- Email: [email protected]
The origins of Western obesity: a role for animal protein?
Medical Hypotheses (2000) 54(3), 488–494 © 2000 Harcourt Publishers Ltd DOI: 10.1054/mehy.1999.0882, available online at http://www.idealibrary.com on
The origins of Western obesity: a role for animal protein? M. F. McCarty Helicon Foundation, San Diego, CA, USA
Summary A reduced propensity to oxidize fat, as indicated by a relatively high fasting respiratory quotient, is a major risk factor for weight gain. Increased insulin secretion works in various ways to impede fat oxidation and promote fat storage. The substantial ‘spontaneous’ weight loss often seen with very-low-fat dietary regimens may reflect not only a reduced rate of fat ingestion, but also an improved insulin sensitivity of skeletal muscle that down-regulates insulin secretion. Reduction of diurnal insulin secretion may also play a role in the fat loss often achieved with exercise training, low-glycemic-index diets, supplementation with soluble fiber or chromium, low-carbohydrate regimens, and biguanide therapy. The exceptional leanness of vegan cultures may reflect an additional factor – the absence of animal protein. Although dietary protein by itself provokes relatively little insulin release, it can markedly potentiate the insulin response to co-ingested carbohydrate; Western meals typically unite starchy foods with an animal proteinbased main course. Thus, postprandial insulin secretion may be reduced by either avoiding animal protein, or segregating it in low-carbohydrate meals; the latter practice is a feature of fad diets stressing ‘food combining’. Vegan diets tend to be relatively low in protein, legume protein may be slowly absorbed, and, as compared to animal protein, isolated soy protein provokes a greater release of glucagon, an enhancer of fat oxidation. The low insulin response to rice may mirror its low protein content. Minimizing diurnal insulin secretion in the context of a low fat intake may represent an effective strategy for achieving and maintaining leanness. © 2000 Harcourt Publishers Ltd
INSULIN ACTIVITY AS A DETERMINANT OF RESPIRATORY QUOTIENT In at least two prospective studies, a relatively high fasting respiratory quotient (RQ) at baseline (after standardized diets) has been shown to be a strong risk factor for the subsequent development of obesity (1,2). Analogously, rats which are resistant to the development of obesity on a high-fat diet have a lower 24-hour RQ than obesity-prone rats (3). It is hardly surprising that a reduced propensity to oxidize fat should be associated with a tendency to increase fat stores. Insulin activity opposes fat oxidation at several levels (4). In adipocytes, insulin promotes storage and retention of fat,
Received 6 April 1999 Accepted 14 April 1999 Correspondence to: Mark F. McCarty MD, NutriGuard Research, 1051 Hermes Avenue, Encinitas, CA 92024, USA
488
thereby limiting the availability of free fatty acids (FFA) and serum triglycerides to other tissues. Insulin enhances synthesis of adipocyte lipoprotein lipase (LPL) at the transcriptional level (5,6), inhibits catecholamine-stimulated lipolysis (by activating a cAMP phosphodiesterase), and promotes FFA re-esterification by increasing adipocyte glucose uptake. In hepatocytes, largely by opposing glucagon-mediated increases in cAMP (7), insulin impedes FFA oxidation and ketogenesis (8,9). In skeletal muscle, insulin’s suppression of FFA oxidation may mainly reflect increased availability of glucose as a competitive fuel (4). In lean subjects, insulin has a quantitatively modest suppressive effect on skeletal muscle LPL activity, whereas it produces a paradoxical increase of this activity in obese subjects (10). Insulin secretion can thus be expected to have a major regulatory impact on RQ. An increase in insulin secretion (or a reduction in insulin clearance) should raise RQ – so long as it does not precipitate hypoglycemia. Measures which selectively promote the efficiency of insulin-stimulated glucose uptake and storage in skeletal muscle (such
The origins of Western obesity
as exercise training, very-low-fat diets, and perhaps bioavailable chromium) (11–14) typically down-regulate insulin secretion (14,15), this may play a role in their documented efficacy for reducing body fat (16–20). A lowglycemic-index diet likewise should be favorable to fat oxidation owing to decreased insulin secretion (21), this dampening of the insulin response reflects not only a decrease in postprandial glycemic excursions, but also a decreased intestinal production of the incretin GIP, a potentiator of glucose-triggered insulin secretion (22). Weight loss associated with soluble fiber supplementation (23) (which slows the digestion of food starch) may thus be largely attributable to a decrease in diurnal insulin secretion. (In some prospective studies, efficient insulin-stimulated glucose uptake has been found to be predictive of subsequent weight gain (4,24,25); this seems paradoxical in light of the foregoing comments. It seems likely that, in the populations examined, good insulin sensitivity of skeletal muscle was not principally a reflection of regular exercise training or very-low-fat diets, but rather served as a marker for insulin-sensitive adipocytes. A common cause of insulin resistance in skeletal muscle is an increased FFA flux secondary to adipocyte insulin resistance (26). As adipocytes hypertrophy, internal mechanisms such as increased production of tumor necrosis factor diminish adipocyte sensitivity to insulin (27,28); this tends to lower the RQ by increasing FFA availability to other tissues – thus preventing a further increase in obesity (27) – but has the further unfortunate consequence of inducing insulin resistance in skeletal muscle. Thus, insulin sensitivity of adipocytes may be the genuine risk factor for weight gain; measures which enhance insulin sensitivity of skeletal muscle without increasing that of adipocytes may in fact promote weight loss by reducing insulin secretion.) A role for excessive beta cell responsiveness in the induction and maintenance of obesity in rodents is suggested by recent studies with the drug RO23-7637 (28). Two days of pre-incubation with this agent in vitro decreased the elevated insulin secretory response to glucose of beta cells obtained from obese rats; oral administration of this drug to obese rats for seven days dose-dependently decreased insulin levels both basally and during i.v. glucose tolerance tests. Chronic ingestion of RO23-7637 induced weight loss or impeded weight gain in Zucker obese rats as well as in rats with dietinduced obesity; reduced food intake played a role in this effect, but treated rats typically lost more weight than pair-fed untreated rats. Whether or not this drug attains clinical utility, these findings provide confirmation for the principle that reduction of diurnal insulin secretion – whether achieved by natural or pharmaceutical measures – may promote leanness. (However, as a caveat, it should © 2000 Harcourt Publishers Ltd
489
be noted that insulin-stimulated de novo lipogenesis is far more significant in rats than in humans.) VEGAN DIETS PROMOTE LEANNESS It is well known that, in comparison to Western populations, the average BMI and prevalence of obesity in Asian cultures consuming low-fat, predominantly vegan traditional diets is low. Recent studies in rural China indicate that neither low calorie consumption nor high physical activity can adequately account for this phenomenon (29). For example, the average BMI of rural Chinese males is 20.5, despite a daily calorie intake that, after downward adjustment to approximate that of men doing light office work, is still 30% higher per unit weight than that of the average American male (where BMI is 25.5). Inasmuch as obesity is not rare in Asian-Americans (30), it is unlikely that genetic factors contribute importantly to the rarity of obesity in rural Asians. Substantial (and often unintended) weight loss is typically seen when Western patients enter treatment programs that stress very-low-fat predominantly vegan diets and moderate daily exercise. During the first year of Ornish’s lifestyle regimen for reversing coronary disease, weight loss averages 25 pounds – despite the fact that most of the patients are not obese, are not explicitly attempting to lose weight, and are not calorically restricted (19). The Pritikin clinics report similar results, as do vegan health spas in Europe (15,31,32). The moderate exercise component of these regimens appears unlikely to account for the dramatic and often rapid weight loss experienced. Cross-sectional studies note that vegans tend to be substantially lighter than either lacto-ovo vegetarians or omnivores (33,34). For example, a survey of macrobiotic vegans in the Boston area found that, on average, they were 15 kg lighter than age/sex-matched controls chosen randomly from offspring of the Framingham cohort (34). Carter states that, on average, vegan adults are 20 pounds lighter than non-vegans of comparable age (35); this estimate accords well with the amount of weight loss seen during vegan therapy regimens. It is commonly assumed that a very low fat intake can account for these observations. This idea has particular intuitive appeal in light of evidence that, under usual circumstances, de novo lipogenesis is minimal in humans – implying that body fat derives predominantly from dietary fat (36). In general, cross-sectional studies demonstrate a correlation between relative weight and the percentage of fat calories in habitual diets (18,37). It is therefore surprising that, within Western societies, ecologic studies fail to note a correlation between habitual fat intake and prevalence of obesity (37,38). Furthermore, although efforts to treat obesity by initiating moderate Medical Hypotheses (2000) 54(3), 488–494
490
McCarty
reductions in dietary fat (reducing fat to 20–25% of calories) typically achieve significant weight loss, the magnitude of the average long-term weight loss on these regimens is often less than impressive (38). Such findings encourage skepticism regarding the role of dietary fat in the Western epidemic of obesity. However, the fat contents of the traditional diets of lean societies, and of very-low-fat therapeutic diets, generally fall in the 8–15% range. This suggests the possibility of a threshold effect – that very-low-fat diets do indeed promote leanness, but that, past some threshold level, percentage fat intake does not have a major impact on body composition; presumably the common ambient variations of fat intake in Western societies (25–45%) lie above this threshold. Such a phenomenon is noted in squirrel monkeys – they remain lean on 13% fat diets, but become obese on diets with twice this level of fat (39). Insulin sensitization might offer a mechanistic basis for such a threshold. Very-low-fat diets, with or without exercise training, are reported to notably enhance the efficiency of insulin-stimulated glucose uptake (13,40,41); this entails a compensatory down-regulation of insulin secretion that is most substantial in subjects who are hyper-insulinemic at baseline (13,15). There is no evidence that such diets (at least in the short-term, prior to substantial weight loss) directly sensitize adipocytes or the liver to insulin (though it seems likely that insulin receptor expression will be up-regulated somewhat in response to decreased insulinemia). Nor is there any evidence that diets with 20–30% fat calories produce clinically useful insulin sensitization relative to higher-fat diets (42). Thus, reduced diurnal insulin secretion may play a significant role in the promotion of leanness by very-low-fat diets – in conjunction with a minimization of fat ingestion (43). When dietary fat is manipulated within the range common in Western society, a moderate reduction in fat intake will be accompanied by a corresponding increase in carbohydrate (and/or protein) intake that increases insulin secretion and in addition impedes fat oxidation through substrate competition; thus, the long-term impact on body composition is likely to be minor. However, when dietary fat intake is slashed from, say, 30%–10% of calories, the down-regulation of insulin secretion consequent to improved insulin sensitivity at least partially compensates for the tendency of increased carbohydrate/protein ingestion to raise RQ – while fat ingestion is cut by two-thirds! Under these circumstances, dynamic fat loss can be expected until the decline in fat stores induces a sufficient increase in RQ. However, whether this offers a complete explanation for the leanness of vegan societies or of American vegans is questionable. Willett emphasizes that obesity is rare in predominantly vegan Chinese provinces whose fat Medical Hypotheses (2000) 54(3), 488–494
consumption is as high as 25% (38). Furthermore, the average fat content of American vegan diets is as high as 30% in some studies (owing to prominent use of cooking or salad oils, nuts or nut butters, and olives and avocados) (44). Some vegan societies or treatment programs emphasize whole, fiber-rich starchy foods that would be expected to have lower glycemic indices than more refined foods. While this factor undoubtedly contributes to the various health benefits of these regimens – including reductions in body weight – it should be noted that the ‘sticky’ white rice which is the staple of many oriental rural diets is not notably low in glycemic index (45). (Parboiled long-grain rice often has a relatively low glycemic index, but this is not the type of rice prevalent in China or Japan (46).) PROTEIN POTENTIATES INSULIN RESPONSE Ð A ROLE IN WESTERN OBESITY? The possible role of a relatively low dietary protein intake – and an absence of ‘high-quality’ animal protein – in the relative leanness of vegans, has so far received little attention. Perhaps this reflects the fact that protein-rich foods are generally viewed as appropriate for dieters, and that low-starch high-protein diet regimens have achieved considerable popularity as strategies for promoting weight loss. The proponents of such regimens point to the fact that protein is far less effective than starch in evoking insulin release, and thus is compatible with the maintenance of a relatively low RQ favorable to fat loss; the insulin excursion above baseline (the ‘insulin index’) after ingestion of pure protein is only about 20% as great as that seen after a comparable caloric load of starch (47). These considerations, however, may not be very relevant to typical ‘balanced’ Western meals in which ample intakes of carbohydrate are combined with significant amounts of animal protein. Although carbohydrate is the most substantial source of calories in most Western diets, major meals traditionally contain a slab of animal flesh that is considered the ‘main course’; dairy products and eggs are also common components of mixed meals. Sandwiches, the staple fare of fast-food restaurants and lunch boxes, usually combine animal flesh with a highglycemic-index starch (wheat bread). A number of reports indicate that co-ingestion of protein markedly potentiates the insulinemic response to starch (48–54). Thus, a bolus of protein which, eaten alone, would produce only a modest insulin excursion, can provoke a substantial increase in insulin output when eaten with a significant quantity of carbohydrate. Most, though not all (54), relevant studies conclude that this potentiation is synergistic rather than merely additive. However, most of these studies use protein doses, © 2000 Harcourt Publishers Ltd
The origins of Western obesity
e.g. (50–60 grams) higher than would be consumed in most meals. A very recent crossover study has assessed the impact of adding a realistic amount of animal protein to a relatively low protein vegetarian regimen (55). The baseline daily diet provided 282 g carbohydrate, 50 g protein, and 101 g fat; the higher protein diet was identical save for the substitution of 32 g egg protein for 14 g of fat. Diurnal insulin secretion was estimated on these regimens by measuring 24-hour urinary C-peptide. The findings were dramatic – on average, C-peptide excretion was 60% higher during the higher protein diet. Perhaps a protein-mediated modulation of renal function contributed in part to the observed increase in C-peptide excretion – but this possibility seems unlikely to explain the full magnitude of the increase. If this finding is replicable it suggests that the relatively high protein content of typical Western diets (as compared to vegan diets) has an important impact on diurnal insulin secretion, and thus may be a significant factor in the Western epidemic of obesity. High-protein diet programs place great emphasis on the avoidance of starchy foods and sugars (low-calorie salads and vegetables are allowed). At the other end of the spectrum, the regimens of Pritikin and Ornish, while carbohydrate-rich, totally proscribe animal flesh and animal fat (in an effort to avoid all dietary cholesterol) and recommend whole starchy foods, many of which have relatively low glycemic indices. While these programs appear to be diametrical opposites, they should each achieve the common goal of reducing diurnal insulin secretion – thus favoring the oxidation of stored fat. A ‘take-home lesson’ from these considerations would be the following: do not combine high-protein animal products with starchy foods in the same meal. Remarkably, this is a cardinal tenet of ‘food combining’ dogma that features prominently in certain popular diet programs (such as that popularized by the bestseller Fit for Life) (56). The proponents of these diet strategies justify them with rationales that, within the context of modern physiology, are frankly laughable. Such reservations, however, have no bearing on whether or not the recommended regimens actually do promote fat loss. (In the case of Fit for Life, the proscription on eating anything other than fruit prior to noon, the emphasis placed on vegan foods, and the restriction of animal flesh, cheese or eggs to evening meals devoid of starchy foods, would seem likely to promote the promised fat loss.) An interesting test of these ideas would be a crossover study comparing diets composed of identical foods combined in different ways, i.e. one diet in which animal protein was included in every meal, as compared to a second diet in which all animal protein was segregated in lowstarch meals or snacks. Certain vegan foods – notably legumes – are relatively protein-rich. There is also a trend toward use of soy protein © 2000 Harcourt Publishers Ltd
491
isolate in foods intended as more healthful substitutes for animal products. In regard to legumes, it is pleasing to observe that, at least when consumed intact, their ‘insulin index’ appears to parallel their very low glycemic index (57,58); could slow digestion of legume protein play a role in this? Studies are needed to address the insulin response to meals in which legumes are combined with starches of higher glycemic index, such as rice. As to isolated soy protein, it appears to be less effective for releasing insulin and more effective for releasing glucagon (which promotes hepatic lipid oxidation and ketogenesis) than certain animal proteins (59–62); this phenomenon may contribute to the many favorable effects of soy protein ingestion (as a substitute for animal protein) noted in clinical and animal studies – including reduced weight gain recently reported in obesity-prone rats (62–64). It has been suggested that, as a general rule, the essential amino acids provided more abundantly by ‘high-quality’ animal protein have greater efficacy for releasing insulin, whereas the non-essential amino acids which predominate in ‘lower quality’ vegan proteins have a greater impact on glucagon (62,63). However, while the substitution of soy protein for animal protein seems likely to be beneficial, what is the effect of adding substantial amounts of soy protein to relatively low protein vegan diets? Clearly, the role of vegan protein in modulating the insulinemic response to vegan diets deserves further attention. Even when the glycemic index of rice approaches that of wheat flour products, the insulinemic response to rice is low in comparison to wheat (45). This has been offered as an explanation for the recent observation that serum levels of sex hormone binding globulin (a protein whose hepatic synthesis is suppressed by insulin) (65) tend to be higher in Chinese provinces where rice (rather than wheat or millet) is the staple food (66). Rice has a notably low protein content – 7% of calories, only about half that of wheat; perhaps this plays a role in the lower insulinemic response to rice, as well as in the low incidence of obesity and of ‘Western’ degenerative disease in Asian societies where rice is the predominant source of calories. An interesting peculiarity of pasta is that, while its glycemic index is lower than that of wheat bread (especially if lightly cooked), its ‘insulin index’ is disproportionately lower – only about 40% that of bread (45,67). If we assume that the protein content of wheat has a significant impact on its ability to evoke insulin secretion (as suggested by the fact that bread has a relatively high insulin index equivalent to that of glucose), then the very low insulin response to pasta may reflect the interaction of two phenomena – delayed absorption of starch and delayed absorption of protein. In any case, habitual consumption of lightly cooked pasta – in preference to other wheat flour products – can be expected to have a favorable impact on weight control. Medical Hypotheses (2000) 54(3), 488–494
492
McCarty
A further question which should be addressed is whether significant intakes of animal protein, even if consumed in meals devoid of starch, might increase fasting insulin secretion by raising fasting levels of ‘insulinogenic’ amino acids (59). When consumed in starch-free meals, protein has the potential to promote hepatic fat oxidation by stimulating glucagon production. Presumably, soy protein isolate, with its greater impact on glucagon, would be most appropriate for this application.
TOWARD A ÔLOW-INSULIN LIFESTYLEÕ An overview of the foregoing discussion can provide insight into the origins of Western obesity. Much of the dietary carbohydrate in Western diets is relatively high glycemic index – wheat flour products (other than pasta), potatoes, sucrose or dextrose. The carbohydrate is typically combined in meals with significant amounts of animal protein, which potentiate the postprandial insulin response. Such diets are almost invariably high enough in fat to impair insulin sensitivity of skeletal muscle (thus up-regulating insulin secretion) and to provide a significant daily load of absorbed fat calories. The resulting ample diurnal insulin secretion impedes the efficiency of fat oxidation and promotes fat storage – and doubtless also plays a crucial role in the genesis of the degenerative diseases characteristic of Western society (63,68). In opposition to this, one may propose a ‘low-insulin lifestyle’ to promote leanness and long-term health. Down-regulation of insulin secretion can be achieved by aiding the efficiency of insulin-mediated glucose storage through exercise training, very-low-fat eating, and supplemental bioactive chromium. Restraining gluconeogenesis (without inhibiting hepatic lipid oxidation) may also be useful in this regard; this may be the clinical basis of metformin’s efficacy (69–71) (for promoting both weight loss and diabetic control), and high-dose biotin may likewise have this effect (72,73). The prandial stimulus to insulin release can be minimized by choosing lowglycemic-index starchy foods and fruit, by avoiding the co-ingestion of animal protein and significant quantities of starch, and – to the extent feasible – by moderating caloric intakes. Adjunctive measures which should aid fat oxidation include prolonged moderate-intensity aerobic exercise during the post-absorptive phase (74,75), and disinhibition of hepatic fatty acid oxidation with hydroxycitrate (76). By achieving an adequately low diurnal RQ in the context of a low rate of fat ingestion, these measures should produce a steady negative fat balance until a new equilibrium is reached at a much leaner body composition.
Medical Hypotheses (2000) 54(3), 488–494
REFERENCES 1. Zurlo F., Lillioja S., Esposito-Del Puente A. et al. Low ratio of fat to carbohydrate oxidation as predictor of weight gain: study 24RQ. Am J Physiol 1990; 259: E650–E657. 2. Seidell J. C., Muller D. C., Sorkin J. D., Andres R. Fasting respiratory exchange ratio and resting metabolic rate as predictors of weight gain: the Baltimore Longitudinal Study on Aging. Int J Obesity 1992; 16: 667–674. 3. Chang S., Graham B., Yakubu F. et al. Metabolic differences between obesity-prone and obesity-resistant rats. Am J Physiol 1990; 259: R1103–R1110. 4. Eckel R. H. Insulin resistance: an adaptation for weight maintenance. Lancet 1992; 340: 1452–1453. 5. Farese R. V. Jr, Yost T. J., Eckel R. H. Tissue-specific regulation of lipoprotein lipase activity by insulin/glucose in normal-weight humans. Metabolism 1991; 40: 214–216. 6. Fried S. K., Russell C. D., Grauso N. L., Brolin R. E. Lipoprotein lipase regulation by insulin and glucocorticoid in subcutaneous and omental adipose tissues of obese women and men. J Clin Invest 1993; 92: 2191–2198. 7. Houslay M. D., Wallace A. V., Marchmont R. J. et al. Insulin controls intracellular cyclic AMP concentrations in hepatocytes by activating specific cyclic AMP phosphodiesterases: phosphorylation of the peripheral plasma membrane enzyme. Adv Cyclic Nucleotide Protein Phosphorylation Res 1984; 16: 159–176. 8. McGarry J. D., Foster D. W. Glucagon and ketogenesis. In: Lefebre P. J., ed. Glucagon I. Berlin: Springer-Verlag, 1983: 383–398. 9. Gamble M. S., Cook G. A. Alteration of the apparent Ki of carnitine palmitoyltransferase for malonyl-coA by the diabetic state and reversal by insulin. J Biol Chem 1985; 260: 9516–9519. 10. Eckel R. H., Yost T. J., Jensen D. R. Alterations in lipoprotein lipase in insulin resistance. Int J Obesity 1995; 19(Suppl 1): S16–S21. 11. Hardin D. S., Azzarelli B., Edwards J. et al. Mechanisms of enhanced insulin sensitivity in endurance-trained athletes: effects on blood flow and differential expression of GLUT 4 in skeletal muscles. J Clin Endocrinol Metab 1995; 80: 2437–2446. 12. O’Rahilly S. O., Hosker J. P., Rudenski A. S. et al. The glucose stimulus-response curve of the beta-cell in physically trained humans, assessed by hyperglycemic clamps. Metabolism 1988; 37: 919–923. 13. Fukagawa N. K., Anderson J. W., Hageman G. et al. Highcarbohydrate, high-fiber diets increase peripheral insulin sensitivity in healthy young and old adults. Am J Clin Nutr 1990; 52: 524–528. 14. Cefalu W. T., Bell-Farrow A., Wang Z. Q. et al. The effect of chromium supplementation on carbohydrate metabolism and body fat distribution. Diabetes 1997; 46(Suppl 1): 55A. 15. Barnard R. J., Ugianskis E. J., Martin D. A., Inkeles S. B. Role of diet and exercise in the management of hyperinsulinemia and associated atherosclerotic risk factors. Am J Cardiol 1992; 69: 440–444. 16. Després J. -P., Pouliot M. -C., Moorjani S. et al. Loss of abdominal fat and metabolic response to exercise training in obese women. Am J Physiol 1991; 261: E159–E167. 17. Sheppard L., Kristal A. R., Kushi L. H. Weight loss in women participating in a randomized trial of low-fat diets. Am J Clin Nutr 1991; 54: 821–828. 18. Lissner L., Heitmann B. L. The dietary fat: carbohydrate ratio in relation to body weight. Curr Opin Lipidol 1995; 6: 8–13. 19. Ornish D., Brown S. E., Scherwitz L. W. et al. Can lifestyle changes reverse coronary heart disease? Lancet 1990; 336: 129–133.
© 2000 Harcourt Publishers Ltd
The origins of Western obesity
20. Kaats G. R., Blum K., Fisher J. A., Adelman J. A. Effects of chromium picolinate supplementation on body composition: a randomized, double-masked, placebo-controlled study. Curr Ther Res 1996; 57: 747–756. 21. Jenkins D. J. A., Wolever T. M. S., Collier G. R. et al. Metabolic effects of a low-glycemic index diet. Am J Clin Nutr 1987; 46: 968–975. 22. Morgan L. M., Goulder T. J., Tsiolakis D. et al. The effect of unabsorbable carbohydrate on gut hormones. Modification of post-prandial GIP secretion by guar. Diabetologia 1979; 17: 85–89. 23. Krotkiewski M. Effect of guar gum on body-weight, hunger ratings and metabolism in these subjects. Br J Nutr 1984; 52: 97–105. 24. Swinburn B. A., Nyomba B. L., Saad M. F. et al. Insulin resistance associated with lower rates of weight gain in Pima Indians. J Clin Invest 1991; 88: 168–173. 25. Yost T. J., Jensen D. R., Eckel R. H. Weight regain following sustained weight reduction is predicted by relative insulin sensitivity. Obesity Res 1995; 3: 583–587. 26. Boden G. Role of fatty acids in the pathogenesis of insulin resistance and NIDDM. Diabetes 1996; 45: 3–10. 27. Kern P. A., Saghizadeh M., Ong J. M. et al. The expression of tumor necrosis factor in human adipose tissue. Regulation by obesity, weight loss, and relationship to lipoprotein lipase. J Clin Invest 1995; 95: 2111–2119. 28. Campfield L. A., Smith F. J., Mackie G. et al. Insulin normalization as an approach to the pharmacological treatment of obesity. Obesity Res 1995; 3(Suppl 4): 591S–603S. 29. Campbell T. C., Junshi C. Diet and chronic degenerative diseases: perspectives from China. Am J Clin Nutr 1994; 59(Suppl): 1153S–1161S. 30. Curb J. D., Marcus E. B. Body fat and obesity in Japanese Americans. Am J Clin Nutr 1991; 53: 1552S–1555S. 31. Lindahl O., Lindwall L., Spångberg A. et al. A vegan regimen with reduced medication in the treatment of hypertension. Br J Nutr 1984; 52: 11–20. 32. Kjeldsen-Kragh J., Brochgrevink C. F., Mowinkel P. et al. Controlled trial of fasting and one-year vegetarian diet in rheumatoid arthritis. Lancet 1991; 338: 899–902. 33. Hardinge M. G., Crooks H., Stare F. J. Nutritional studies of vegetarians. J Am Diet Assoc 1966; 48: 25–28. 34. Sacks F.M., Castelli W.P., Donner A., Kass E.H. Plasma lipids and lipoproteins in vegetarians and controls. N Engl J Med 1975; 292: 1148–1151. 35. Carter J. P., Furman T., Hutcheson H. R. Preeclampsia and reproductive performance in a community of vegans. Southern Med J 1987; 80: 692–697. 36. Acheson K. J., Schutz Y., Bessard T. et al. Nutritional influences on lipogenesis and thermogenesis after a carbohydrate meal. Am J Physiol 1984; 246: E62–E70. 37. Lissner L., Heitmann B. L. Dietary fat and obesity: evidence from epidemiology. Eur J Clin Nutr 1995; 49: 79–90. 38. Willett W. C. Is dietary fat a major determinant of body fat? Am J Clin Nutr 1998; 67(suppl): 556S–562S. 39. Ausman L. M., Rasmussen K. M., Gallinam D. L. Spontaneous obesity in maturing squirrel monkeys fed semipurified diets. Am J Physiol 1981; 241: R316–R321. 40. Anderson J. W., Herman R. H., Zakim D. Effect of high glucose and high sucrose diets on glucose tolerance of normal men. Am J Clin Nutr 1973; 26: 600–607. 41. Anderson J. W., Ward K. High-carbohydrate, high-fiber diets for insulin-treated men with diabetes mellitus. Am J Clin Nutr 1979; 32: 2312–2331. 42. Borkman M., Campbell L. V., Chisholm D. J., Storlien L. H. Comparison of the effects on insulin sensitivity of high carbohydrate and high fat diets in normal subjects. J Clin Endocrinol Metab 1991; 72: 432–437.
© 2000 Harcourt Publishers Ltd
493
43. McCarty M. F. Letter to the editor. Am J Clin Nutr 1998; 67: in press. 44. Resnicow K., Barone J., Engle A. et al. Diet and serum lipids in vegan vegetarians: a model for risk reduction. J Am Diet Assoc 1991; 91: 447–453. 45. Miller J. B., Pang E., Bramall L. Rice: a high or low glycemic index food? Am J Clin Nutr 1992; 56: 1034–1036. 46. Wolever T. M. S., Jenkins D. J. A., Kalmusky J. et al. Comparison of regular and parboiled rices: explanation of discrepancies between reported glycemic responses to rice. Nutr Res 1986; 6: 349–357. 47. Berger S., Vongaraya N. Insulin response to ingested protein in diabetes. Diabetes 1966; 15: 303–306. 48. Rabinowitz D., Merimee T. J., Maffezzoli R., Burgess J. A. Patterns of hormonal release after glucose, protein, and glucose plus protein. Lancet 1966; 2: 454–456. 49. Pallotta J. A., Kennedy P. J. Response of plasma insulin and growth hormone to carbohydrate and protein feeding. Metabolism 1968; 17: 901–908. 50. Nuttall F. Q., Mooradian A. D., Gannon M. C. et al. Effect of protein ingestion on the glucose and insulin response to a standardized oral glucose load. Diabetes Care 1984; 7: 465–470. 51. Spiller G. A., Jensen C. D., Pattison T. S. et al. Effect of protein dose on serum glucose and insulin response to sugars. Am J Clin Nutr 1987; 46: 474–480. 52. Gulliford M. C., Bicknell E. J., Scarpello J. H. Differential effect of protein and fat ingestion on blood glucose responses to highand low-glycemic-index carbohydrates in noninsulin-dependent diabetic subjects. Am J Clin Nutr 1989; 50: 73–777. 53. Zawadzki K. M., Yaspelkis B. B. III, Ivy J. L. Carbohydrate-protein complex increases the rate of muscle glycogen storage after exercise. J Appl Physiol 1992; 72: 1854–1859. 54. Krezowski P. A., Nuttall F. Q., Gannon M. C., Bartosh N. H. The effect of protein ingestion on the metabolic response to oral glucose in normal individuals. Am J Clin Nutr 1986; 44: 847–856. 55. Remer T., Peitrzik K., Manz F. A moderate increase in daily protein intake causing an enhanced endogenous insulin secretion does not alter circulating levels or urinary excretion of dehydroepiandrosterone sulfate. Metabolism 1996; 45: 1483–1486. 56. Diamond H., Diamond M. Fit for Life. New York: Warner Book, 1985. 57. Golay A., Coulston A. M., Hollenbeck C. B. et al. Comparison of metabolic effects of white beans processed into two different physical forms. Diabetes Care 1986; 9: 260–266. 58. Granfeldt Y., Björck I., Drews A., Tovar J. An in vitro procedure based on chewing to predict metabolic response to starch in cereal and legume products. Eur J Clin Nutr 1992; 46: 649–660. 59. Descovich G. C., Benassi M. S., Cappelli M. et al. Metabolic effects of lecithinated and non-lecithinated textured soy protein in hypercholesterolaemia. In: Noseda G., Fragiacomo C., Fumagalli R., Paoletti R., eds. Lipoproteins and Coronary Atherosclerosis, Amsterdam: Elsevier Biomedical Press, 1982; 279–288. 60. Sugano M., Ishiwaki N., Nakashima K. Dietary proteindependent modifications of serum cholesterol level in rats. Ann Nutr Metab 1984; 28: 192–199. 61. Hubbard R., Kosch C. L., Sanchez A. et al. Effect of dietary protein on serum insulin and glucagon levels in hyper- and normocholesterolemic men. Atherosclerosis 1989; 76: 55–61. 62. Sanchez A., Hubbard R. W. Plasma amino acids and the insulin/glucagon ratio as an explanation for the dietary protein modulation of atherosclerosis. Med Hypotheses 1991; 36: 27–32.
Medical Hypotheses (2000) 54(3), 488–494
494
McCarty
63. McCarty M. F. Vegan proteins may reduce risk for cancer, obesity, and cardio-vascular disease by promoting increased glucagon activity. Med Hypotheses 1999; 53: 459–485. 64. Iritani N., Hosomi H., Fukuda H. et al. Soybean protein suppresses hepatic lipogenic enzyme gene expression in Wistar fatty rats. J Nutr 1996; 126: 380–388. 65. Pugeat M., Crave J. C. C., Elmidani M. et al. Pathophysiology of sex hormone binding globulin (SHBG): relation to insulin. J Steroid Biochem Mol Biol 1991; 40: 841–849. 66. Gates J. R., Parpia B., Campbell T. C., Junshi C. Association of dietary factors and selected plasma variables with sex hormonebinding globulin in rural Chinese women. Am J Clin Nutr 1996; 63: 22–31. 67. Granfeldt Y., Björck I., Hagander B. On the importance of processing conditions, product thickness and egg addition for the glycaemic and hormonal responses to pasta: a comparison with bread made from ‘pasta ingredients’. Eur J Clin Nutr 1991; 45: 489–499. 68. McCarty M. F. Glycemic index may be a determinant of cancer risk. Submitted for publication, 1998. 69. Stumvoll M., Nurjhan N., Perriello G. et al. Metabolic effects of metformin in non-insulin-dependent diabetes mellitus. N Engl J Med 1995; 333: 550–554.
Medical Hypotheses (2000) 54(3), 488–494
70. Argaud D., Roth H., Weirnsperger N., Leverve X. M. Metformin decreases gluconeogenesis by enhancing the pyruvate kinase flux in isolated rat hepatocytes. Eur J Biochem 1993; 213: 1341–1348. 71. Clarke B. F., Campbell I.W. Comparison of metformin and chlorpropamide in non-obese, maturity-onset diabetics uncontrolled by diet. Br Med J 1977; 2: 1576–1578. 72. Zhang H., Osada K., Maebashi M. et al. A high biotin diet improves the impaired glucose tolerance of long-term spontaneously hyperglycemic rats with non-insulin-dependent diabetes mellitus. J Nutr Sci Vitaminol 1996; 42: 517–526. 73. McCarty M. F. High-dose biotin, an inducer of glucokinase expression, may synergize with chromium picolinate to enable a definitive nutritional therapy of type II diabetes. Med Hypotheses 1998; 00: 000–000. 74. Flatt J. P. Dietary fat, carbohydrate balance, and weight maintenance: effects of exercise. Am J Clin Nutr 1987; 45: 296–306. 75. McCarty M. F. Optimizing exercise for fat loss. Med Hypotheses 1995; 44: 325–330. 76. McCarty M. F. Promotion of hepatic lipid oxidation and gluconeogenesis as a strategy for appetite control. Med Hypotheses 1994; 42: 215–225.
© 2000 Harcourt Publishers Ltd
The origins of Western obesity: a role for animal protein? M. F. McCarty Helicon Foundation, San Diego, CA, USA
Summary A reduced propensity to oxidize fat, as indicated by a relatively high fasting respiratory quotient, is a major risk factor for weight gain. Increased insulin secretion works in various ways to impede fat oxidation and promote fat storage. The substantial ‘spontaneous’ weight loss often seen with very-low-fat dietary regimens may reflect not only a reduced rate of fat ingestion, but also an improved insulin sensitivity of skeletal muscle that down-regulates insulin secretion. Reduction of diurnal insulin secretion may also play a role in the fat loss often achieved with exercise training, low-glycemic-index diets, supplementation with soluble fiber or chromium, low-carbohydrate regimens, and biguanide therapy. The exceptional leanness of vegan cultures may reflect an additional factor – the absence of animal protein. Although dietary protein by itself provokes relatively little insulin release, it can markedly potentiate the insulin response to co-ingested carbohydrate; Western meals typically unite starchy foods with an animal proteinbased main course. Thus, postprandial insulin secretion may be reduced by either avoiding animal protein, or segregating it in low-carbohydrate meals; the latter practice is a feature of fad diets stressing ‘food combining’. Vegan diets tend to be relatively low in protein, legume protein may be slowly absorbed, and, as compared to animal protein, isolated soy protein provokes a greater release of glucagon, an enhancer of fat oxidation. The low insulin response to rice may mirror its low protein content. Minimizing diurnal insulin secretion in the context of a low fat intake may represent an effective strategy for achieving and maintaining leanness. © 2000 Harcourt Publishers Ltd
INSULIN ACTIVITY AS A DETERMINANT OF RESPIRATORY QUOTIENT In at least two prospective studies, a relatively high fasting respiratory quotient (RQ) at baseline (after standardized diets) has been shown to be a strong risk factor for the subsequent development of obesity (1,2). Analogously, rats which are resistant to the development of obesity on a high-fat diet have a lower 24-hour RQ than obesity-prone rats (3). It is hardly surprising that a reduced propensity to oxidize fat should be associated with a tendency to increase fat stores. Insulin activity opposes fat oxidation at several levels (4). In adipocytes, insulin promotes storage and retention of fat,
Received 6 April 1999 Accepted 14 April 1999 Correspondence to: Mark F. McCarty MD, NutriGuard Research, 1051 Hermes Avenue, Encinitas, CA 92024, USA
488
thereby limiting the availability of free fatty acids (FFA) and serum triglycerides to other tissues. Insulin enhances synthesis of adipocyte lipoprotein lipase (LPL) at the transcriptional level (5,6), inhibits catecholamine-stimulated lipolysis (by activating a cAMP phosphodiesterase), and promotes FFA re-esterification by increasing adipocyte glucose uptake. In hepatocytes, largely by opposing glucagon-mediated increases in cAMP (7), insulin impedes FFA oxidation and ketogenesis (8,9). In skeletal muscle, insulin’s suppression of FFA oxidation may mainly reflect increased availability of glucose as a competitive fuel (4). In lean subjects, insulin has a quantitatively modest suppressive effect on skeletal muscle LPL activity, whereas it produces a paradoxical increase of this activity in obese subjects (10). Insulin secretion can thus be expected to have a major regulatory impact on RQ. An increase in insulin secretion (or a reduction in insulin clearance) should raise RQ – so long as it does not precipitate hypoglycemia. Measures which selectively promote the efficiency of insulin-stimulated glucose uptake and storage in skeletal muscle (such
The origins of Western obesity
as exercise training, very-low-fat diets, and perhaps bioavailable chromium) (11–14) typically down-regulate insulin secretion (14,15), this may play a role in their documented efficacy for reducing body fat (16–20). A lowglycemic-index diet likewise should be favorable to fat oxidation owing to decreased insulin secretion (21), this dampening of the insulin response reflects not only a decrease in postprandial glycemic excursions, but also a decreased intestinal production of the incretin GIP, a potentiator of glucose-triggered insulin secretion (22). Weight loss associated with soluble fiber supplementation (23) (which slows the digestion of food starch) may thus be largely attributable to a decrease in diurnal insulin secretion. (In some prospective studies, efficient insulin-stimulated glucose uptake has been found to be predictive of subsequent weight gain (4,24,25); this seems paradoxical in light of the foregoing comments. It seems likely that, in the populations examined, good insulin sensitivity of skeletal muscle was not principally a reflection of regular exercise training or very-low-fat diets, but rather served as a marker for insulin-sensitive adipocytes. A common cause of insulin resistance in skeletal muscle is an increased FFA flux secondary to adipocyte insulin resistance (26). As adipocytes hypertrophy, internal mechanisms such as increased production of tumor necrosis factor diminish adipocyte sensitivity to insulin (27,28); this tends to lower the RQ by increasing FFA availability to other tissues – thus preventing a further increase in obesity (27) – but has the further unfortunate consequence of inducing insulin resistance in skeletal muscle. Thus, insulin sensitivity of adipocytes may be the genuine risk factor for weight gain; measures which enhance insulin sensitivity of skeletal muscle without increasing that of adipocytes may in fact promote weight loss by reducing insulin secretion.) A role for excessive beta cell responsiveness in the induction and maintenance of obesity in rodents is suggested by recent studies with the drug RO23-7637 (28). Two days of pre-incubation with this agent in vitro decreased the elevated insulin secretory response to glucose of beta cells obtained from obese rats; oral administration of this drug to obese rats for seven days dose-dependently decreased insulin levels both basally and during i.v. glucose tolerance tests. Chronic ingestion of RO23-7637 induced weight loss or impeded weight gain in Zucker obese rats as well as in rats with dietinduced obesity; reduced food intake played a role in this effect, but treated rats typically lost more weight than pair-fed untreated rats. Whether or not this drug attains clinical utility, these findings provide confirmation for the principle that reduction of diurnal insulin secretion – whether achieved by natural or pharmaceutical measures – may promote leanness. (However, as a caveat, it should © 2000 Harcourt Publishers Ltd
489
be noted that insulin-stimulated de novo lipogenesis is far more significant in rats than in humans.) VEGAN DIETS PROMOTE LEANNESS It is well known that, in comparison to Western populations, the average BMI and prevalence of obesity in Asian cultures consuming low-fat, predominantly vegan traditional diets is low. Recent studies in rural China indicate that neither low calorie consumption nor high physical activity can adequately account for this phenomenon (29). For example, the average BMI of rural Chinese males is 20.5, despite a daily calorie intake that, after downward adjustment to approximate that of men doing light office work, is still 30% higher per unit weight than that of the average American male (where BMI is 25.5). Inasmuch as obesity is not rare in Asian-Americans (30), it is unlikely that genetic factors contribute importantly to the rarity of obesity in rural Asians. Substantial (and often unintended) weight loss is typically seen when Western patients enter treatment programs that stress very-low-fat predominantly vegan diets and moderate daily exercise. During the first year of Ornish’s lifestyle regimen for reversing coronary disease, weight loss averages 25 pounds – despite the fact that most of the patients are not obese, are not explicitly attempting to lose weight, and are not calorically restricted (19). The Pritikin clinics report similar results, as do vegan health spas in Europe (15,31,32). The moderate exercise component of these regimens appears unlikely to account for the dramatic and often rapid weight loss experienced. Cross-sectional studies note that vegans tend to be substantially lighter than either lacto-ovo vegetarians or omnivores (33,34). For example, a survey of macrobiotic vegans in the Boston area found that, on average, they were 15 kg lighter than age/sex-matched controls chosen randomly from offspring of the Framingham cohort (34). Carter states that, on average, vegan adults are 20 pounds lighter than non-vegans of comparable age (35); this estimate accords well with the amount of weight loss seen during vegan therapy regimens. It is commonly assumed that a very low fat intake can account for these observations. This idea has particular intuitive appeal in light of evidence that, under usual circumstances, de novo lipogenesis is minimal in humans – implying that body fat derives predominantly from dietary fat (36). In general, cross-sectional studies demonstrate a correlation between relative weight and the percentage of fat calories in habitual diets (18,37). It is therefore surprising that, within Western societies, ecologic studies fail to note a correlation between habitual fat intake and prevalence of obesity (37,38). Furthermore, although efforts to treat obesity by initiating moderate Medical Hypotheses (2000) 54(3), 488–494
490
McCarty
reductions in dietary fat (reducing fat to 20–25% of calories) typically achieve significant weight loss, the magnitude of the average long-term weight loss on these regimens is often less than impressive (38). Such findings encourage skepticism regarding the role of dietary fat in the Western epidemic of obesity. However, the fat contents of the traditional diets of lean societies, and of very-low-fat therapeutic diets, generally fall in the 8–15% range. This suggests the possibility of a threshold effect – that very-low-fat diets do indeed promote leanness, but that, past some threshold level, percentage fat intake does not have a major impact on body composition; presumably the common ambient variations of fat intake in Western societies (25–45%) lie above this threshold. Such a phenomenon is noted in squirrel monkeys – they remain lean on 13% fat diets, but become obese on diets with twice this level of fat (39). Insulin sensitization might offer a mechanistic basis for such a threshold. Very-low-fat diets, with or without exercise training, are reported to notably enhance the efficiency of insulin-stimulated glucose uptake (13,40,41); this entails a compensatory down-regulation of insulin secretion that is most substantial in subjects who are hyper-insulinemic at baseline (13,15). There is no evidence that such diets (at least in the short-term, prior to substantial weight loss) directly sensitize adipocytes or the liver to insulin (though it seems likely that insulin receptor expression will be up-regulated somewhat in response to decreased insulinemia). Nor is there any evidence that diets with 20–30% fat calories produce clinically useful insulin sensitization relative to higher-fat diets (42). Thus, reduced diurnal insulin secretion may play a significant role in the promotion of leanness by very-low-fat diets – in conjunction with a minimization of fat ingestion (43). When dietary fat is manipulated within the range common in Western society, a moderate reduction in fat intake will be accompanied by a corresponding increase in carbohydrate (and/or protein) intake that increases insulin secretion and in addition impedes fat oxidation through substrate competition; thus, the long-term impact on body composition is likely to be minor. However, when dietary fat intake is slashed from, say, 30%–10% of calories, the down-regulation of insulin secretion consequent to improved insulin sensitivity at least partially compensates for the tendency of increased carbohydrate/protein ingestion to raise RQ – while fat ingestion is cut by two-thirds! Under these circumstances, dynamic fat loss can be expected until the decline in fat stores induces a sufficient increase in RQ. However, whether this offers a complete explanation for the leanness of vegan societies or of American vegans is questionable. Willett emphasizes that obesity is rare in predominantly vegan Chinese provinces whose fat Medical Hypotheses (2000) 54(3), 488–494
consumption is as high as 25% (38). Furthermore, the average fat content of American vegan diets is as high as 30% in some studies (owing to prominent use of cooking or salad oils, nuts or nut butters, and olives and avocados) (44). Some vegan societies or treatment programs emphasize whole, fiber-rich starchy foods that would be expected to have lower glycemic indices than more refined foods. While this factor undoubtedly contributes to the various health benefits of these regimens – including reductions in body weight – it should be noted that the ‘sticky’ white rice which is the staple of many oriental rural diets is not notably low in glycemic index (45). (Parboiled long-grain rice often has a relatively low glycemic index, but this is not the type of rice prevalent in China or Japan (46).) PROTEIN POTENTIATES INSULIN RESPONSE Ð A ROLE IN WESTERN OBESITY? The possible role of a relatively low dietary protein intake – and an absence of ‘high-quality’ animal protein – in the relative leanness of vegans, has so far received little attention. Perhaps this reflects the fact that protein-rich foods are generally viewed as appropriate for dieters, and that low-starch high-protein diet regimens have achieved considerable popularity as strategies for promoting weight loss. The proponents of such regimens point to the fact that protein is far less effective than starch in evoking insulin release, and thus is compatible with the maintenance of a relatively low RQ favorable to fat loss; the insulin excursion above baseline (the ‘insulin index’) after ingestion of pure protein is only about 20% as great as that seen after a comparable caloric load of starch (47). These considerations, however, may not be very relevant to typical ‘balanced’ Western meals in which ample intakes of carbohydrate are combined with significant amounts of animal protein. Although carbohydrate is the most substantial source of calories in most Western diets, major meals traditionally contain a slab of animal flesh that is considered the ‘main course’; dairy products and eggs are also common components of mixed meals. Sandwiches, the staple fare of fast-food restaurants and lunch boxes, usually combine animal flesh with a highglycemic-index starch (wheat bread). A number of reports indicate that co-ingestion of protein markedly potentiates the insulinemic response to starch (48–54). Thus, a bolus of protein which, eaten alone, would produce only a modest insulin excursion, can provoke a substantial increase in insulin output when eaten with a significant quantity of carbohydrate. Most, though not all (54), relevant studies conclude that this potentiation is synergistic rather than merely additive. However, most of these studies use protein doses, © 2000 Harcourt Publishers Ltd
The origins of Western obesity
e.g. (50–60 grams) higher than would be consumed in most meals. A very recent crossover study has assessed the impact of adding a realistic amount of animal protein to a relatively low protein vegetarian regimen (55). The baseline daily diet provided 282 g carbohydrate, 50 g protein, and 101 g fat; the higher protein diet was identical save for the substitution of 32 g egg protein for 14 g of fat. Diurnal insulin secretion was estimated on these regimens by measuring 24-hour urinary C-peptide. The findings were dramatic – on average, C-peptide excretion was 60% higher during the higher protein diet. Perhaps a protein-mediated modulation of renal function contributed in part to the observed increase in C-peptide excretion – but this possibility seems unlikely to explain the full magnitude of the increase. If this finding is replicable it suggests that the relatively high protein content of typical Western diets (as compared to vegan diets) has an important impact on diurnal insulin secretion, and thus may be a significant factor in the Western epidemic of obesity. High-protein diet programs place great emphasis on the avoidance of starchy foods and sugars (low-calorie salads and vegetables are allowed). At the other end of the spectrum, the regimens of Pritikin and Ornish, while carbohydrate-rich, totally proscribe animal flesh and animal fat (in an effort to avoid all dietary cholesterol) and recommend whole starchy foods, many of which have relatively low glycemic indices. While these programs appear to be diametrical opposites, they should each achieve the common goal of reducing diurnal insulin secretion – thus favoring the oxidation of stored fat. A ‘take-home lesson’ from these considerations would be the following: do not combine high-protein animal products with starchy foods in the same meal. Remarkably, this is a cardinal tenet of ‘food combining’ dogma that features prominently in certain popular diet programs (such as that popularized by the bestseller Fit for Life) (56). The proponents of these diet strategies justify them with rationales that, within the context of modern physiology, are frankly laughable. Such reservations, however, have no bearing on whether or not the recommended regimens actually do promote fat loss. (In the case of Fit for Life, the proscription on eating anything other than fruit prior to noon, the emphasis placed on vegan foods, and the restriction of animal flesh, cheese or eggs to evening meals devoid of starchy foods, would seem likely to promote the promised fat loss.) An interesting test of these ideas would be a crossover study comparing diets composed of identical foods combined in different ways, i.e. one diet in which animal protein was included in every meal, as compared to a second diet in which all animal protein was segregated in lowstarch meals or snacks. Certain vegan foods – notably legumes – are relatively protein-rich. There is also a trend toward use of soy protein © 2000 Harcourt Publishers Ltd
491
isolate in foods intended as more healthful substitutes for animal products. In regard to legumes, it is pleasing to observe that, at least when consumed intact, their ‘insulin index’ appears to parallel their very low glycemic index (57,58); could slow digestion of legume protein play a role in this? Studies are needed to address the insulin response to meals in which legumes are combined with starches of higher glycemic index, such as rice. As to isolated soy protein, it appears to be less effective for releasing insulin and more effective for releasing glucagon (which promotes hepatic lipid oxidation and ketogenesis) than certain animal proteins (59–62); this phenomenon may contribute to the many favorable effects of soy protein ingestion (as a substitute for animal protein) noted in clinical and animal studies – including reduced weight gain recently reported in obesity-prone rats (62–64). It has been suggested that, as a general rule, the essential amino acids provided more abundantly by ‘high-quality’ animal protein have greater efficacy for releasing insulin, whereas the non-essential amino acids which predominate in ‘lower quality’ vegan proteins have a greater impact on glucagon (62,63). However, while the substitution of soy protein for animal protein seems likely to be beneficial, what is the effect of adding substantial amounts of soy protein to relatively low protein vegan diets? Clearly, the role of vegan protein in modulating the insulinemic response to vegan diets deserves further attention. Even when the glycemic index of rice approaches that of wheat flour products, the insulinemic response to rice is low in comparison to wheat (45). This has been offered as an explanation for the recent observation that serum levels of sex hormone binding globulin (a protein whose hepatic synthesis is suppressed by insulin) (65) tend to be higher in Chinese provinces where rice (rather than wheat or millet) is the staple food (66). Rice has a notably low protein content – 7% of calories, only about half that of wheat; perhaps this plays a role in the lower insulinemic response to rice, as well as in the low incidence of obesity and of ‘Western’ degenerative disease in Asian societies where rice is the predominant source of calories. An interesting peculiarity of pasta is that, while its glycemic index is lower than that of wheat bread (especially if lightly cooked), its ‘insulin index’ is disproportionately lower – only about 40% that of bread (45,67). If we assume that the protein content of wheat has a significant impact on its ability to evoke insulin secretion (as suggested by the fact that bread has a relatively high insulin index equivalent to that of glucose), then the very low insulin response to pasta may reflect the interaction of two phenomena – delayed absorption of starch and delayed absorption of protein. In any case, habitual consumption of lightly cooked pasta – in preference to other wheat flour products – can be expected to have a favorable impact on weight control. Medical Hypotheses (2000) 54(3), 488–494
492
McCarty
A further question which should be addressed is whether significant intakes of animal protein, even if consumed in meals devoid of starch, might increase fasting insulin secretion by raising fasting levels of ‘insulinogenic’ amino acids (59). When consumed in starch-free meals, protein has the potential to promote hepatic fat oxidation by stimulating glucagon production. Presumably, soy protein isolate, with its greater impact on glucagon, would be most appropriate for this application.
TOWARD A ÔLOW-INSULIN LIFESTYLEÕ An overview of the foregoing discussion can provide insight into the origins of Western obesity. Much of the dietary carbohydrate in Western diets is relatively high glycemic index – wheat flour products (other than pasta), potatoes, sucrose or dextrose. The carbohydrate is typically combined in meals with significant amounts of animal protein, which potentiate the postprandial insulin response. Such diets are almost invariably high enough in fat to impair insulin sensitivity of skeletal muscle (thus up-regulating insulin secretion) and to provide a significant daily load of absorbed fat calories. The resulting ample diurnal insulin secretion impedes the efficiency of fat oxidation and promotes fat storage – and doubtless also plays a crucial role in the genesis of the degenerative diseases characteristic of Western society (63,68). In opposition to this, one may propose a ‘low-insulin lifestyle’ to promote leanness and long-term health. Down-regulation of insulin secretion can be achieved by aiding the efficiency of insulin-mediated glucose storage through exercise training, very-low-fat eating, and supplemental bioactive chromium. Restraining gluconeogenesis (without inhibiting hepatic lipid oxidation) may also be useful in this regard; this may be the clinical basis of metformin’s efficacy (69–71) (for promoting both weight loss and diabetic control), and high-dose biotin may likewise have this effect (72,73). The prandial stimulus to insulin release can be minimized by choosing lowglycemic-index starchy foods and fruit, by avoiding the co-ingestion of animal protein and significant quantities of starch, and – to the extent feasible – by moderating caloric intakes. Adjunctive measures which should aid fat oxidation include prolonged moderate-intensity aerobic exercise during the post-absorptive phase (74,75), and disinhibition of hepatic fatty acid oxidation with hydroxycitrate (76). By achieving an adequately low diurnal RQ in the context of a low rate of fat ingestion, these measures should produce a steady negative fat balance until a new equilibrium is reached at a much leaner body composition.
Medical Hypotheses (2000) 54(3), 488–494
REFERENCES 1. Zurlo F., Lillioja S., Esposito-Del Puente A. et al. Low ratio of fat to carbohydrate oxidation as predictor of weight gain: study 24RQ. Am J Physiol 1990; 259: E650–E657. 2. Seidell J. C., Muller D. C., Sorkin J. D., Andres R. Fasting respiratory exchange ratio and resting metabolic rate as predictors of weight gain: the Baltimore Longitudinal Study on Aging. Int J Obesity 1992; 16: 667–674. 3. Chang S., Graham B., Yakubu F. et al. Metabolic differences between obesity-prone and obesity-resistant rats. Am J Physiol 1990; 259: R1103–R1110. 4. Eckel R. H. Insulin resistance: an adaptation for weight maintenance. Lancet 1992; 340: 1452–1453. 5. Farese R. V. Jr, Yost T. J., Eckel R. H. Tissue-specific regulation of lipoprotein lipase activity by insulin/glucose in normal-weight humans. Metabolism 1991; 40: 214–216. 6. Fried S. K., Russell C. D., Grauso N. L., Brolin R. E. Lipoprotein lipase regulation by insulin and glucocorticoid in subcutaneous and omental adipose tissues of obese women and men. J Clin Invest 1993; 92: 2191–2198. 7. Houslay M. D., Wallace A. V., Marchmont R. J. et al. Insulin controls intracellular cyclic AMP concentrations in hepatocytes by activating specific cyclic AMP phosphodiesterases: phosphorylation of the peripheral plasma membrane enzyme. Adv Cyclic Nucleotide Protein Phosphorylation Res 1984; 16: 159–176. 8. McGarry J. D., Foster D. W. Glucagon and ketogenesis. In: Lefebre P. J., ed. Glucagon I. Berlin: Springer-Verlag, 1983: 383–398. 9. Gamble M. S., Cook G. A. Alteration of the apparent Ki of carnitine palmitoyltransferase for malonyl-coA by the diabetic state and reversal by insulin. J Biol Chem 1985; 260: 9516–9519. 10. Eckel R. H., Yost T. J., Jensen D. R. Alterations in lipoprotein lipase in insulin resistance. Int J Obesity 1995; 19(Suppl 1): S16–S21. 11. Hardin D. S., Azzarelli B., Edwards J. et al. Mechanisms of enhanced insulin sensitivity in endurance-trained athletes: effects on blood flow and differential expression of GLUT 4 in skeletal muscles. J Clin Endocrinol Metab 1995; 80: 2437–2446. 12. O’Rahilly S. O., Hosker J. P., Rudenski A. S. et al. The glucose stimulus-response curve of the beta-cell in physically trained humans, assessed by hyperglycemic clamps. Metabolism 1988; 37: 919–923. 13. Fukagawa N. K., Anderson J. W., Hageman G. et al. Highcarbohydrate, high-fiber diets increase peripheral insulin sensitivity in healthy young and old adults. Am J Clin Nutr 1990; 52: 524–528. 14. Cefalu W. T., Bell-Farrow A., Wang Z. Q. et al. The effect of chromium supplementation on carbohydrate metabolism and body fat distribution. Diabetes 1997; 46(Suppl 1): 55A. 15. Barnard R. J., Ugianskis E. J., Martin D. A., Inkeles S. B. Role of diet and exercise in the management of hyperinsulinemia and associated atherosclerotic risk factors. Am J Cardiol 1992; 69: 440–444. 16. Després J. -P., Pouliot M. -C., Moorjani S. et al. Loss of abdominal fat and metabolic response to exercise training in obese women. Am J Physiol 1991; 261: E159–E167. 17. Sheppard L., Kristal A. R., Kushi L. H. Weight loss in women participating in a randomized trial of low-fat diets. Am J Clin Nutr 1991; 54: 821–828. 18. Lissner L., Heitmann B. L. The dietary fat: carbohydrate ratio in relation to body weight. Curr Opin Lipidol 1995; 6: 8–13. 19. Ornish D., Brown S. E., Scherwitz L. W. et al. Can lifestyle changes reverse coronary heart disease? Lancet 1990; 336: 129–133.
© 2000 Harcourt Publishers Ltd
The origins of Western obesity
20. Kaats G. R., Blum K., Fisher J. A., Adelman J. A. Effects of chromium picolinate supplementation on body composition: a randomized, double-masked, placebo-controlled study. Curr Ther Res 1996; 57: 747–756. 21. Jenkins D. J. A., Wolever T. M. S., Collier G. R. et al. Metabolic effects of a low-glycemic index diet. Am J Clin Nutr 1987; 46: 968–975. 22. Morgan L. M., Goulder T. J., Tsiolakis D. et al. The effect of unabsorbable carbohydrate on gut hormones. Modification of post-prandial GIP secretion by guar. Diabetologia 1979; 17: 85–89. 23. Krotkiewski M. Effect of guar gum on body-weight, hunger ratings and metabolism in these subjects. Br J Nutr 1984; 52: 97–105. 24. Swinburn B. A., Nyomba B. L., Saad M. F. et al. Insulin resistance associated with lower rates of weight gain in Pima Indians. J Clin Invest 1991; 88: 168–173. 25. Yost T. J., Jensen D. R., Eckel R. H. Weight regain following sustained weight reduction is predicted by relative insulin sensitivity. Obesity Res 1995; 3: 583–587. 26. Boden G. Role of fatty acids in the pathogenesis of insulin resistance and NIDDM. Diabetes 1996; 45: 3–10. 27. Kern P. A., Saghizadeh M., Ong J. M. et al. The expression of tumor necrosis factor in human adipose tissue. Regulation by obesity, weight loss, and relationship to lipoprotein lipase. J Clin Invest 1995; 95: 2111–2119. 28. Campfield L. A., Smith F. J., Mackie G. et al. Insulin normalization as an approach to the pharmacological treatment of obesity. Obesity Res 1995; 3(Suppl 4): 591S–603S. 29. Campbell T. C., Junshi C. Diet and chronic degenerative diseases: perspectives from China. Am J Clin Nutr 1994; 59(Suppl): 1153S–1161S. 30. Curb J. D., Marcus E. B. Body fat and obesity in Japanese Americans. Am J Clin Nutr 1991; 53: 1552S–1555S. 31. Lindahl O., Lindwall L., Spångberg A. et al. A vegan regimen with reduced medication in the treatment of hypertension. Br J Nutr 1984; 52: 11–20. 32. Kjeldsen-Kragh J., Brochgrevink C. F., Mowinkel P. et al. Controlled trial of fasting and one-year vegetarian diet in rheumatoid arthritis. Lancet 1991; 338: 899–902. 33. Hardinge M. G., Crooks H., Stare F. J. Nutritional studies of vegetarians. J Am Diet Assoc 1966; 48: 25–28. 34. Sacks F.M., Castelli W.P., Donner A., Kass E.H. Plasma lipids and lipoproteins in vegetarians and controls. N Engl J Med 1975; 292: 1148–1151. 35. Carter J. P., Furman T., Hutcheson H. R. Preeclampsia and reproductive performance in a community of vegans. Southern Med J 1987; 80: 692–697. 36. Acheson K. J., Schutz Y., Bessard T. et al. Nutritional influences on lipogenesis and thermogenesis after a carbohydrate meal. Am J Physiol 1984; 246: E62–E70. 37. Lissner L., Heitmann B. L. Dietary fat and obesity: evidence from epidemiology. Eur J Clin Nutr 1995; 49: 79–90. 38. Willett W. C. Is dietary fat a major determinant of body fat? Am J Clin Nutr 1998; 67(suppl): 556S–562S. 39. Ausman L. M., Rasmussen K. M., Gallinam D. L. Spontaneous obesity in maturing squirrel monkeys fed semipurified diets. Am J Physiol 1981; 241: R316–R321. 40. Anderson J. W., Herman R. H., Zakim D. Effect of high glucose and high sucrose diets on glucose tolerance of normal men. Am J Clin Nutr 1973; 26: 600–607. 41. Anderson J. W., Ward K. High-carbohydrate, high-fiber diets for insulin-treated men with diabetes mellitus. Am J Clin Nutr 1979; 32: 2312–2331. 42. Borkman M., Campbell L. V., Chisholm D. J., Storlien L. H. Comparison of the effects on insulin sensitivity of high carbohydrate and high fat diets in normal subjects. J Clin Endocrinol Metab 1991; 72: 432–437.
© 2000 Harcourt Publishers Ltd
493
43. McCarty M. F. Letter to the editor. Am J Clin Nutr 1998; 67: in press. 44. Resnicow K., Barone J., Engle A. et al. Diet and serum lipids in vegan vegetarians: a model for risk reduction. J Am Diet Assoc 1991; 91: 447–453. 45. Miller J. B., Pang E., Bramall L. Rice: a high or low glycemic index food? Am J Clin Nutr 1992; 56: 1034–1036. 46. Wolever T. M. S., Jenkins D. J. A., Kalmusky J. et al. Comparison of regular and parboiled rices: explanation of discrepancies between reported glycemic responses to rice. Nutr Res 1986; 6: 349–357. 47. Berger S., Vongaraya N. Insulin response to ingested protein in diabetes. Diabetes 1966; 15: 303–306. 48. Rabinowitz D., Merimee T. J., Maffezzoli R., Burgess J. A. Patterns of hormonal release after glucose, protein, and glucose plus protein. Lancet 1966; 2: 454–456. 49. Pallotta J. A., Kennedy P. J. Response of plasma insulin and growth hormone to carbohydrate and protein feeding. Metabolism 1968; 17: 901–908. 50. Nuttall F. Q., Mooradian A. D., Gannon M. C. et al. Effect of protein ingestion on the glucose and insulin response to a standardized oral glucose load. Diabetes Care 1984; 7: 465–470. 51. Spiller G. A., Jensen C. D., Pattison T. S. et al. Effect of protein dose on serum glucose and insulin response to sugars. Am J Clin Nutr 1987; 46: 474–480. 52. Gulliford M. C., Bicknell E. J., Scarpello J. H. Differential effect of protein and fat ingestion on blood glucose responses to highand low-glycemic-index carbohydrates in noninsulin-dependent diabetic subjects. Am J Clin Nutr 1989; 50: 73–777. 53. Zawadzki K. M., Yaspelkis B. B. III, Ivy J. L. Carbohydrate-protein complex increases the rate of muscle glycogen storage after exercise. J Appl Physiol 1992; 72: 1854–1859. 54. Krezowski P. A., Nuttall F. Q., Gannon M. C., Bartosh N. H. The effect of protein ingestion on the metabolic response to oral glucose in normal individuals. Am J Clin Nutr 1986; 44: 847–856. 55. Remer T., Peitrzik K., Manz F. A moderate increase in daily protein intake causing an enhanced endogenous insulin secretion does not alter circulating levels or urinary excretion of dehydroepiandrosterone sulfate. Metabolism 1996; 45: 1483–1486. 56. Diamond H., Diamond M. Fit for Life. New York: Warner Book, 1985. 57. Golay A., Coulston A. M., Hollenbeck C. B. et al. Comparison of metabolic effects of white beans processed into two different physical forms. Diabetes Care 1986; 9: 260–266. 58. Granfeldt Y., Björck I., Drews A., Tovar J. An in vitro procedure based on chewing to predict metabolic response to starch in cereal and legume products. Eur J Clin Nutr 1992; 46: 649–660. 59. Descovich G. C., Benassi M. S., Cappelli M. et al. Metabolic effects of lecithinated and non-lecithinated textured soy protein in hypercholesterolaemia. In: Noseda G., Fragiacomo C., Fumagalli R., Paoletti R., eds. Lipoproteins and Coronary Atherosclerosis, Amsterdam: Elsevier Biomedical Press, 1982; 279–288. 60. Sugano M., Ishiwaki N., Nakashima K. Dietary proteindependent modifications of serum cholesterol level in rats. Ann Nutr Metab 1984; 28: 192–199. 61. Hubbard R., Kosch C. L., Sanchez A. et al. Effect of dietary protein on serum insulin and glucagon levels in hyper- and normocholesterolemic men. Atherosclerosis 1989; 76: 55–61. 62. Sanchez A., Hubbard R. W. Plasma amino acids and the insulin/glucagon ratio as an explanation for the dietary protein modulation of atherosclerosis. Med Hypotheses 1991; 36: 27–32.
Medical Hypotheses (2000) 54(3), 488–494
494
McCarty
63. McCarty M. F. Vegan proteins may reduce risk for cancer, obesity, and cardio-vascular disease by promoting increased glucagon activity. Med Hypotheses 1999; 53: 459–485. 64. Iritani N., Hosomi H., Fukuda H. et al. Soybean protein suppresses hepatic lipogenic enzyme gene expression in Wistar fatty rats. J Nutr 1996; 126: 380–388. 65. Pugeat M., Crave J. C. C., Elmidani M. et al. Pathophysiology of sex hormone binding globulin (SHBG): relation to insulin. J Steroid Biochem Mol Biol 1991; 40: 841–849. 66. Gates J. R., Parpia B., Campbell T. C., Junshi C. Association of dietary factors and selected plasma variables with sex hormonebinding globulin in rural Chinese women. Am J Clin Nutr 1996; 63: 22–31. 67. Granfeldt Y., Björck I., Hagander B. On the importance of processing conditions, product thickness and egg addition for the glycaemic and hormonal responses to pasta: a comparison with bread made from ‘pasta ingredients’. Eur J Clin Nutr 1991; 45: 489–499. 68. McCarty M. F. Glycemic index may be a determinant of cancer risk. Submitted for publication, 1998. 69. Stumvoll M., Nurjhan N., Perriello G. et al. Metabolic effects of metformin in non-insulin-dependent diabetes mellitus. N Engl J Med 1995; 333: 550–554.
Medical Hypotheses (2000) 54(3), 488–494
70. Argaud D., Roth H., Weirnsperger N., Leverve X. M. Metformin decreases gluconeogenesis by enhancing the pyruvate kinase flux in isolated rat hepatocytes. Eur J Biochem 1993; 213: 1341–1348. 71. Clarke B. F., Campbell I.W. Comparison of metformin and chlorpropamide in non-obese, maturity-onset diabetics uncontrolled by diet. Br Med J 1977; 2: 1576–1578. 72. Zhang H., Osada K., Maebashi M. et al. A high biotin diet improves the impaired glucose tolerance of long-term spontaneously hyperglycemic rats with non-insulin-dependent diabetes mellitus. J Nutr Sci Vitaminol 1996; 42: 517–526. 73. McCarty M. F. High-dose biotin, an inducer of glucokinase expression, may synergize with chromium picolinate to enable a definitive nutritional therapy of type II diabetes. Med Hypotheses 1998; 00: 000–000. 74. Flatt J. P. Dietary fat, carbohydrate balance, and weight maintenance: effects of exercise. Am J Clin Nutr 1987; 45: 296–306. 75. McCarty M. F. Optimizing exercise for fat loss. Med Hypotheses 1995; 44: 325–330. 76. McCarty M. F. Promotion of hepatic lipid oxidation and gluconeogenesis as a strategy for appetite control. Med Hypotheses 1994; 42: 215–225.
© 2000 Harcourt Publishers Ltd