightly different conversation with him. We’ll talk about the normal stuff — like lifestyle changes, statins, and other medications. But I’ll also describe the trial results and say something like this:
“The vitamin D research is messy and there isn’t a clear answer. The new TARGET-D study offers us some justification to go ahead with a vitamin D blood test, but more research is required. If it’s a financial option for you, and you’re comfortable with the uncertainty of the research, go ahead. Equally, it’s not essential and we can stick with the Endocrine Society’s current guidelines.”
I wouldn’t have brought that up before.
And personally, I’m considering doing something I haven’t bothered with until now: testing my own vitamin D level. Again, this new study doesn’t give us definitive evidence. But out of curiosity, I want to see where my level stands. If it’s clearly low, I will consider boosting my vitamin D intake. I currently take 1,000 IU as part of MicroVitamin. But just because I take a supplement, that doesn’t mean you also have to.
