Tuesday, June 12, 2012

Pasta-eating, beer-drinking Swedes and LCHF diets

A large new 25 year observational study from Sweden published in Nutrition Journal this month has found that Swedes started increasing their fat consumption from about 2004 onwards at the same time that low carbohydrate, high fat (LCHF) diets became popular.   This was a reversal of a trend towards lower fat options that started in the mid-1980's.

In addition to a distinct switch away from low fat margarine to butter, the following tables show a growing trend towards increased fat in the diet:

Far from losing weight as a result of the increased fat consumption (something the low carb community might expect), the body mass indexes (BMI) of the Swedes participating in the study continued to increase throughout the 25 year study period:

What was particularly worrisome to researchers was an apparent reversal in the long term downward trend of serum cholesterol levels starting around 2004, leading them to believe that Swedes were putting themselves at increasing risk of cardiovascular disease:

At this point you are probably thinking that this does not bode well for proponents of low carbohydrate eating.  After all, what could be clearer than the following trends away from carbohydrates in favour of fats?

As it happens, this study leaves me with more questions and skeptical observations than constructive conclusions:
  • While the percentage of energy from carbohydrates has been steadily falling in recent years, I would like to know whether the total calories consumed changed in this period, and if so, whether they increased or decreased.
  • On page 6, the study notes that there were 23 observations of BMI measurements under 10, and 3 above 100.  I don't know if any living person has ever had a BMI above 100.  Below 10 is also very unlikely.  One person was also recorded as measuring 2.7 meters (that's almost 9 feet!).  It's well known that the Swedes are some of the tallest people in the world, but that's rather extreme.  If the trained nurses who recorded these measurements made these 27 mistakes, I hope we can assume the others were more or less accurate.
  • This was an observational study.  There are many valid criticisms of observational studies drawn from self-reported questionnaires.  Portion sizes are notoriously variable, and people are often far from accurate in what they report that they eat.
  • It is very unfortunate that only total cholesterol was reported (it does not appear to have been measured).  Without knowing the subjects' triglyceride levels, or their HDL, the cholesterol trends in the table above are predictive of nothing. This must have been a terrifically expensive study, and it's a shameful waste not to have at least measured HDL.  
  • A very interesting chart shows that alcohol consumption showed a clear and steady upward trend.  In particular, men's consumption of export beer showed a steady increase that appeared to have had absolutely no correlation to the popularity of LCHF dieting.  Do you think those beer drinkers public realize what a carb-load beer provides?