Critical insights into the etiology of insulin resistance have been gained
by the use of animal models where insulin action has been modulated by stri
ctly controlled dietary interventions not possible in human studies. Overal
l, the literature has moved from a focus on macronutrient proportions to un
derstanding the unique effects of individual subtypes of fats, carbohydrate
s and proteins. Substantial evidence has now accumulated for a major role o
f dietary fat subtypes in insulin action. Intake of saturated fats is stron
gly linked to development of obesity and insulin resistance, while that of
polyunsaturated fats (PUFAs) is not. This is consistent with observations t
hat saturated fats are poorly oxidized for energy and thus readily stored,
are poorly mobilized by lipolytic stimuli, impair membrane function, and in
crease the expression of genes associated with adipocyte profileration (mak
ing their own home). PUFAs have contrasting effects in each instance. It is
therefore not surprising that increased PUFA intake in animal models is as
sociated with improved insulin action and reduced adiposity. Less informati
on is available for carbohydrate subtypes. Early work clearly demonstrated
that diets high in simple sugars (in particular fructose) led to insulin re
sistance. However, again attention has rightly shifted to the very interest
ing issue of subtypes of complex carbohydrates. While no differences in ins
ulin action have yet been shown, differences in substrate flux suggest ther
e could be long-term beneficial effects on the fat balance of diets enhance
d in slowly digested/resistant starches. A new area of major interest is in
protein subtypes. Recent results have shown that rats fed high-fat diets w
here the protein component was from casein or soy were insulin-resistant, b
ut when the protein source was from cod they were not. These are exciting t
imes in our growing understanding of dietary factors and insulin action. Wh
ile it has been clear for some time that 'oils ain't oils', the same is now
proving true for carbohydrates and proteins.