particle diameter tends to be decreased in patients with Type 2 diabetes (23), and the postprandial accumulation of TG-rich lipoproteins accentuated (24).
The close association of the latter two abnormalities with hypertriglycerid- aemia makes it less necessary to directly measure LDL particle diameter or postprandial lipaemia; if the plasma TG concentration of a patient with Type 2 diabetes exceeds 2 mM, it is almost certain that the individual will have smaller and denser LDL particles and day-long increases in remnant lipoprotein (RLP) concentration.
The fact that LDL cholesterol concentrations are not higher in patients with Type 2 diabetes than in non-diabetic individuals (22) does not preclude the need to consider how dietary recommendation would affect LDL cholesterol concentrations.
Indeed, there is evidence that lowering of LDL cholesterol concentrations by pharmacological means will decrease risk of CHD (25,26).
Based on the above considerations, it seems clear that dietary recommenda- tions for patients with Type 2 diabetes, above and beyond weight loss, must take into account the effects of a given intervention on both LDL cholesterol metabolism, and the atherogenic lipoprotein phenotype of hypertriglycerid- aemia, low HDL cholesterol concentration, smaller and denser LDL particles, and an exaggerated degree of postprandial lipaemia.
The goal is to maximally decrease CHD risk factors attributed to both forms of abnormal lipoprotein metabolism.
LDL Cholesterol There is abundant evidence in non-diabetic individuals that replacing SF with either unsaturated fat or CHO will lower LDL cholesterol concentrations to a similar degree (27,28), and this appears to be the case in patients with Type 2 diabetes (15,17, 20). Given the evidence that pharmacological lowering of LDL cholesterol decreases CHD risk in patients with Type 2 diabetes (25,26), the intake of SF should be limited to less than 10% of total calories in these individuals.
TG-rich Lipoproteins The decision to replace SF with either unsaturated fat or CHO will have an enormous impact on the circulating concentration of TG-rich lipoproteins.
There is substantial evidence in patients with Type 2 diabetes that diets low in SF and high in CHO will increase fasting plasma TG concentration as compared to substitution of SF with MUFA/PUFA (15,17, 20).
Evidence has recently been published that fasting RLP concentrations are also increased in patients with Type 2 diabetes (29), and the postprandial accumulation of RLPs of both endogenous (hepatic) and exogenous (intestine) origin are increased in patients with Type 2 diabetes (30).
This latter observation is not simply a
194 NUTRITIONAL MANAGEMENT OF DIABETES MELLITUS
function of the increase in fasting TG-pool size characteristic of patients with Type 2 diabetes, but appears to also involve a decrease in the removal rate from plasma of TG-rich lipoproteins following a mixed meal.
HDL Cholesterol Although a low HDL cholesterol concentration in patients with Type 2 diabetes is usually associated with a high plasma TG concentration, the difficulty in raising HDL cholesterol concentration with dietary manipulation is in marked contrast to the relative responsiveness of plasma TG concentra- tions.
The reason for the difference is not clear.
There is evidence that HDL cholesterol concentrations are inversely related to the fractional catabolic rate (FCR) of apo lipoprotein A-1 in patients with Type 2 diabetes (31), the more rapid the FCR of apo A-1, the lower the HDL cholesterol concentration.
Furthermore, the higher the plasma insulin response to an oral glucose challenge, the faster the apo A-1 FCR (31).
Perhaps the changes in circulating insulin concentration resulting from relatively minor variations in macro- nutrient composition in patients with Type 2 diabetes are not sufficient to modulate the FCR of apo A-1.
Irrespective of the explanation, it appears that dietary manipulations have relatively little effect on HDL cholesterol concentrations in patients with Type 2 diabetes.
Dyslipidaemia and CHD Risk As discussed above, there is evidence that drug-induced decreases in LDL cholesterol concentration decrease risk of CHD.
Although there is ongoing debate as to whether or not hypertriglyceridaemia is an ‘independent’ risk factor in non-diabetic individuals (32), the importance of increases in plasma TG for predicting CHD in patients with Type 2 diabetes seems less controversial (32, 34).
There is little reason to question the importance of a low HDL cholesterol as a CHD risk factor, and there is increasing evidence in non-diabetic subjects of the atherogenic potential of postprandial lipaemia (the accumulation of RLPs throughout the day), and the appearance of smaller and denser LDL particles (35, 37).
Thus, it seems prudent to suggest that dietary recommendations for patients with Type 2 diabetes should take into account the predictable effects on lipoprotein metabolism.
METABOLIC EFFECTS OF VARIATIONS IN THE KIND OF CARBOHYDRATE IN ISOCALORIC DIETS
Evidence presented to this point has focused on the effects of variations in the relative amounts of dietary fat and CHO in isocaloric diets, and emphasised the
VARIATIONS IN DIETARY FAT AND CARBOHYDRATE 195
untoward effects of replacing SF with CHO versus PUFA/MUFA. Another possible approach to this issue would be to continue the practice of replacing SF with CHO, but doing this with the kind of CHO that would maintain the beneficial effects of low SF, high CHO diets on LDL cholesterol concentration, without leading to the harmful impact on glucose, insulin and lipoprotein metabolism that has been observed with CHO-enriched diets. In this context, two different, but somewhat related, approaches have been evaluated , emphasising either the glycaemic index of the high CHO diets, or their fibre content.
VARIATIONS IN GLYCAEMIC INDEX
Perhaps the best example of the ability of differences in glycaemic index of CHO-enriched diets to modify glycaemic control and lipoprotein metabolism in patients with Type 2 diabetes is the report by Jarvi and colleagues (38).
These investigators compared the metabolic effects of two diets, each containing 55% of total calories as CHO, in 20 patients, consuming each of the test diets for 24 days.
The glycaemic indices were calculated to vary from 57 to 83 as compared to white wheat bread.
The two test diets were compared to each other, as well as to baseline values obtained on an uncontrolled diet.
Of considerable interest was the observation that fasting plasma glucose, TG, and LDL cholesterol concentrations fell on both diets, supporting the general belief that essentially any prescribed diet is better than no diet plan.
On the other hand, the degree of improvement in all of these variables was the same, irrespective of the difference in glycaemic index of the diet.
Furthermore, the improvements in day-long plasma glucose and insulin concentration appeared to be of somewhat lesser magnitude than in the study in which the CHO intake was reduced, and unsaturated fat intake increased (20).
Brand and associates (39) conducted a somewhat similar study in 16 patients with Type 2 diabetes, comparing the metabolic effects of two diets, differing in their glycaemic index from 77 to 91.
However, both of the diets were relatively low in CHO (*45%).
The fasting plasma glucose concentrations were similar after the high glycaemic index and low glycaemic index diets, as were the fasting concentrations of plasma insulin, TG, LDL cholesterol and HDL cholesterol.
However, the total integrated postprandial plasma glucose response following the low glycaemic index breakfast and lunch was lower by 14% as compared to the high glycaemic meals.
Two other papers are often cited as showing the benefits of differences in the glycaemic index of CHO-rich foods: in one of these (40), only six patients with Type 2 diabetes were studied, and their data were not presented separately; while in the other paper the patients were only followed for two weeks (41).
196 NUTRITIONAL MANAGEMENT OF DIABETES MELLITUS
VARIATIONS IN FIBRE CONTENT In the most general sense, it is deemed useful for patients with Type 2 diabetes to increase their dietary fibre intake, particularly of soluble fibre.
At the same time, the clinical utility of this intervention is not clear.
For example, the recommendation of the American Diabetes Association is for patients to increase their dietary fibre intake to 20, 35 g/day, without clear evidence of the importance of this amount of fibre on glycaemic control (42).
Chandalia and associates (43) have pursued this issue further, and evaluated the metabolic effect of essentially doubling the daily fibre intake of 13 patients with Type 2 diabetes.
In this study they compared two diets, each containing *55% of daily calories as CHO, with one test diet having twice as much total fibre (50 vs 24 g/day), and threefold the soluble fibre content (25 vs 8 g/day).
The patients were studied at the end of two randomly assigned diet periods of six weeks in duration.
Even if it is assumed that patients will be willing to consume diets made up almost entirely of oranges (300 g/day), green peas (110 g/day), zucchini (195 g/day), papaya (250 g/day), peaches (300 g/day), fruit cocktail (200 g/day) and cherries (100 g/day), the improvement in day-long plasma glucose concentrations was no greater, and the decline in plasma TG concentration of lesser magnitude, than when the low fat, high CHO diet was compared to a diet in which MUFA was increased and CHO decreased (20).
When put into the context of results of earlier studies evaluating variations in fibre content, it appears that the clinical benefit is closely related to the increment in daily fibre intake.
More specifically, increases in fibre intake of *15 g/day did not lead to any significant metabolic changes (44,45), whereas dietary increases up to 23 g/day (46) and 30 g/day (47) resulted in a modest improvement in glycaemic control, without any decrease in plasma TG concentration.
Thus, there appears to be no study in which the untoward metabolic effects of CHO-enriched diets have been shown to be attenuated to the degree seen when MUFA/PUFA fat is used to replace SF (20), and the only instance in which the effects were even comparable involved the daily intake of 50 g fibre (43).
SUMMARY AND SUGGESTIONS
The most useful dietary intervention, by far, is the initiation of a calorie- restricted diet in patients with Type 2 diabetes who are overweight: the beneficial effects of weight loss on both glucose and lipid abnormalities in patients with Type 2 diabetes are both unequivocal and dramatic. Although there may be something to be gained by avoiding the use of low fat, high CHO diets in this context, the major benefits will depend upon the decrease in total calories, not in variations in kind of macronutrient.