Should you believe the Chocolate as "Health Food" Hype?

You've read the headlines:

• Chocolate Lovers May Be Lowering Their Risk of Heart Disease (International Business Times, August 29, 2011)
• Heart Disease, Diabetes and Stroke: More Chocolate Less Risk? (ABC News, August 29, 2011)
• Scientists announce amazing findings: chocolate offers huge protection from heart disease and stroke (Natural News, August 30, 2011)
• High Chocolate Consumption May Reduce Heart Disease Risk By One Third (Medical News Today, August 30, 2011)

These recent news stories touting the health-promoting potential of chocolate were instigated by a research study published in the British Medical Journal.  The original story had a less memorable title: Chocolate consumption and cardiometabolic disorders: systematic review and meta-analysis.  While the authors found that the highest levels of chocolate consumption were associated with a 37% reduction in cardiovascular disease compared to the lowest levels; they concluded that "corroboration is now required from further studies, especially experimental studies to test causation rather than just association." 


One of the most confounding factors of this research (and most investigations into the links between diet and health) was the method by which data was collected.  Six of the seven studies included in the analysis relied on food intake questionnaires, which are at the mercy of participant recall and subjective estimations.  The most thorough statistical methods and the most robust formulas cannot compensate for a data set inherently flawed by human error.  So what the research really tells us is that people who report eating the most chocolate have the least signs and symptoms of cardiovascular disease.


In the full text of the BMJ article, the authors admit: 


Chocolate intake is likely to be underestimated by consumers, and may be underestimated to a larger extent by those with a higher body mass index. As people with a higher body mass index are also more likely to have a cardiovascular disease outcome, then the underestimation of their chocolate intake may induce an artificial inverse association between chocolate and risk of cardiovascular disease.


Other factors that might hamper the quality of recording chocolate consumption also need consideration. These include the potential effect of recall bias and the challenges of recording snacks (which might include chocolate) as these are generally under-reported compared with meals.


In other words, fat people underestimate how much they eat (which may explain why they are fat).   Since excess weight predisposes a person to heart problems and excess weight is generally a result of inflated portion sizes, then someone with a BMI above the normal, healthy range is likely to eat more chocolate because he/she eats more in general than someone who weighs less.  When a skinny person, like myself, reports eating "a piece" of chocolate, she is probably referring to a single square imprinted in a chocolate bar.  An overweight consumer is more likely to consider an entire chocolate bar to be a  single "piece."  What many fat people don't understand is that you can't deny eating a candy bar just because no one saw you raid the hidden stash in your underwear drawer (the scale doesn't lie).


The authors of the BMJ article also noted the "heterogeneity in reporting and measuring chocolate consumption."  Four of the five food intake questionnaires used did not probe portion size at all.  They simply asked about frequency of chocolate consumption in multiple choice format.
  
Study A: (a) Never, (b) Once a month, (c) Once a week or more
Study B: (a) Less than once a week, (b) Once a week or more
Study C: (a) Never, (b) Less than once a month to less than once a week, (c) Once a week, (d) more than once a week
Study D: (a) None, (b) 1-3/month, (c) 1-4/week, (d) >5/week


Due to the variation, only the lowest and highest categories could be used to associate chocolate consumption with heart disease.  In my opinion, complex statistics and scholarly considerations are useless given the vague multiple choice answers.  You might as well ask a bunch of people on the street if they eat "a lot" or "a little" chocolate. 


And to add insult to injury, subjects reported consumption of chocolate in a wide variety of forms: chocolate bars, chocolate drinks, chocolate desserts, nutritional supplements, biscuits, etc.  Researchers did not distinguish between a slice of chocolate layer cake and a Snickers or a dark chocolate bar and handful of Hersey's kisses.  This fact alone discredits the results of the analysis for me.  You might as well interpret the data as suggesting that people who prefer chocolate over vanilla things have less heart disease.  And the BMJ article does include the "Necessary Cautions":


Beyond the caution needed in interpretation of data from observational studies, one must also consider other aspects associated with chocolate consumption. For instance, the high energy density of commercially available chocolate (about 2100 kJ (500 kcal)/100 g) means excessive consumption will probably induce weight gain, a risk factor for hypertension, dyslipidaemia, diabetes, and cardiometabolic disorders in general. Hence the high sugar and fat content of commercially available chocolate should be considered, and initiatives to reduce it might permit an improved exposure to the beneficial effect of chocolate. However, the articles included in our analysis did not provide the information needed to evaluate any potential differences between different types of chocolate in the associations with cardiometabolic disorders.


I love chocolate as much as the next guy and do crave it sometimes.  But I'm not going to eat it daily in an attempt to increase my longevity.  Listen to that annoying, common sense voice in your head. If it sounds too good to be true, it probably is...