Junkfood Science: Science says...

May 04, 2007

Science says...

“Science says....” one thing today and something entirely different the next. Or so it seems.

We read in the New York Times on Tuesday that too little vitamin D may cause weakness and disability in the elderly. Yet, All Headline News reported that research “found that elderly people who consume high levels of calcium and vitamin D have a greater risk of harming their brains through larger brain lesions. They are also more likely to suffer from cognitive impairment, depression or stroke.”

The same day, news reported on a study finding that, despite everything we hear, “calcium and vitamin D supplements given to elderly...does not appear to reduce their risk of bone fractures.” While just last week, the Globe said that vitamin D has been linked to preventing cancer and that a vitamin D study will be coming out soon showing results that “are nothing short of astounding.”

And on and on. Good heavens. That’s just the past week’s news on a single vitamin, let alone the multitude of other things we eat. But it provides an opportunity to sort out a few basic scientific principles to help guide us in what to consider when we hear all of these things.

Basically, like so many of the popular dietary cures or demons of the day, the claims surrounding vitamin D’s role in cancer and health come mostly from epidemiological observational studies — looking for correlations among a group of people — and speculations on possible mechanisms to explain them as seen in the laboratory. There is a long history of people jumping to endorse a vitamin, food, ingredient or lifestyle based on these types of studies, only to discover later that carefully-conducted clinical trials completely disprove such beliefs. Remember beta-carotene was thought to prevent cancers yet randomized trials found it increased the risks instead? Then, there was vitamin E long touted for its cancer and heart disease prevention, only to be debunked in multiple randomized clinical trials. Low-fat went the same way. And on and on. The sobering reality is that correlations found in such observational studies, no matter how solid, strong and consistent, simply cannot substitute for actual clinical studies on real people because those correlations do not mean they are the cause. There could be countless other explanations. Most often, they are simply markers for social and economic factors that affect health.

Let’s take those studies on the elderly, as an example. The study reported by the New York Times had been published in the Journals of Gerontology Series. The researchers culled through a database called the InCHIANTI (aging in the Chianti area) survey of seniors living in two towns of Tuscany, Italy. The dietary information in the InCHIANTI database is from food-frequency questionnaires on food consumptions during the previous year. These researchers were looking for associations between physical performance and handgrip strength and vitamin D levels. They reported an inverse relationship between poor physical performance and strength and vitamin D status.

This same InCHIANTI database has been used for countless other studies which have pulled out all sorts of correlations. Another published last summer in the American Journal of Clinical Nutrition said that low magnesium levels were correlated with weaker muscle strength and performance among the elderly. Another before that published in the Journal of Gerontology Series said that low hemoglobin levels and anemia were associated with decreased muscular strength in the elderly.

You’re getting the picture. Besides the ‘seek and ye shall find’ phenomenon with these sorts of studies (which statistically-derive correlations from a huge database), by jumping to believe each one provides the answer almost always leads us to miss the real reason. Of course, those with an interest in selling vitamin D, magnesium or iron supplements or “healthy” food might be eager to use these types of studies to get us to believe that popping a pill or eating more or less of a certain food will make us stronger or help us live longer! But these studies never warrant changing the way we eat, no matter how convincing they might seem.

Here’s one illustration of how they can steer us in the wrong direction.

A closer examination of the InCHIANTI survey, reveals that individual nutrients are merely tag-along associations with what was really occurring among the elderly in this population. A study, led by Dr. Benedretta Bartali at the National Institute of Research and Care on Aging in Florence, Italy, provides a much more complete picture. As it turns out, those other correlations appear most explained by advancing age, and difficulties among the very oldest >85 year-olds in shopping, cooking, chewing and having the dexterity needed to cook (grasp handles, use a knife, etc.). As a result, those over age 85 were not eating as much — only 67% of the calories as those at age 65 — and weren’t eating as much variety because they’d resorted to things easier for them to get and eat, namely soups, and less of virtually everything else. So, they were short on calories and almost all macro- and micronutrients. They were hungry! These researchers noted that more provisions to assist elderly and more processed and prepared foods could be the greatest help in reducing the percentage suffering from poor nutrition.

Not that a special vitamin might prevent muscular weakness.

And what about those cancer-preventive claims in last week’s Globe story? Reporting on “astounding findings” from a not-yet published study is marketing, not science, and we’ll just have to wait and see what their next study actually found. Past studies by the same Harvard researchers have consistently concluded that vitamin D has “potent anticancer properties” and that higher and higher amounts appear needed to reduce fractures, prevent prostate cancer, and reduce overall cancers and mortality.

We can’t rush to presume that the reason these researchers advocate for vitamin D supplementation well above the RDA recommendations has anything to do with the fact that one of them is a paid consultant for DiaSorin, a company which describes itself as “a worldwide leader in Osteoporosis and Calcium Regulation diagnostic market, DiaSorin offers a full line of products for the management of Bone and Calcium-related diseases, including advanced diagnostic tools, such as 25-OH Vitamin D and 1,25(OH)2 Vitamin D assays.”

Let’s look at their studies. Each has sorted through databases for associations, rather than being clinical trials examining actual long-term health outcomes.

The study in the Journal of the National Cancer Institute, for example, culled through the Health Professionals Follow-up Study, a database of questionnaires gathered from 51,529 male healthcare professionals who had signed up in 1986. For this study, the Harvard researchers used the food-frequency questionnaires — where the men recorded how often they’d eaten 130 different food and beverages over the past year — to estimate vitamin D intakes, and the 25-OH Vitamin D levels that had been done on 1,095 of the men. They did not assess any of the men’s actual exposure to sunlight but, instead, used the average UV-B radiation in the geographic region where they lived based on the address they resided in 1986.

[Note: Vitamin D is actually from D3 (cholecalciferol) that’s produced in the skin when it’s exposed to UV-B radiation and D2 (ergocalciferol) which comes from our diet. Both of these precursors are hydroxylated in our livers to form 25-OH Vitamin D (the nonactive form stored in our body) and then converted by our kidneys to 1,25(OH)2 Vitamin D (the biologically active form). Vitamin D deficiency can result from inadequate intake, reduced absorption, abnormal metabolism, or resistance to the effects of vitamin D. There's more to it than just what we eat. People with kidney problems, for example, often develop rickets or osteomalacia because their kidneys aren’t producing enough of the active form (1,25(OH)2 Vitamin D).]

Using the 25-OH Vitamin D measurements on the 1,095 men, the researchers then:

quantified the relation of these six determinants to plasma 25-hydroxy-vitamin D [25(OH)D] level by use of a multiple linear regression model. We used results from the model to compute a predicted 25(OH)D level for each of 47 800 men in the cohort based on these characteristics. We then prospectively examined this variable in relation to cancer risk with multivariable Cox proportional hazards models.

Did you follow that? They concluded that “high levels of vitamin D may be associated with a 17% reduced cancer incidence and mortality in men.”

We’ve already learned that such minor associations are not even tenable for these types of studies. As the National Cancer Institute said: “In epidemiologic research, [increases in risk of less than 100 percent] are considered small and... may be due to chance, statistical bias, or the effects of confounding factors that are sometimes not evident.”

By giving the relative risk, the researchers also didn’t reveal the actual difference. It wouldn’t have been nearly so impressive sounding. The actual annual cancer deaths among the men at the bottom of the predicted 25(OH)D levels compared to the top levels was a difference of 0.084% (yes, a number that tiny).

So, this study was a nonfinding. By the way, body mass index and physical activity didn’t make any difference, either.

But even clinical trials can give us inconsistent findings and it can be hard to sort them out. Dr. Gerbrand Izakes at the University Medical Centre Groningen, The Netherlands, recently examined the randomized, controlled clinical trials on vitamin D supplementation in fracture prevention. Two early studies reported a significant decrease in fracture risk among the treated group, one trial found only a difference of borderline statistical significance; and eight found no positive effects at all, he said. He analyzed these differences and found that the only benefits from high doses of vitamin D depended on the people who had been studied, being only notable among institutionalized people suffering from deficiencies. A treatment beneficial for someone with a nutritional deficiency or specific medical condition is very different from a preventative for otherwise healthy people. He concluded that inconsistencies among recent trials are at least partially due to publication bias and differences in the target population. “High dose vitamin D may be effective in institutionalized persons but probably is not effective in the general population,” he said.

Consistently, the strongest nutritional science supports the prevention of actual deficiencies, but beyond that, claims of any food or ingredient as having magical health benefits or harmful qualities in the ranges typically eaten have yet to be supported in research we can bring to the table.

The bottom line is that by eating normally and enjoying a wide variety of everything and eating enough, most of us are likely to get the nutrients our bodies need. Endless fretting about every microgram not required.


© 2007 Sandy Szwarc

Bookmark and Share