31 January, 2012

The diabetes boogeyman: Does diabetes make you fat rather than the other way around?

A boogeyman often invoked by the food fascists in their efforts to get control over what we eat

That obesity causes diabetes is one of the most entrenched medical mantras that there is. Even the (mythical) lifespan benefits of statins and antioxidants are not as often invoked as the relationship between obesity and diabetes. I think I see the relationship touted at least once a day somewhere.

But think!

* Most overweight people don't get diabetes and some diabetics are slim!

* It is true that certain diets can help control diabetes but that does not mean that diet causes diabetes. A roof can shelter me from the rain but that does not mean that roofs cause rain!

* There is undoubtedly a correlation between being overweight and having obesity but correlation is not causation and it could well be that at least in some people diabetes causes you to put on weight. I have certainly seen instances of the latter, where an overweight person discovered they had diabetes, went on a diabetic diet and promptly lost a lot of weight. It was apparently diabetes that was making them fat. Excessive appetite is a known symptom of diabetes so that is hardly a surprise. And all the so-often-quoted research findings are just correlations, so prove nothing about which way the causal arrow points.

* There are various studies of diabetes in mice but mice are a very short-lived species whereas we are a very long-lived species. We have obviously evolved life-preserving mechanisms that mice have not so what is a problem in mice might be quite easily coped with by the human body. And that is one reason why rodent findings often do not generalize to people.

So I am going to reproduce below something I wrote in 2008. I know of nothing which would alter my conclusions since. If anybody can tell me something important that I have overlooked, however, I would be delighted to hear it:
I decided to look at the evidence behind the claim. I looked at what appear to be the two most cited articles on the question — by Seidell and by Mokdad et al..

Neither article goes any where near proving the claim. Seidell, in fact, notes the differing relationship between weight and diabetes in Asia versus the West and makes the entirely sensible observation that the two things are “common consequences of changing lifestyles” — NOT directly linked, in other words. Both, for instance, could be a consequence of (say) reduced exercise.

And the Mokdad article is quite naive. It shows that fatties are more likely to have diabetes but again enables no causal inferences. Additionally, it does not allow for the curvilearity that is known to feature in relationships with obesity. In other words, it combines moderately overweight people with grossly overweight people — which flies in the face of the fact that it is people of middling weight who live longest. It could be just the real fatties who tend to get diabetes at an accelerated rate.

And genetic effects are, of course, not mentioned anywhere, despite all we know (and have known for a long time) about the genetic influence on body weight. It could be that a genetic difference causes both diabetes AND a larger fat mass. So even severe dieting would not chase that pesky diabetes-causing gene away.

The fact that prevalence of diabetes has been increasing would seem at first to discount a genetic influence but it does not, of course. Many genetic influences need environmental “triggers” to become dominant and we just don’t know what environmental triggers might have come to the fore in recent years. How about increased crime causing both stress and overeating as a response to stress? Who knows?

As far as I can see, then, the alleged effect of fat on diabetes is just a guess. Ho hum! Just another instance of crap medical “wisdom”.

I wonder do pigs get diabetes? Fat pigs are a byword. And pigs are a pretty good animal model for human beings. Pig tissue is even used for direct implantation into human hearts! Rodent models always have dubious generalizability but I think I would believe a double-blind study with pigs.

So, you see, I am not like those (such as the Global Warmists) for whom no evidence will count. I have just specified precisely what evidence would convince me. And nor would the evidence concerned be hard to gather. You might even get some good bacon at the end of it! Yum!


An example here ("Alec Baldwin reveals he has lost 30 pounds after discovering that he is prediabetic") of diabetes being associated with weight-gain and the gain being reversed by dealing with the diabetes. That train of events DOES suggest a causal inference: That it was the diabetes that caused the weight gain and not vice versa.

Update 2: Aborigines

I probably should say something about diabetes among Australian Aborigines. Within living memory, diabetes was virtually unknown among Aborigines but it is now very common. This is of course very interesting epidemiology and dietary change is the conventional explanation. At the beginning of the period Aborigines still fed themselves mostly by traditional hunter/gatherer activities and so got plenty of exercise and a varied diet of minimally processed food.

Nowadays with more generous government welfare provision they subsist largely on the most disparaged of Western foods --such as Coke and potato chips. That this switch from an aspirational diet to a disparaged one has been accompanied by a rise in diabetes tends to give epidemiologists erections. It should not, however, as there have been other large changes at the same time -- in particular the availability of alcohol has greatly increased.

At the beginning of the period it was in fact illegal to supply Aborigines with alcohol. Unlike Westerners, they have not had millennia to adapt to it and so handle it very badly. But paternalistic restrictions are out of fashion now so there is now a very large alcohol abuse problem in Aboriginal communities and it seems entirely possible that this is behind the diabetes upsurge.

To make matters worse, there is today a definite problem with "metho", which was at least much less in the past. "Metho" is a combination of ethyl alcohol and methyl alcohol, a quite poisonous product, but it is cheap and very intoxicating -- and some stores frequented by Aborigines are even reputed to keep it in the fridge for sale to Aborigines as a beverage. It even has a pet name among Aborigines: "White lady". How they survive drinking it at all is a mystery.

So diet is only one candidate for explaining health problems among Aborigines.

I should perhaps end up with a clarifying note: In all my remarks above I had in mind the form of diabetes that accounts for at least 90% of the cases: Diabetes mellitus type 2.

Update 3: An actual cause of diabetes -- Statins?

The study below is large so the results should be fairly robust. And the effect was fairly large as such effects go. These results may go some way towards explaining the apparent rise in diabetes incidence in recent years. The doctors hand out statins like jellybeans these days.

Given the pervasive side-effects of statins, it seems likely that the statins did cause the increase in diabetes observed but it must be noted that we are again here dealing with correlational data -- which is not capable of supporting causal inferences by itself
Arch Intern Med. 2012;172(2):144-152.

Statin Use and Risk of Diabetes Mellitus in Postmenopausal Women in the Women's Health Initiative

By Annie L. Culver et al.


Background: This study investigates whether the incidence of new-onset diabetes mellitus (DM) is associated with statin use among postmenopausal women participating in the Women's Health Initiative (WHI).

Methods: The WHI recruited 161 808 postmenopausal women aged 50 to 79 years at 40 clinical centers across the United States from 1993 to 1998 with ongoing follow-up. The current analysis includes data through 2005. Statin use was captured at enrollment and year 3. Incident DM status was determined annually from enrollment. Cox proportional hazards models were used to estimate the risk of DM by statin use, with adjustments for propensity score and other potential confounding factors. Subgroup analyses by race/ethnicity, obesity status, and age group were conducted to uncover effect modification.

Results: This investigation included 153 840 women without DM and no missing data at baseline. At baseline, 7.04% reported taking statin medication. There were 10 242 incident cases of self-reported DM over 1 004 466 person-years of follow-up. Statin use at baseline was associated with an increased risk of DM (hazard ratio [HR], 1.71; 95% CI, 1.61-1.83). This association remained after adjusting for other potential confounders (multivariate-adjusted HR, 1.48; 95% CI, 1.38-1.59) and was observed for all types of statin medications. Subset analyses evaluating the association of self-reported DM with longitudinal measures of statin use in 125 575 women confirmed these findings.

Conclusions: Statin medication use in postmenopausal women is associated with an increased risk for DM. This may be a medication class effect. Further study by statin type and dose may reveal varying risk levels for new-onset DM in this population.


Go to John Ray's Main academic menu
Go to Menu of longer writings
Go to John Ray's basic home page
Go to John Ray's pictorial Home Page
Go to Selected pictures from John Ray's blogs