Macronutrients and Weight Gain
The integrated impact of the differential regulation of macronutrient intake, digestion, absorption, storage and oxidation generally supports the hypothesis that dietary fat may be particularly associated with weight gain. This has been examined in a number of studies conducted over several days or weeks. Here the period of study must be related to the precision of the measurements of changes in macronutrient balance. Tightly controlled experimental studies performed in calorimeters, where changes in fat stores can be measured to + 9g fat/day, can be conducted over just a few days, whereas measurements made in free-living conditions, using in vivo body composition measurements require a period of several weeks or even months, since the precision is of the order of + 1kg fat (28). Community studies, which rely solely on changes in body weight as an index of changes in fat stores, must be conducted over several months in order to provide a reliable indicator of long-term changes in fat stores, rather than acute fluctuations caused by shifts in water or carbohydrate balance.
Other studies have investigated the regulation of macronutrient balance using hour-by-hour measurements of the subjects' self-selected intake and subsequent substrate oxidation whilst continuously confined to a whole body calorimeter. In one experiment six lean young healthy men were studied on three occasions in which each individual meal and snack was covertly manipulated to provide approximately 13% dietary energy as protein, 20, 40 or 60% energy as fat and the remainder as carbohydrate (29). Hence diets which were relatively high in fat were low in carbohydrate, with a high energy density and vice versa. On the 60% fat diet subjects were in marked positive energy and fat balance, but there was no significant change in carbohydrate balance. Body weight increased by 0.9 kg over 7 days. On the 20% fat diet energy and fat balance was negative, although there was a modest positive carbohydrate balance. Despite the negative energy balance subjects gained 0.2 kg over 7 days, due to the increase in glycogen stores.
In a subsequent study conducted under free-living conditions absolute energy balance was lower on each dietary treatment, probably reflecting the greater energy needs of subjects outside the confines of the calorimeter. However, the inter-treatment effect was remarkably similar (30). Here the 20% fat diet elicited spontaneous weight loss of 0.74 kg in 14 days, relative to a gain of 0.09 kg on the 60% fat diet, despite eating ad libitum. This is consistent with a study in women, who lost 0.4 kg in 2 weeks on a 15-20% fat diet, but gained 0.32 kg over a similar period on a diet providing 45-50% fat (31).
This phenomenon of 'high fat hyperphagia' is a plausible explanation for the association of high fat diets with obesity (see Chapter 10). It is a robust and readily reproducible effect that is observed across different groups of subjects and in different laboratories. Given the increasing prevalence of obesity, a number of strategies have been considered to counterbalance the effect of high-fat diets. In a public health context physical activity offers an attractive option which will also provide independent health benefits. In experimental studies subjects who are confined to a calorimeter with a sedentary protocol, readily over-eat on high fat diets relative to high carbohydrate diets. However, a recent study has shown that the imposition of 3 x 40 minute periods of cycling, designed to raise the physical activity level (TEE/BMR) from 1.25 to 1.61, resulted in a significant reduction in energy balance, relative to the high fat-sedentary protocol. This was attributable to both an increase in energy needs and a reduction in intake (32) (Figure 9.5). This interaction between macronutrient and physical activity was also observed in the epidemiological analysis of the Gothenburg (Goteborg) Women's Study where the risk of weight gain was significantly increased only in those women consuming a high fat diet who were also classified as sedentary (33).
In a clinical setting pharmacotherapy has been proposed to curb hyperphagia. In a study of six obese women offered either a 25 or 50% fat diet, along with either a centrally acting appetite suppressant (dexfenfluramine) or placebo, the drug reduced energy intake on the high fat diet relative to
the placebo by 10%, but it did not significantly decrease intake on the low fat diet. It is noteworthy, however, that the effect of decreasing the fat content of the diet was more than three times greater than the impact of the drug (34) (Figure 9.6).
One of the most convincing mechanisms to explain the phenomenon of high fat hyperphagia relates to the energy density of fat (39 kJ/g), which is disproportionately high relative to carbohydrate (16kJ/g) or protein (23kJ/g). In studies where the energy density of the diets was held constant in the face of macronutrient manipulations the high fat hyperphagia is frequently abolished. For example, in a crossover study using predominantly liquid diets of similar energy density but containing either 24 or 47% fat, there was no difference in energy intake over a 2-week period (35). In the studies of Stubbs and colleagues described above, it was apparent that subjects had consumed a similar bulk of food on each occasion, despite the macronutrient manipulation (29,30). Thus the hyperphagia observed on high fat diets may be viewed as 'passive over-consumption', i.e. increases in energy intake in the absence of increases in the volume of food consumed. This was investigated further in a study where the energy density of the diets was fixed, whilst still providing 20, 40 or 60% energy as fat (36). Here high fat hyperphagia was not observed, although it is notable that there was a persistent decrease in energy intake on the high carbohydrate/ low fat diet, suggesting that the suppressive effect of carbohydrate on food intake is over and above its effects on the energy density of the food.
The importance of the energy density of food as a determinant of energy intake has previously been reviewed in detail (37). Traditional high fat foodstuffs frequently have a higher energy density than low fat/high carbohydrate foods. However, the advent of new food processing methods means that an increasing number of items may be low in fat but retain an energy density comparable to the high fat variety. Their effects on body weight regulation are unclear. Likewise the role of non-caloric sweeteners (38) and fat substitutes in the relation of macronut-rient balance in children (39) and adults remains controversial with conflicting results obtained using different experimental paradigms (40-42).
The primary conclusion of the experimental studies described above is that diets containing a high proportion of fat disrupt the physiological processes that regulate macronutrient balance and frequently result in positive fat balance and weight gain. In contrast, in less controlled studies, especially in the community, it is the balance between physiological processes and external and/or cognitive influences which determines the overall effect. Here the effects of macronutrients on energy intake cannot be easily segregated from other factors controlling appetite (see Chapter 8). Accordingly it does not inevitably follow that low fat diets lead to weight loss.
There is currently a clear divide in scientific opinion regarding the merits of low fat/high carbohydrate diets in the context of weight loss (43,44). We conclude that although there is evidence of spontaneous weight loss associated with low fat diets relative to high fat control diets, most intervention studies have shown only a small reduction in body weight (up to 0.6 kg/month). Moreover this loss occurs mostly in the first 3-6 months, after which weight may be gradually regained. This precludes the use of low fat diets as the sole strategy for weight reduction.
Nonetheless it is pertinent to note that most ad libitum, low fat intervention studies currently in the literature were not primarily designed to examine the impact of macronutrient manipulations on body weight. Indeed most subjects were not overweight and may therefore be more likely to protect their body weight, through either physiological or cognitive processes. Some studies gave specific advice to subjects to increase carbohydrate intake to maintain body weight (45) or advised on other nonspecific strategies to equalize energy intake to control values (46,47). Some studies used a low fat diet as part of a broader management plan (48,49) including advice to stop smoking, which would tend to lead to weight gain. The notable exception is the study of Ornish, in which there is a decrease of 11.5 kg in weight (relative to controls) over 1 year (48). These patients were mostly overweight and angiographic evidence of coronary artery disease, which may have provided a very significant motivational factor to enhance compliance to the comprehensive management programme.
In studies of the treatment of obese patients there is little evidence of any macronutrient-specific effects on weight loss (50-53). Here the rate of weight loss is closely related to the energy deficit that is achieved, suggesting that the macronutrient effects are more subtle than gross differences in energy intake. However, low fat diets are an effective method to decrease total energy intake and two studies have shown that subjects in the low fat group perceived the diet to be more palatable and showed significant improvements in quality of life scores which may increase the likelihood of greater compliance in the longer term (50,51). Two studies have assessed the value of ad libitum low fat diets, relative to energy-restricted diets for weight maintenance over 1 year following acute weight loss. Schlundt found no significant difference between groups in terms of weight regain, while Toubro and Astrup demonstrated significantly enhanced weight maintenance in the low fat group (54,55).
These studies raise a number of general issues regarding the efficacy of low fat diets in community studies. Firstly, compliance to the low fat regime is obviously a prerequisite for effective weight loss. An individual's perception of their personal fat intake relative to the population average does not correlate well with actual fat intake (56). Indeed in one survey and when questioned, most people tend to believe that they eat less fat than the average person! Thus subjects who routinely incorporate some low fat products in their diet (e.g. low fat milks or spreads) may believe that they have reduced their fat intake, yet consciously or subconsciously compensate, by consuming additional fat in other items. In studies in the UK the majority of consumers report that in recent years they have reduced their personal fat intake (57), yet estimates of fat consumption from the National Food Survey have remained rather constant (58). In a study which incorporated independent estimates of compliance, using 13C-labelled glucose and subsequent measurement of i3C in expired air, there was a positive relationship between individual adherence to the low fat diet and the extent of weight loss (59). This confirms that different level of compliance is a major determinant of the success of low fat interventions.
Secondly, the actual change in dietary fat intake may be less than reported by dietary surveys. This is difficult to assess since self-reporting of food intake is notoriosly unreliable and dietary education may increase any bias in the reported macronutrient intake. Attempts to educate subjects how to consume less fat may also serve to educate subjects how to report less fat. Alternatively subjects may accurately report their intake on specific measurement days, when they may closely adhere to the dietary prescription, but this may be a poor reflection of their typical eating habits (60).
Finally, there is growing concern that subjects may reduce their fat intake but not necessarily reduce their total energy intake. Even in one of the most effective of the intervention studies where there was a striking reduction in dietary fat from 31.5 to 6.8%, this was accompanied by only a 0.6 MJ/day decrease in energy intake over one year (48). Apparently three-quarters of the decrease in fat was counterbalanced by increases in energy from other dietary constituents. This compensation may represent a physiological system which recognizes the fall in energy intake and endeavours to restore the status quo by stimulating consumption (26). Even if subjects adhere to the low fat prescription, weight loss may be attenuated by increases in other macronutrients. The overt nature of a study where subjects must self-select their own food (unlike the mostly covert laboratory manipulations) may trigger unpredictable cognitive responses, as observed in the preload/test meal paradigms (23). Covert manipulations may therefore be more effective in producing spontaneous weight loss because there is no obstructive cognitive response. However, the effect of reductions in dietary energy density on innate appetite control systems are insufficient to overcome total dietary disinhibition. The emphasis on low fat foods may contribute to a perception that such foods will not cause weight gain regardless of the amount consumed, thus liberating subjects from exerting any dietary restraint (61). Low fat foods which substitute for the high fat equivalent will tend to lead to a cut in energy intake, but the addition of low fat foods to the diet will simply increase energy intake.
The rather limited impact of reductions in the proportion of fat in the diet on body weight is in contrast with the effect of increases in protein, where the community data is generally stronger than derived from the laboratory studies described previously. In observational studies of food intake over a 9-day period, dietary protein was the most efficient macronutrient at suppressing subsequent intake, independent of its contribution to net energy intake (62). In a larger study of 160 women with 16 days of weighed food records over a 1-year period the energy consumed as protein was inversely related to total energy intake (63). Most convincingly, in a dietary intervention study for the treatment of obesity two groups were randomized to a 30% fat diet. A high protein group consumed 25% energy as protein and 45% energy as carbohydrate, whilst a high carbohydrate group consumed 12% as protein and 58% as carbohydrate. Over a 6-month period the mean weight loss in the high protein group was
8.9 kg compared to only 5.1 kg in the high carbohydrate group (P < 0.0001) (64). The principal mechanism of this effect appeared to be a greater reduction in overall energy intake in the high protein group, which would be consistent with increased satiety reported previously following high protein preloads. However, further research is required to assess the long-term impact of high protein diets on morbidity and mortality before high protein diets can be considered as a viable public health intervention for the control of macronutrient balance and body weight.
Was this article helpful?