Monday, June 13, 2011

Nutrient Intake and Substrate Oxidation part 2

Nutrient Oxidation and Nutrient Balance Results in Hill et al. "Nutrient balance in humans: effects of diet composition", AJCN, 1991, vol. 54, pp. 10-17.

The experiment provides convincing evidence that oxidation of substrates varies with intake. Fig. 2* can also be used to verify roughly (by printing out the graph, estimating the nutrient oxidation of each bar using the scale and adding them up) that the total nutrient oxidation does not differ significantly between diets and between diet and baseline.

Nutrient Intake and Substrate Oxidation

or some comments on Hill et al. "Nutrient balance in humans: effects of diet composition", AJCN, 1991, vol. 54, pp. 10-17.

Recently the study above was analysed and presented (on Primal Wisdom blog q.v.) as support for arguments that:
  1. the body does not manage 'energy balance', it manages substrate stores and
  2. high fat diets may not be good for weight loss because they may make it too easy to (paraphrasing the end of the blog post) consume more fat than your body burns daily, (so that) you will increase your body fat day by day.

Now, point no. 1 is both complicated and simple. It is complicated in that it seems clear from much emerging evidence about hormones such as leptin and ghrelin and brain functions that the brain and the body does have extremely complicated mechanisms for controlling 'energy balance' (see Whole Health Source for many articles and discussions on this topic). It may do this by either controlling input e.g. by modulating appetite or by controlling output e.g. by modulating metabolic rate, activity levels (inclination to be more or less active) or waste heat production. It is simple only in the fact that there are indeed three basic macronutrient categories which correspond roughly to 'strucutural' (i.e. protein), 'limited storage' (i.e. glycogen (carbohydrate) and 'unlimited storage' (i.e. adipose (fat)). It is within this very limited perspective that this particular paper explores the issue which amounts to peering at metabolism through a keyhole. This paper succeeds in showing that nutrient intake is an independent factor influencing substrate oxidation so that body/brain control of 'energy balance' has to work within the constraints of the hormonal responses to nutrient intake but it doesn't prove that it trumps all other influences.

The second point of the blog article is that high fat diets may be an impediment to fat loss because if you eat high fat it is too easy to store any fat you accidentally overeat. I do not want to imply that the blog article is stating that high fat diets will inevitably cause weight gain or impede fat loss, but it comes very close to claiming this:
You might be able to achieve this on a low carbohydrate diet, and you might not. If eating a low carb diet allows you to eat less fat than you burn daily, you will lose fat, and if it doesn’t you will not. On the other hand, regardless of theoretical "energy" intake, if eating a low carbohydrate diet results in your consuming more fat than your body burns daily, you will increase your body fat day by day.
So let's look at the study, what it set out to do, how it did it and what the results were.

There were two main aims of the study stated by the authors: to examine the effect of diet [macronutrient] composition on [overall] energy expenditure and on nutrient balance. Their conclusions were that diet composition did not affect total daily energy expenditure but did affect nutrient balance as the substrate oxidation shifted to match the diet composition.

The subjects in the study were overweight to moderately obese: BMIs ranged from 26 to 34 and body fat% in the 30% range in the males and high 30% to high 40% in the females. The study methods on the whole are precise and careful. For example, body composition was measured using underwater weighing and the functioning and calibration of the indirect room calorimeter are described in detail. Activity or movement in the room calorimeter was estimated using a radar monitoring system and estimates of sleeping metabolic rate (SMR), as distinct from resting metabolic rate (RMR) were also made. Body composition, RMR and 24 hour energy expenditure were measured repeatedly throughout the experiment:
"1-2 d before beginning each experimental feeding period."
"Body composition was determined at the beginning and end of each experimental feeding week..."
but this data is not reported.
The paper is also less than clear when it describes caloric intake. From the abstract:
"For each subject, total caloric intake was identical on all diets and was intended to provide the subject's maintenance energy requirements."
To do this the amount of food given to the subjects was estimated as 1.5 x RMR where the RMR (which includes TEF (thermic effect of food)) was calculated based on a 24 hour session in the room calorimeter. It was noted by the authors that the resulting intake was
"for a level of activity typically seen during 24 h in the room calorimeter"
and pointed out that this
"might have resulted in negative energy balance on other days when activity outside the calorimeter may have been greater."

It is stated that all food was consumed on site at the research centre and food that was not eaten was weighed. This information is used later to calculate the difference between intake and oxidation for the different substrates. However caloric intake on any of the diets is not reported.

The results for daily energy expenditure (DEE) support the assertion that caloric intake was identical for each diet for each subject as the authors state that total daily energy expenditure did not change according to diet. It has to be assumed then that any changes shown in Table 3 are non-significant (although this isn't explicitly stated), as there is some variation within each subject from measurement to measurement (it would be surprising if there wasn't). Mean results for RMR, SMR and energy expenditure due to activity as shown in Table 4 are more convincing as it is easy to see that the standard error on the means puts all the results for baseline and the different diet periods within reach of each other.

There are two exceptions:
  1. the authors note that the values for the mixed diet period are higher and that this is because two subjects, who happened to have the lowest activity levels, were excluded from this part of the experiment
  2. the mean energy expenditures due to activity during the high carb diet were noticeably (although again presumably not significantly) lower than for the high fat diet.
However, again it seems odd that given that the information for caloric intake and energy expenditure was available, that overall caloric balance was not reported.

In the follow-on post, the nutrient balance figures will be discussed.