The Journal of Nutrition Vol. 127 No. 9 September 1997, pp. 1875S-1883S
Copyright ©1997 by the American Society for Nutritional Sciences

Body Weight Set-Points: Determination and Adjustment¹

Richard E. Keesey² and Matt D. Hirvonen

Department of Psychology, University of Wisconsin, Madison, WI 53706

ABSTRACT

BODY WEIGHT: EVIDENCE FOR ITS REGULATION

COORDINATED CONTROL OF ENERGY INTAKE AND EXPENDITURE IN BODY WEIGHT

OBESITY AS A PHYSIOLOGICALLY REGULATED CONDITION

IS OBESITY IN HUMANS REGULATED?

SUMMARY

FOOTNOTES

LITERATURE CITED

ABSTRACT

It is proposed that body weight, like body water and body temperature, is physiologically regulated. In the case of body weight, coordinated adjustments in both the intake and expenditure of energy serve to stabilize the weights of individuals at a specified level and to resist their displacement from this level. Obese individuals also display these behavioral and metabolic adjustments to weight perturbations and thus appear to actively resist efforts to reduce their weight from the elevated levels they ordinarily display. Experimental studies of genetically transmitted and diet-induced forms of obesity in animals similarly suggest a view of obesity as a condition of body energy regulation at an elevated set-point. An individual's set-point for regulated body weight is apparently adjustable, shifting over a lifespan in conjunction with naturally occurring but still unspecified physiologic changes. Experimentally, the set-point for body weight can be adjusted by manipulation of specific hypothalamic sites. Lesions of the lateral hypothalamus, for example, cause a chronic reduction in the level at which laboratory animals regulate body weight. It thus appears that hypothalamic mechanisms play a primary role in setting the level at which individuals regulate body weight, and it is likely that the genetic, dietary and other lifespan influences on body weight are expressed through these mechanisms.

BODY WEIGHT: EVIDENCE FOR ITS REGULATION

Although weight stability is suggestive of homeostatic control, regulation implies the active defense of a particular physiologic condition. Thus, if weight is regulated, we should expect body energy perturbations to be met by compensatory adjustments in the intake and/or expenditure of energy. Indeed, such evidence for the active defense of body weight is seen in both the animal and human research literatures. Animals displaced experimentally from the body weights they ordinarily maintain are quick to restore weight to the usual level upon removal of the perturbing conditions. Shown in Figure 1, for example, are the body weights of a group of male rats whose growth was arrested by restricting their daily caloric intake for several weeks. Upon being permitted to feed freely, these rats quickly restored weight to a level appropriate to their age and gender. A similar rapid and precise restoration to a normal level is seen after a rat's body weight has been experimentally elevated (Steffens 1975).

The human research literature is replete with similar reports. Individuals whose caloric intake has been constrained as a result of famine or war, or those who diet, typically regain whatever weight they lost within a relatively short time after the intake restraints are removed (Keys et al. 1950). As can be seen in Figure 2, this is unfortunately true for the obese as well as for those of normal body weight (Johnson and Drenick 1977).

COORDINATED CONTROL OF ENERGY INTAKE AND EXPENDITURE IN BODY WEIGHT

Traditionally, food intake has been regarded as the key controlled factor in the process of regulating body weight. We know, for example, that compensatory adjustments in food intake occur when an individual's weight is displaced from the normally maintained level. If one's weight is reduced, food intake elevates. When the previously deprived rats depicted in Figure 1 were once again permitted to feed freely, they displayed supranormal levels of daily intake until their body weight was restored to the level of nonrestricted rats. Conversely, if one's weight is experimentally elevated, eating is sharply curtailed.

It is often the case, however, that control of intake is insufficient to account for the stability of body weight. Increases or declines in food intake frequently fail to produce the changes in body weight expected on the basis of the caloric excess or deficit. The weight losses produced by most diets, for example, are typically less than those expected from the apparent caloric deficit; similarly, overconsumption often fails to produce weight gains commensurate with the apparent caloric excess. In other cases, weight gains occur in the absence of hyperphagia, whereas weight loss can be seen without a reduction in intake (Levitsky et al. 1976). Effects of this sort would not occur unless energy expenditure were also undergoing adjustment.

Indeed, it can be shown that adjustments in expenditure accompany those in intake when body energy is perturbed. For example, when weight loss occurs, resting metabolism declines by an amount significantly in excess of that expected from the loss in metabolically active tissue. We have observed a drop of 24.6% in daily resting energy expenditure when the body weight of rats was reduced (by caloric restriction) by 14.9% (Corbett et al. 1985). A decline in resting energy expenditure disproportionately larger than the associated loss in body mass indicates that less energy is required to maintain a gram of tissue in an individual who is weight reduced rather than at the normally maintained body weight.

In a converse fashion, there is an exaggerated increase in energy expenditure when one overconsumes and body weight rises above the level normally maintained (Rothwell and Stock 1982). That is, the increase in daily energy expenditure with an elevation in body weight is considerably greater than that expected from the added body tissue. At least in some species, this increase in energy expenditure can be traced to the activation of a specific thermogenic organ, viz., the brown adipose tissue (Rothwell and Stock 1979).

What these observations demonstrate is that perturbations of body energy initiate a coordinated pattern of compensatory intake and expenditure adjustments. Weight declines from the normally maintained level produce increases in intake accompanied by decreases in daily energy expenditure, thereby blunting further weight loss while providing the conditions for a rapid restoration of the lost weight. Elevations in weight from the regulated level cause intake to be curtailed while rates of energy expenditure are enhanced, thereby forestalling further gain and facilitating weight loss. In this manner, the weight of an individual is stabilized at a particular level.

Max Kleiber, a noted animal nutritionist, studied the nature of the relationship between daily resting energy expenditure and body mass in various animal species, ranging in size from small birds and rodents to large mammals. The relationship Kleiber observed can be seen in Figure 3 (Kleiber 1975) in which the log of daily resting energy expenditure of each species is plotted as a function of the log of its body mass. The slope of this function, relating log daily energy expenditure (kcal/d) to log body mass is 0.75, indicating that daily energy expenditure increases at a rate three fourths that of body mass (kg) as species increase in size. The best-fit equation for this relationship between daily resting energy expenditure and body mass is kcal/d = kBW_kg^0.75, with 69 often cited as the value for k.

An important related issue is whether Kleiber's daily energy expenditure-body mass relationship, based upon interspecies comparisons, can account for the daily resting expenditures of different-sized members of the same species. Kleiber favored such an extension of his interspecies relationship, including an application of the 0.75 power exponent. Others, citing empirical and/or theoretical considerations, contend that the mass exponent of 0.66 is more appropriate for intraspecific comparisons (Donhoffer 1986). It is worth noting in this regard that the range of weight differences among members of the same species tends to be relatively small. Similarly, the energy expenditures predicted by body mass raised to the 0.66 or 0.75 power do not differ greatly within a restricted weight range. These circumstances render somewhat problematic a determination of whether the mass exponent for any particular interspecies comparison is the same or different from the interspecific value of 0.75. Nevertheless, there is no question that the daily resting expenditures of different-sized members of the same species are tightly related to the maintained level of body weight. In Figure 4, for example, the measured 24-h resting energy expenditures of 70 male rats, all of the same age but ranging in body weight from 340 to 430 g, have been expressed relative to the body weight each is naturally maintaining. Clearly, as with interspecies comparisons, body mass can account for a very significant portion of the variance in the daily resting expenditure of individual rats.

As an illustration of this point, consider once again the daily energy expenditure of the 70 rats depicted in Figure 4. Note that the daily energy expenditures of these different-sized rats tend to cluster around the best-fit line relating daily resting expenditure to body weight. Yet, if the body weight of any rat is displaced from the spontaneously maintained level, its resting expenditure would change by a greater amount than this best-fit function would predict. Note in Figure 4 that, when the intake of one of the average-sized rats was restricted so as to bring its weight to a lower level, its daily resting expenditure declined to an extent that it was substantially below the expenditure of rats naturally maintaining body weight at that level.

We might conclude from this example that the daily resting energy expenditure of each rat will be consonant with the value predicted from the body mass-expenditure function (Fig. 4) only when its body weight is at the physiologically preferred level or "set-point." The set-point for each rat can thus be taken as the particular body weight at which its daily energy expenditure is congruent with the value predicted by the best-fit function describing the body mass-expenditure relationship for that species. Thus, for each rat, there is at any point in its lifespan only one body weight at which its daily resting expenditure will be at the expected value.

OBESITY AS A PHYSIOLOGICALLY REGULATED CONDITION

Further consideration of the daily energy expenditure of the rats depicted in Figure 4 can provide a rationale for considering obesity as a physiologically regulated condition. Consider this time not rats of average size but rather the heavier of the 70 rats represented in Figure 4. Given their deviation from the average weight for age and sex, the rats found at the upper end of this distribution would, if not obese, certainly qualify as overweight. Yet, note that these heavier rats also metabolize energy at a rate conforming to the body mass-expenditure function. Note also, however, that they do so only when they remain heavier than the other rats. When the intake of one of these rats was restricted so as to lower its body weight to an average level, its resting energy expenditure dropped substantially below that of rats that naturally maintained body weight at average levels. This rat apparently metabolizes at a rate appropriate to its body mass only when at a higher body weight, suggesting that being heavy is as natural for this particular rat as being of average weight appeared to be for the one considered earlier. That is to say, "overweight" appears to be the natural state of this rat.

It could be argued, of course, that the rat used to illustrate this point was not obese and that frank obesity may not fit this pattern of seemingly normal regulation at elevated body weights. But, observations from clearly obese animals can also be shown to support a view of obesity as a condition of regulation at an elevated set-point.

Two forms of obesity in rats have served as animal models of human obesity. In one, the obesity is genetic in origin; in the other, the obesity can be traced to dietary influences. Both forms are marked by adipocyte hypertrophy and hyperplasia, conditions that similarly characterize obese humans.

Our understanding of the mechanisms responsible for the heritable components of obesity has been aided by studies of genetically obese strains of rodents. This understanding has been advanced by the recent identification of leptin, the product of a defective Lep gene in the obese mouse strain, Lep^ob/Lep^ob (Halaas et al. 1995). Injection of leptin has been shown to restore the weight of Lep^ob/Lep^ob mice to normal levels, thus raising the possibility that this metabolite may provide the signal the system for body weight regulation relies upon to index body energy status.

In the widely studied Zucker rat, obesity is transmitted as a single Mendelian recessive gene (fa) from the mating of heterozygous (Fa/fa) lean rats (Zucker and Zucker 1961). The development of obesity in the Zucker fa/fa rat results from a pattern in young rats of high levels of food intake coupled with lower than normal rates of energy expenditure. Yet, as weight-stable adults, the intake of obese Zucker rats is controlled in ways appropriate to sustaining their obesity. Similarly, adult obese Zucker rats appropriately adjust their daily energy expenditure so as to appear to defend their obesity. For example, when dieted and caused to drop their weight from the elevated levels they typically maintain, obese Zucker rats display the same sharp reduction in resting energy expenditure seen in normal weight rats following dietary restriction and weight loss (Keesey and Corbett 1990). To appreciate how substantial this expenditure adjustment can be in obese Zucker rats, consider the body weight and expenditure observations depicted in Figure 5. With their large body tissue mass, the daily caloric expenditure of unrestricted fa/fa rats is normally 26% higher than that of lean Fa/- littermates. But, with only a modest weight loss (from 623 to 583 g), the decline in the obese rats' metabolic rate is such that their daily expenditure is now comparable to that of unrestricted lean rats weighing only 285 g. That is, a diet-induced weight loss of 6% caused the daily energy needs of these obese Zuckers to decline to the level of lean Zucker rats weighing less than half as much! There clearly appears to be a strong metabolic resistance to weight loss in these obese rats. Were such a mechanism to operate in obese humans, their frequent claim that they eat the same or less than their lean friends, but lose no weight, must be given more credence than it is ordinarily accorded.

To summarize, a reduced rate of energy expenditure, coupled with overconsumption, predisposes developing Zucker fa/fa rats to the obese condition they display as adults. However, as adults, a reduced expenditure no longer conveys any energetic advantage to obese Zuckers. As seen in Figure 5, their daily expenditure (kcal/d) is actually higher than that of lean Zuckers. Other reports indicate that the energy that obese Zuckers expend on general activity is also greater, although normal when expressed as a percentage of total expenditure. When reduced from the elevated body weights they typically maintain, however, their intake increases and their metabolic rate declines, suggesting that they defend their elevated body weight in the same way (and, apparently, as effectively) as their lean littermates defend normal body weights.

A specific example of an apparently irreversible dietary influence on body weight is found in a study in which rats were maintained on a palatable, high fat diet for 6 mo (Corbett et al. 1986). As seen in Figure 6, the body weights of these rats became progressively higher than those of rats fed a standard laboratory diet. After 6 mo, the rats fed the high fat diet weighed 26% more than those eating a standard diet. Near the end of this 6-mo period, the intake of half of the now-obese rats and half of the normal-weight control rats was restricted so as to lower their body weights from the level maintained by the rats still fed either the high fat or regular diet ad libitum (see Fig. 6).

When body composition was subsequently analyzed, the rats fed the high fat diet were found to have more than double the mass of adipose tissue of rats fed a standard diet. Further analysis revealed that changes in both adipose cell size and cell number contributed to this increase in body fat. Average fat cell size in the high fat-fed rats increased by 61%, and the total number of adipocytes increased by 48%.

When the body weights of some of the rats were lowered (by restricting their intake) after 5 mo of exposure to the high fat diet, the adjustments in body tissues, just as during the gain phase, were seen largely in the adipose mass. However, although total body fat declined, the gain in adipocyte number that had occurred in the preceding 5 mo of high fat feeding did not reverse. Rather, the fat loss in these rats was achieved almost entirely by reducing the amount of lipid per adipocyte. Consistent with other reports, the diet-induced increases in fat cell number were apparently irreversible.

Just before the analyses of body composition, the daily resting energy expenditure of both the obese and normal-weight rats was assessed, both at the body weights each now spontaneously maintained and at the reduced weights brought about by the caloric restriction (Fig. 6). For the nonrestricted obese rats, we found resting energy expenditure, expressed relative to BW_kg^0.75, to be normal and appropriate to the larger tissue mass these rats maintained. The conditions responsible for producing the typical elevation in energy expenditure in the initial stages of diet-induced weight gain thus seemed not to persist over the 6-mo exposure to the high fat diet. Rather, the resting expenditure of the obese rats at this time was quite comparable to that of normal-weight rats [63.9 vs. 64.7 kcal/(d·BW_kg^0.75)].

Of further interest were the results obtained from the calorically restricted obese and normal-weight rats. Diet-induced weight loss in the normal-weight rats led to the expected decline in daily resting energy expenditure [from 64.7 to 55.2 kcal/(d·BW_kg^0.75)]. However, the obese rats showed a similar adjustment to weight loss [from 63.9 to 57.3 kcal/(d·BW_kg^0.75)]. In fact, the total daily caloric expenditure of weight-reduced obese rats actually declined to a level below that of the nonrestricted normal-weight rats (31.7 vs. 32.7 kcal/d) despite the fact that the obese rats, though weight-reduced, still weighed 53 g more.

The findings in dietary obese rats of 1 ) apparently normal rates of resting metabolism and 2 ) sharp declines in resting metabolism when body weight is lowered from these higher weight levels suggest that prolonged maintenance on high fat diets can elevate the body weight set-point. The finding that these changes are seemingly irreversible is consistent with the morphologic and physiologic changes that appear in animals chronically maintained on such weight-promoting diets. Among these are apparently nonreversible increases in adipocyte number (Faust et al. 1978), which can be detected some weeks after exposure to such diets. Another is the pattern of change in tissue norepinephrine (NE) turnover rates following exposure to weight-promoting diets. Although initially elevated, NE turnover rates decline after prolonged exposure to such diets (Levin et al. 1983) and apparently return to near normal levels after several months. Whether the changes in fat cell number and NE turnover simply covary with other internal adjustments crucial to elevating the level of regulated body weight or are themselves responsible for this diet-induced regulatory adjustment is not presently known. The outcome, however, appears to be a nonreversible elevation in the level at which body weight is regulated.

IS OBESITY IN HUMANS REGULATED?

It certainly is the case that obese persons, like individuals of normal weight, are capable of maintaining stable body weights. Of course, given the pressure exerted on obese people to lose weight, it would be surprising if their body weights were not somewhat more variable than normal. Yet, even when the obese are successful in shedding weight, their tendency to regain it subsequently is amply documented.

Speaking more directly to this issue is evidence that obese individuals, just as those of normal body weight, display metabolic adjustments to caloric restriction that act both to limit the loss of weight and to favor its recovery. Some of the most compelling observations bearing on this issue derive from the research of Leibel and Hirsch. In one study (Leibel and Hirsch 1984), it was found that daily calories required to maintain the body weight of dieted, obese patients declined by an amount disproportionate to the weight loss they displayed. The daily calories initially required to maintain the body weight of these obese patients was comparable to that of normal weight control subjects[(1432 vs. 1341 kcal/(m²·d), respectively]. However, after undergoing weight loss, the maintenance requirements of the obese dropped to 1,021 kcal/(m²·d). In fact, the total calories for maintenance of the weight-reduced obese patients was actually less than that of the control subjects (2171 vs. 2280 kcal/d), in spite of the fact that the former still weighed 60% more! Moreover, this enhanced metabolic efficiency of the weight-reduced obese was not a transient effect. Three formerly obese individuals who were successful in maintaining a reduced body weight for 4-6 y still displayed daily maintenance requirements of only 1031 kcal/(m²·d). A subsequent report (Leibel et al. 1995) both confirmed and extended these effects in weight-reduced obese patients.

There is evidence that adjustments in resting energy expenditure underlie, at least in part, the sharply reduced maintenance requirements of obese individuals who have undergone weight loss. One study (Garrow and Warwick 1978) measured the resting metabolic rates of 27 obese women both before and after a diet-induced weight loss. On the basis of the resting expenditure scores before the weight loss (open circles, Fig. 7), a best-fit line was derived to describe the usual relationship between resting metabolism and the body weight of these 27 women. The investigators then used this line (dashed line in Fig. 7) to assess the effects of weight loss on resting metabolism. It is evident that, just as with the weight-reduced rats in Fig. 4, weight loss produced a decline in resting energy expenditure in most obese women that was substantially in excess of that expected on the basis of the actual weight loss. Similarly, Bray (1969) noted that obese patients, who lost 3% of their initial body weight while eating a low calorie diet for 4 wk, displayed a decline of 17% in their resting rates of energy expenditure. In a similar fashion, Leibel et al. (1995) reported that the energy savings of weight-reduced obese patients were realized by reductions in both the resting and nonresting components of their energy expenditure. Thus, the obese, just as those of normal body weight, display metabolic adjustments to dieting that will substantially reduce the observed weight loss from the amount expected on the basis of the apparent calorie deficit.

It is to be expected that an individual with a regulated form of obesity will display natural resistance to diet-induced weight loss. Compensatory metabolic adjustments to caloric restriction will not only diminish initial weight loss but facilitate the restoration of weight previously lost. Sustained weight reduction, if achieved by dieting, will therefore require a lifelong commitment to a daily caloric intake not only less than satisfying, but possibly lower than that of individuals of normal body weight. Some observers (Wooley and Wooley 1984) have questioned whether such eating and other life-style adjustments, in conjunction with the possible nutritional inadequacies and depressed metabolism that chronic dieting can produce, might not be too great a price for the modest weight losses that most obese individuals are able to achieve and sustain.

For at least two reasons, this may be too pessimistic a view. If obesity is indeed a physiologically regulated phenomenon, it is understandable that a person attempting to lose weight by dieting will encounter physiologic resistance. Yet, it is important to recognize that the physiologic adjustments that form the basis of this resistance to weight loss can concurrently provide certain benefits. They can, for example, bring significant improvement in individuals suffering from obesity-related hypertension or hyperinsulinemia. Furthermore, as noted previously, even a modest decline from the body weight set-point can produce these benefits. This means that the obese need not reduce body weight to the normal range to realize significant benefits. A small decline in body weight from the regulated level should thus produce in the obese essentially the same physiologic adjustments, with their associated benefits, as it would in a normal-weight individual.

Second, one should not lose sight of an essential feature of regulatory systems, viz., that the system set-point can apparently be adjusted. Were means to become available to lower the level an individual's energy-regulating system was set to maintain, the very same physiologic adjustments that otherwise impede diet-induced weight loss could serve to facilitate and then sustain such weight changes. This is to suggest that future obesity research should have as an aim first, the identification of the factors responsible for setting the level at which such individuals regulate body weight and, second, the application of this knowledge to the design of procedures that would allow the system set-point to be readjusted. To this end, the final section examines ways in which alterations in the level of regulated body weight can be achieved and considers the central nervous mechanisms thought to be involved in mediating these adjustments.

There are reasons to believe that body weight changes produced by certain pharmacologic agents may exert this influence by altering the set-point for regulated body energy. "Anorectic" drugs such as fenfluramine offer one such example. Although it is generally assumed that these drugs act by directly suppressing appetite, some observations suggest an alternative explanation (Levitsky et al. 1981, Stunkard 1982). Anorectic drugs, for example, are effective in suppressing food intake only until body weight declines to a particular level. Intake then returns to essentially normal, although body weight remains at a reduced level. Traditionally, tolerance to the anorectic agent has been offered as the explanation for the drug's failure to continue suppressing food intake. However, an alternative explanation is that anorectic drugs produce their effects by lowering the body weight set-point. According to this interpretation, the initial food intake suppression with these drugs should be seen as secondary to lowering body weight to a new (reduced) level of regulation. Upon achieving this reduced regulation level, however, food intake is restored to levels appropriate to the stable maintenance of the new weight. The demonstration that fenfluramine fails to suppress food intake in rats whose body weight has been reduced to this new level before the start of drug administration certainly favors such an interpretation of this drug's mode of action. It might be noted in this regard that the increases in energy intake and the reductions in expenditure initially displayed upon smoking cessation, followed by the restoration of energy balance and subsequent stable maintenance of body weight at a higher body weight, has prompted the suggestion that nicotine similarly produces its effects upon body energy by lowering the body weight set-point (Schwid et al. 1992).

Lesions of the lateral hypothalamus (LH) produce what appears to be a downregulation in body energy (Powley and Keesey 1970). Rats with LH lesions chronically maintain body weight at a reduced percentage of normal (see Fig. 8). Furthermore, when challenged, they defend these reduced body weights. It can be seen in Figure 9 that, just as nonlesioned rats seen in Fig. 1 quickly restored body weight to its proper level after a period of food restriction and weight loss, so do LH-lesioned rats quickly restore weight to a reduced level of maintenance after a similar diet-induced weight decline (Mitchel and Keesey 1977). Also, just as nonlesioned rats restore their body weight to normal levels after it has been elevated by force-feeding, so do LH-lesioned rats quickly lower weight to a reduced level after being force-fed to normal levels (Keesey 1978).

The general pattern of daily energy expenditure in LH-lesioned rats provides further evidence that they are regulating body energy normally, though at a reduced set-point. In a study of energy balance, it was observed that the daily resting expenditure of LH rats was normal in that it was appropriate for their reduced body sizes. That is, their total daily resting expenditure was indistinguishable from that of nonlesioned, normal-weight rats when adjusted for their smaller body size (i.e., BW_kg^0.75). Thus, LH-lesioned rats do not appear to be displaced from their regulated or set-point energy level at the reduced body weights they maintain; rather, they display rates of resting metabolism similar to those of nonlesioned rats spontaneously maintaining body weight at this level. Normal levels of daily energy flux at a reduced body weight are indicative of a reduced set-point, not displacement from a higher set-point.

The results of other experiments add further support to this conclusion. The underlying rationale of these studies was that, if the set-point for body energy has indeed been reduced, LH-lesioned rats should display a metabolic resistance to being displaced from their lower body weights of the same sort that nonlesioned rats display upon being displaced from normal body weights. To this end, rats maintaining stable but reduced body weights following LH lesions were given either a highly palatable liquid diet or restricted amounts of the regular diet. The palatable diet produced substantial weight gain, restoring the weight of the lesioned rats to the levels of nonlesioned rats, whereas caloric restriction produced a further 12% weight loss. Similar weight gains or losses were produced by these dietary procedures in nonlesioned rats.

Data from these experiments (Corbett et al. 1985) reconfirmed that LH-lesioned rats maintaining stable but reduced body weights display daily rates of energy expenditure that are normal and virtually the same as those of nonlesioned rats [kcal/d = 69.4 vs. 70.3 kcal/(d·BW_kg^0.75)]. However, elevating the body weight of LH rats to the level maintained by nonlesioned animals caused their daily expenditure to rise substantially to 77.5 kcal/(d·BW_kg^0.75). Thus, restoring the weight of LH-lesioned rats to normal levels caused them to become hypermetabolic, just as nonlesioned rats are hypermetabolic when they overeat and gain weight (Rothwell and Stock 1982).

In a converse fashion, lowering the body weight of LH-lesioned rats from the reduced levels they already maintain caused a sharp decline in daily expenditure. Nonlesioned rats displayed virtually the same adjustment in expenditure when their weight was reduced to the level the LH-lesioned rats spontaneously maintained. The difference, of course, is that the LH-lesioned rats displayed this response only when their weight declined from its already reduced level of maintenance.

The preceding observations support the view that lateral hypothalamic lesions cause a chronic down-regulation in body energy. By displaying 1 ) a normal expenditure of energy at a reduced body weight, 2 ) a disproportionately lower than expected energy expenditure when their weight declines from this reduced level, and 3 ) a disproportionately higher than expected heat production when raised to the body weights of nonlesioned controls, rats with lesions of this hypothalamic area give every indication of regulating body energy normally, but at a reduced set-point. Clearly, this region of the hypothalamus plays a role in determining the level at which an individual's body weight is physiologically regulated.

SUMMARY

If the goal is substantial and sustainable weight loss in the obese, it is proposed that a more promising approach would be one based upon a strategy of directly altering the set-point of the energy-regulating system. Were it possible to lower the system set-point, the physiologic adjustments that ordinarily act to resist weight change (such as when the system is perturbed by dieting) would instead facilitate the achievement and subsequent maintenance of a lower weight. The last sections of the paper thus considered instances in which the level of regulated body weight is altered as the result of naturally occurring but unspecified physiologic changes. This was followed by a consideration of the hypothalamic mechanisms thought to be responsible for setting the levels at which body energy is regulated. Experiments were described demonstrating that the set-point for regulated body energy can be chronically lowered by direct manipulation of these hypothalamic mechanisms. While such experimental procedures for reducing the body weight set-point are not suitable for obesity treatment, it is proposed that a better understanding of mechanisms by which they bring about these adjustments can point the way to safe and effective means of lowering the level an individual's energy-regulating system is set to maintain.

FOOTNOTES

LITERATURE CITED

• Bray G. A. Effect of caloric restriction on energy expenditure in obese patients. Lancet 1969; 2:397-398 [Medline][Medline]
• Corbett S. W., Stern J. S., Keesey R. E. Energy expenditure in rats with diet-induced obesity. Am. J. Clin. Nutr. 1986; 44:173-180 [Medline][Abstract/Free Full Text]
• Corbett S. W., Wilterdink E. J., Keesey R. E. Resting oxygen consumption in over- and underfed rats with lateral hypothalamic lesions. Physiol. & Behav. 1985; 35:971-977 [Medline][Medline]
• Donhoffer S. Body size and metabolic rate exponent and coefficient of the allometric equation: the role of units. J. Theor. Biol. 1986; 119:125-137 [Medline][Medline]
• Faust, I. M., Johnson, P. R., Stern, J. S. & Hirsch, J. (1978) Diet-induced adipocyte number increase in adult rats: a new model of obesity. Am. J. Physiol. 235: E279-E286.
• Garrow, J. S. & Warwick, P. M. (l978) Diet and obesity. In: Diet of Man: Needs and Wants (Yudkin, J., ed.), pp. 127-144. Applied Science Publishers, London, UK.
• Goodner C., Oglive J. Homeostasis of body weight in a diabetes clinic population. Diabetes 1974; 23:318-326 [Medline][Medline]
• Halaas, J. C., Coajiwala, K. S., Maffer, M., Cohen, S. L., Chait, B. T., Rabinowitz, D., Lallone, R. L., Burley, S. K. & Friedman, J. M. (1995) Weight-reducing effects of the plasma protein encoded by the obese gene. Science (Washington, DC) 269: 543-546.
• Johnson D., Drenick E. J. Therapeutic fasting in morbid obesity: long-term fellow-up. Arch. Intern. Med. 1977; 137:1381-1382 [Medline][Abstract/Free Full Text]
• Keesey R. E. Set-points and body weight regulation. Symposium on Obesity: Basic Mechanisms and Treatment. The Psychiatric Clinics of North America 1978; 1:523-543
• Keesey R. E., Corbett S. W. Adjustments in daily energy expenditure to caloric restriction and weight loss by adult obese and lean Zucker rats. Int. J. Obes. 1990; 14:1079-1084 [Medline][Medline]
• Keys, A., Brozek, J. & Henschel, A. (1950) The Biology of Human Starvation. University of Minnesota Press, Minneapolis, MN.
• Khosha T., Billewicz W. Z. Measurement of changes in body weight. Br. J. Nutr. 1964; 18:227-239[Medline]
• Kleiber, M. (1975) The Fire of Life: An Introduction to Animal Energetics. Robert E. Krieger Co., New York, NY.
• Leibel R. L., Hirsch J. Diminished energy requirements in reduced-obese patients. Metabolism 1984; 2:164-170
• Leibel R. L., Rosenbaum M., Hirsch J. Changes in energy expenditure resulting from altered body weight. N. Engl. J. Med. 1995; 332:621-628 [Medline][Abstract/Free Full Text]
• Levin, B. E., Triscari, J. & Sullivan, A. C. (1983) Altered sympathetic activity during development of diet-induced obesity in rats. Am. J. Physiol. 244: R347-R255.
• Levitsky D. A., Faust I., Glassman M. The ingestion of food and the recovery of body weight following fasting in the naive rat. Physiol. & Behav. 1976; 17:575-580 [Medline][Medline]
• Levitsky D., Strupp B., Lupoll J. H. Tolerance to anoretic drugs: pharmacological or artificial. Pharmacol. Biochem. Behav. 1981; 14:661-667 [Medline][Medline]
• Mitchel J. S., Keesey R. E. Defense of a lowered weight maintenance level by lateral hypothalamically lesioned rats: evidence from a restriction-refeeding regimen. Physiol. & Behav. 1977; 18:1121-1125 [Medline][Medline]
• Powley T. L., Keesey R. E. Relationship of body weight to the lateral hypothalamic feeding syndrome. J. Comp. Physiol. Psychol. 1970; 70:25-36 [Medline][Medline]
• Rolls B. J., Rowe E. A., Turner R. C. Persistent obesity in rats following a period of consumption of a mixed high energy diet. J. Physiol. (London) 1980; 298:415-427 [Medline][Abstract/Free Full Text]
• Rothwell N. J., Stock M. J. A role for brown adipose tissue in diet-induced thermogenesis. Nature (Lond.) 1979; 281:31-35 [Medline][Medline]
• Rothwell N. J., Stock M. J. Energy expenditure of 'cafeteria'-fed rats determined from measurements of energy balance and indirect calorimetry. J. Physiol. 1982; 328:371-377[Abstract/Free Full Text]
• Schwid S. R., Hirvonen M. D., Keesey R. E. Nicotine effects on body weight: a regulatory perspective. Am. J. Clin. Nutr. 1992; 55:878-884 [Medline][Abstract/Free Full Text]
• Sclafani, A. (l993) Dietary obesity. In: Obesity: Theory and Therapy (Stunkard, A. J. & Wadden, T. A., eds.), pp. 125-136. Raven Press, New York, NY.
• Steffens A. B. Influence of reversible obesity on eating behavior, blood glucose, and insulin in the rat. Am. J. Physiol. 1975; 228:1738-1744 [Medline][Abstract/Free Full Text]
• Stunkard A. J. Anorectic agents lower a body weight set-point. Life Sci. 1982; 30:2043-2055 [Medline][Medline]
• Stunkard, A. J. (1991) Genes, environment and human obesity. In: Progress in Obesity Research 1990 (Oomura, Y., Tarui, S., Inoue, S. & Shimazu, T., eds.), pp. 669-674. John Libby, London, UK.
• Wooley, S. C. & Wooley, O. W. (1984) Should obesity be treated at all? In: Eating and Its Disorders (Stunkard, A. J. & Stellar, E., eds.), pp. 185-195. Raven Press, New York, NY.
• Zucker L. M., Zucker T. F. Fat, a new mutation in the rat. J. Heredity 1961; 52:275-278

0022-3166/97 $3.00 ©1997 American Society for Nutritional Sciences