July 17, 2009

Very Bad Science: Nitrates Cause Type 2 Diabetes

As my readers know, after reading a lot of research I've become convinced that toxic pollutants in our environment are a major explanation for the growth of obesity and the rise of Type 2 diagnoses over the past decades.

So you might expect when reports appeared all over the media recently, linking the increase in diabetes to nitrates in our food, I would have welcomed it as yet another factor pointing to an environmental cause for diabetes.

I might have, had I not read a detailed description of the research in question. Unfortunately, this study turns out to be yet another example of faith-based research--research that interprets correlations as causation. That is, researchers determine that two things happen at the same time and conclude one must have caused the other. As is often the case, this kind of conclusion is based not on hard scientific fact but on the fact that the correlation supports a a religious belief: "Meat and fast foods are evil."

Only the abstract of the study is available for free, here:

Epidemiological Trends Strongly Suggest Exposures as Etiologic Agents in the Pathogenesis of Sporadic Alzheimer's Disease, Diabetes Mellitus, and Non-Alcoholic Steatohepatitis . Journal of Alzheimer's Disease,Suzanne M. de la Monte et al., Volume 17, Number 3 / 2009. doi: 10.3233/JAD-2009-1070

But fortunately, a news report gives the salient details of this study. You can read it here:

Medical News Today: Researchers Find Possible Environmental Causes For Alzheimer's, Diabetes

To explain why this study is so flawed you need to know something about what has made other studies linking environmental substances to diabetes so compelling. In those studies, researchers take a group of people who developed diabetes and look for hard evidence that they have had more exposure to the possible environmental toxin than did a group of controls similar to these people who did not develop the condition.

That kind of well-designed study discovered elevated rates of diabetes in people in the Native American population of Upstate New York who had elevated concentrations of PCBs in their blood stream compared to peers who did not. Another such study linked the concentrations of arsenic found in NHANES subject's urine to the likelihood those specific people would develop diabetes over time.

The next step in linking an environmental toxin to a disease is to find a mechanism that explains why the toxin might cause the specific disease. Researchers did that when they did rodent experiments proving that the pesticide Atrazine, whose use is most prevalent in the regions in the US that have the highest levels of obesity, caused insulin resistance and mitochondiral dysfunction.

(You can find links to the research cited above and others pointing to toxins in the environment causing genetic damage leading to diabetes HERE.)

With that in mind, lets look at what the researchers did in this new nitrates study. It turns out that all they did was compare "age adjusted increases in death rate from Alzheimer's, Parkinson's, and diabetes and the progressive increases in human exposure to nitrates, nitrites and nitrosamines through processed and preserved foods as well as fertilizers" over the period from 1968 to 2005. Both grew.

As a control they compared the rise in nitrate use with the death rate from AIDS, Stroke, and Leukemia. These did not rise in tandem with increasing nitrate use.

QED? I don't think so. In fact, once again I am left wondering how someone trained in science could have considered this study worthy of publication.

For starters one gaping hole in the researchers logic is that they compared the death rate of one group of diseases which have no cure to the death rate of three diseases for which dramatic cure rates have occurred during the period from 1968 to 2005.

It takes about 25 seconds to realize that the death rate for AIDS, Stroke, and Leukemia have all plummeted since 1968 due to the advent of powerful and highly effective medications.

Leukemia was a death sentence in 1968 as anyone who ever wept their way through "Love Story" will remember.

AIDS was almost universally fatal when it first emerged in the early 1980s. Powerful drugs have turned it into a chronic condition.

The incidence of fatal stroke has decreased dramatically since the late 1960s due to the wide spread adoption of blood pressure screening and improvements in blood pressure medication.

Contrast the situation with the three conditions whose death rate the research would have you believe correlates with the rise in the use of nitrates. The researchers, rather surprisingly and, I would add, incorrectly, describe Alzheimers, Parkinsons, and Type 2 Diabetes thusly: "All of these diseases are associated with increased insulin resistance and DNA damage." What they neglect to mention is that not only are these three diseases still incurable, they have also undergone dramatic changes in how they are diagnosed, increases that mean many more people will have these conditions listed on their death certificates.

There is not a single drug for Alzheimers which delays its inevitable progression through deterioration to death. The drugs sold to the public with the claim they treat Alzheimers at best produce tiny "improvements" in functioning on specific cognitive tests which do not translate into changes observed in actual day to day function or behavior.

The researchers mention that the death rate from Alzheimers has skyrocketed over their period. But the explanation for this is two fold. First, improved health care means that people are living a lot longer--long enough to get Alzheimers which is relatively rare in people under the age of 80. Far fewer people made it into their mid-or late 80s in 1968. In addition, doctors did not diagnose "Alzheimers" in the 1960s for anything but unusual forms of dementia that occurred in relatively young people, aged 40-60. The use of the term "Alzheimers" for what used to be called "senile dementia" is relatively new.

Diabetes, as my readers know does have effective treatments. Unfortunately, they are not prescribed by the family doctors who treat patients. In fact, NHANES data found that the retreat from using insulin, which was more common in the 1960s to using heavily marketed oral drugs has led to a rise in the average A1c of the American diabetic over the 1990s, not a drop.

In addition, the diagnostic criteria for diagnosing Type 2 diabetes have undergone radical changes between 1968 and the present, to the point where comparing the death rate from diagnosed Type 2 in 1968 to the death rate from Diabetes diagnosed today is an apples-to-oranges situation. Most people who died from diabetes-related complications in the 1970s did so without ever getting a diabetes diagnosis.

So the fundamental methodology used in this study is deeply flawed. Beyond that, there is no attempt to go beyond vague hypothesis to explain the "correlation equals causation" conclusion here. Even if the correlation is real, which the study does not in any way prove, there are a lot of other things that have risen in tandem with the death rate from Type 2 diabetes: The number of bathrooms in the average house. The number of teams in Major League Baseball. The monthly premium for health insurance.

There are also some other far more scientifically valid environmental factors that have risen with the incidence of Type 2 diabetes--including the use of high fructose corn syrup in foods, the use of soy by products in foods, and the use of omega-6 vegetable oils in foods.

But to link anything in a causative way to a disease, you need to take that next step and find evidence that people who have the disease have more exposure to the supposed cause than other people otherwise similar to those people.

An elegant study recently showed much higher weight gain in a group of people drinking fructose-sweetened beverages to those drinking sugar sweetened beverages. You can read the details of that study HERE. This would take us a step closer to linking the invasion of our food supply by high fructose corn syrup to the advent of the obesity epidemic. Though because there are are so many different environmental correlations with the growth of both obesity and Type 2 diabetes, it is hard to assign the blame to any one.

But returning the the nitrate study, we see that its authors can point to no data showing that people who actually get Alzheimers or Type 2 diabetes consumed more nitrates than those who didn't. They don't show that people with high exposure to nitrates from fertilizer or industrial processes have more diabetes than people who did not have this exposure. No one has looked at the blood, urine, or tissues of people with Type 2 diabetes to see if they have higher levels of nitrosamines than people without the condition.

There is no proof that exposure to nitrates causes insulin resistance offered. And, for that matter, there isn't solid evidence showing a causative link between insulin resistance and Alzheimers and Type 2 Diabetes. Two thirds of the adult population is insulin resistant. Only about 10% ever develop diabetes.

The link is even more tenuous for Alzheimers. Fully one half of all people develop dementia if they live to be 80 or more including people who have normal insulin sensitivity. Since true Alzheimers can only be diagnosed on autopsy, many people whose cause of death is listed as "Alzheimers" actually have vascular dementia which is not clearly related to insulin resistance.

The scientific acceptance of this nitrate study--and the reason it got wide distribution in the media, is because the researchers linked nitrate exposure to eating processed foods and meat. In fact, among the factors they used to establish the rate of nitrate exposure was "annual sales at popular fast food chains, and sales for a major meat processing company."

So the religious belief: "meat is evil" and the sociological snob belief "people who eat fast food are fat slobs who cause their own diabetes" converge to give this woefully bad study credibility.

Mind you, it is possible that nitrates do cause diabetes, just as it is equally possible they don't. My point here is that this study doesn't begin to answer the question of whether there is such a relationship.

And its publication drops my respect for the people who peer review journal articles even further. They get paid huge amounts to provide their supposed scientific expertise. I do this for free. Yet over and over I am able to point out flaws in study design that should have been evident to anyone capable of logical thinking.

Something is wrong with this picture.



Khurt said...

My goodness. Science education in this country needs updating. That research was terrible.

Anna said...

Great post and excellent review of the bad science that floods the media and pollutes the public's perception of science.

But my jaw dropped at "they get paid huge amounts to provide their supposed scientific expertise".

Really? I want my husband in on this gig, because he participates in NIH grant peer review in "basic science" and considers it the equivalent to "jury duty for scientists" - important to do for the greater good, but definitely NOT lucrative.

When he is on a study section, 4 times a year for several years (maybe 5?) he spends several long evenings reviewing grants (I consider that jury-duty for our family), then has to fly cross-country for the study section meeting and back again. His travel expenses are covered and he usually blows the minimal honorarium on a good meal. The honorarium doesn't even come close to covering his time, the jet lag, and the disruption of his own research. Some study sections now operate via teleconferencing, which eliminates the travel time, but still takes 3-4 days out of his work schedule.

And the government regulations regarding "contractors" is so cumbersome and confusing since the war in Iraq that it's hardly worth filing the paperwork for his honorarium and misc travel expenses (airport parking, taxis, etc.). Peer-review, at least for the NIH, is essentially volunteer work. The value to the participant is experience in how the peer-review system works - and it can make one a better grant writer.

I'm always in awe at the money spent on some pretty goofy research areas or bad studies (like this one you reviewed) that are reported in the media as if they are significant, yet basic research, the foundation of hard science is woefully underfunded and losing ground (and talented people, who go into more lucrative fields or are recruited into industry).

So I want to know more about these well-paid peer reviewers and how to sign up my husband. NIH's pie piece of the stimulus deal in his area of research worked out to about 200 new grants a year - a drop in the bucket, really. There are a heck of a lot more than 200 senior levels scientists (and their staff, graduate students, etc.) who have been slowly squeezed dry by the tight grant situation since our money was diverted into "spreading democracy and freedom"...

Nick said...

Jenny, I'm curious about your thoughts on fructose. Dr Davis interpreted the study by implicating fructose in any form, rather than the fructose from HFCS used in beverages. It seems that the processing of HFCS could have had some impact on the health of the people in the study vs. if they would have eaten an equivalent amount of fructose from fruit.

I think there is pretty compelling evidence that fructose in all forms should be limited, but I don't think I have seen anything like the study using beverages that that used fruit instead of beverages.

Jenny said...
This comment has been removed by the author.
Jenny said...

don't know enough to answer your question. I eat fruit and I think it is wrong to demonize it. I think it is an issue of the quantity of fructose involved when people are sucking down huge sodas and developing fatty livers etc. But there is data in both directions about whether HFCS is worse than Sucrose or not.

If you keep your blood sugar under control, it is not possible to eat huge amounts of fructose filled naturally occurring foods as they also have starch or sucrose. But to give up tomatoes or fresh berries seems a shame to me. There are lots of significant micronutrients in them and to lump them with Pepsi seems to me, at least, mistaken.

trinkwasser said...

Here's another pretty large confounding factor. Why has the use of nitrogen fertiliser grown so rapidly during this period? Because modern wheat varieties require high amounts of feed to produce the huge yields they are now capable of producing.

The wheats are specifically bred for three things: large heads, short stems and disease resistance. It's not at all unlikely this, especially the latter factor, increases the amount of lectins in the grain. This is just as likely to be causal, and the increased nitrogen use just an indicator of the type of wheat grown.

It may also relate to the corn used for HFCS in the US, and the beet used for sugar in the UK. Most high yielding crop varieties are dependent on high nitrogen input.

carriecz said...

N is THE limiting factor for plant growth. Other than water, the nitrogen levels available will determine how tall a plant will grow. Although it is mostly for the stem, for some plants, too much N will cause little to zero fruit. For grains, which are grown in fields and mostly in monoculture, ammonia is usually applied to replace the N taken from the soil from the previous crop. A plant cannot survive without nitrogen, it is the major building block. I don't see a viable way to avoid consumption of nitrogen. Although rest assured, ammonia is VERY expensive and farmers will not waste a drop. Even in a good case of corn following a soy crop, where the soybean fixes N in the soil, or if you have green manure (plants overwinter), there may not be a need for added ammonia but the N in the soil has not been removed from the equation. It just got there in another way. And "organic" application of manure, well sure the N level is what drives the amount of manure applied. (nevermind the heavy metals in the manure, that's another topic)
Thank you for this story. It has set my mind at ease. I wish the answers were easier to find.