Tuesday, October 18, 2005

The Deception of Statistics

In a review study in the July 2005 British Journal of Medicine: Changes in dietary fat and declining coronary heart disease in Poland: population based study, Zatonski and Willett explore the impact of dietary changes in Poland on coronary heart disease.

They suggest that the reduction of over a third of coronary heart disease in Poland between 1990 and 2002 is attributed to increased consumption of polyunsaturated fats, with a reduction in saturated fats. They conclude this after accounting for and including estimates that show a small reduction in smoking and an increase in the intake of imported fruit, neither of which they believe were large enough to have made the impact on the significant declines in cardiovascular disease seen.

Before I continue, let's be clear here - the reserachers concluded it was a decrease in saturated fat intake and an increase in polyunsaturated fat intake that was the reason for the decrease in cardiovascular disease in men (38%) and women (42%) in Poland.

So how much of a decrease in saturated fat are we talking about here?

What the researchers found was that a decrease in saturated fat, coupled with an increase in polyunsaturated fat (which changed the ratio of the two combined by 70%) was the main reason for the improvements in cardiovascular health.

The researchers build their case to reduce saturated fat by pointing out that statistically saturated fat intake decreased by 7% in the time period investigated. Statistically, 7% seems like a significant amount, however, in real grams of saturated fat intake this was a very minor decrease from an average of 44.8g per day to 41.5g per day, or 30 less calories from saturated fat per day.

They continue to build their case by showing that the statistical increase in consumption of polyunsaturated fat intake happened at the same time - it rose 57% in the time period, but in real terms was an increase from 14.89g per day to 23.3g per day, or an additional 75 calories from polyunsaturated fats.

Now here is where things get a bit interesting.

The data isn't "consumption data" from what people actually eat, but "availability data" on what foods are available to eat. And, even with what is best a "minor" decrease in saturated fat intake, at the lowered intake of 41.5g per day, such an intake of saturated fat is still HIGH compared with our dietary guideline of less than 10% of calories from saturated fat.

Within our own guidelines we have maximum levels of saturated fat intake provided at various calorie intake to use as a guide to meal planning - for example, at 2000-calories, 10% or less saturated fat is 20g or less in a day; at 2500-calories, 10% or less saturated fat is 25g or less. Unless these folks were eating more than 3735-calories per day, they were consuming more than 10% of their calories from saturated fat if they're eating 41.5g a day!

In this study, the researchers never tell us what the percentage of calories are from total or saturated fat, but based on how many calories must be eaten to maintain less than 10% of calories from saturated fat as is recommended in the United States, it's easy to calculate that this population is eating more than 10% of their calories from saturated fat due to the high calorie intake that must be eaten to eat less than 10% of calories from saturated fat.

So, even though it is unstated by the researchers, this population is eating a high fat diet. There is just no way around it when you do the math.

Now, even with the minor decrease in saturated fat, the actual fat intake from saturated fat and polyunsaturated fat increased when the two sources are taken together. In fact, the increase is from 59.6g per day from saturated and polyunsaturated to 64.8g per day from both. So this population increased their intake of fat from saturated and polyunsaturated fats combined.

The researchers do not reveal what the monounsaturated fat intake is, so we have no way of knowing if total fat increased or not. What we do know however, is that generally monounsaturated fat intake is usually almost twice that of polyunsaturated due to the monounsaturated fats in oils typically higher in polyunsaturated fats.

In this study, the researchers noted that the increased oils eaten were rapeseed (canola) and soya (soybean) oils and a quick look at the fatty acid profile of these oils shows that 100g of these oils contain 55.3g of monounsaturated fat and 32.4g of polyunsaturated fat along with 7.8g of saturated fat. So if the population consumed 23.3g of polyunsaturated fat, it is safe to assume they consumed at least 39g of monounsaturated fats in their overall dietary intake of fats if we use the canola and soybean oil profile.

Let's do some quick math...

  • 41.5g known from saturated fat
  • 23.3g known from polyunsaturated fat
  • approximately 39g of monounsaturated fat
  • Total = 103.8g of fat in a day, or 934.2 calories from fat.

The researchers stated in their conclusion, "The experience of Poland is consistent with epidemiological and clinical evidence(4) indicating that mortality due to coronary heart disease can be reduced by partly replacing dietary saturated fats with polyunsaturated fats while maintaining a low intake of trans fatty acids."

Oh really?

If this population is indeed eating 2,500-calories each day on average for both men and women, approximately 934 from fat represents 37.3% of calories from fat and 15% of calories from saturated fat.

Now some may say, but you simply don't know the calories or the monounsaturated fat intake. Well, guess what? The researchers don't know either - they used food availability data to build their case, not actual food consumption data, so my guess here is as good as theirs is! And no matter how you slice it, 41.5g from saturated fat per day is (when compared with our dietary guidelines) a high intake of saturated fat!

I wondered how far off I might be in declaring this a high-fat diet and looked to see if any other studies are available based on consumption data and that give some hard percentage of calories from fat(s) in the diet. Sure enough, in a Polish medical journal (Zaklad Epidemiologii) data was published in 2003 that looked at dietary habits during the period of 1993-2001.

The findings support my assertions:

  • Total fat intake 37.8% in men 2001
  • Total fat intake 35.9% in women 2001

If anything, the data here supports eating a nutrient-dense diet, rich with animal fats and fruits and vegetables.

The data is best viewed by what is not explored - total fruit consumption (the researchers looked only at imported fruits), vegetable consumption, reduction in stress levels over the period of time due to governmental policy changes, living standard changes, etc. - each of these is a confounding factor that remains unexplored and each of these confounding factors has indeed been highlighted in critical citations of this review study.

In this month's journal, Uffe Ravnskov states in his letter, "Zatonski's and Willett's statement that their finding is supported by epidemiological and clinical evidence is not true either. In a review of all cohort and case-control studies, heart patients had eaten more saturated fat than had healthy controls in three cohorts, but in one cohort they had eaten less, and in 22 cohorts and in six case-control studies no difference was found."

In July's journal, another letter pointed to the economic changes not included in the review data, "Lock and McKee challenge these claims and discuss the likelihood of Europe's agricultural policy damaging the progress achieved between 1991 and 2002. "

In July's journal, a commentary stated that the braod range of fruits and vegetables consumed must be accounted for - "The paper clearly shows that dietary change had a major impact on cardiovascular mortality, but the authors, by considering changes in the intake of imported fruit only, have underestimated changes in consumption of the full range of fruits and vegetables known to have significant cardioprotective effects. The introduction of a market economy had major consequences for domestic production and the retail sector. There has been large investment in agriculture, with central European producers now integrated with wider European agricultural production. Much non-traditional produce is now grown locally, and fruits and vegetables are available to consumers throughout the year. Consequently, to capture the full impact of changing availability of food since the 1990s, the changes in both the total and the seasonal consumption of all fruits and vegetables need to be taken into account."

So, while some researchers continue to insist that saturated fats are a villan in our diet and will grasp at straws for some way to prove that reducing saturated fat is cadioprotective, this particular study is no more than data manipulation and statistical deception. 41.5g of saturated fat consumed daily is not low-fat and is not 10% or less calories each day from saturated fat. In this paper saying such is avoided by sticking with the statistics rather than the cold hard real numbers in grams and calories each day.

Who are they trying to kid? Are we really supposed to believe that reducing saturated fat intake by a mere 3.3g per day is going to lead to a decline in heart disease by 38-42%? This reduction while still consuming what could well be 15% of total calories from saturated fat?

I don't know about you, but this data strongly suggests that a change in the quality of the diet and imporvements in living standard - even with a high intake of saturated fat - is cardioprotective! This populations diet is not low-fat, is not low in saturated fat - so, don't let the statistics blind you into thinking it is!

1 comment:

  1. Anonymous10:17 PM

    Link exchange is nothing else however it is just placing the other person's web site link on your page at suitable place and other person will also do similar in support of you.
    Also visit my web-site : arm