The Journal of Nutrition
Recent Advances in Nutritional Sciences
Metabolic Effects of Dietary Fiber Consumption
and Prevention of Diabetes1
Martin O. Weickert* and Andreas F. H. Pfeiffer
Department of Clinical Nutrition, German Institute of Human Nutrition, Potsdam-Rehbruecke, Germany and Department of
Endocrinology, Diabetes, and Nutrition, Charité-University-Medicine-Berlin, Campus Benjamin Franklin, Berlin, Germany
Abstract
A high dietary fiber (DF) intake is emphasized in the recommendations
of most diabetes and nutritional associations. It is accepted that vis-
cous and gel-forming properties of soluble DF inhibit macronutrient
absorption, reduce postprandial glucose response, and beneficially in-
fluence certain blood lipids. Colonic fermentation of naturally available
high fiber foods can also be mainly attributed to soluble DF, whereas no
difference between soluble and insoluble DF consumption on the
regulation of body weight has been observed. However, in prospective
cohort studies, it is primarily insoluble cereal DF and whole grains, and
not soluble DF, that is consistently associated with reduced diabetes
risk, suggesting that further, unknown mechanisms are likely to be
involved. Recent research indicates that DF consumption contributes
to a number of unexpected metabolic effects independent from
changes in body weight, which include improvement of insulin sensi-
tivity, modulation of the secretion of certain gut hormones, and effects
on various metabolic and inflammatory markers that are associated
with the metabolic syndrome. In this review, we briefly summarize
novel findings from recent interventions and prospective cohort
studies. We discuss concepts and potential mechanisms that might
contribute to the further understanding of the involved processes.
Background
Consumption of soluble dietary fiber (DF)2 reduces postprandial
glucose responses after carbohydrate-rich meals, as well as
lowering total and LDL cholesterol levels (1). These effects are
likely explained the viscous and/or gel-forming properties of
soluble DF, which thereby slow gastric emptying and macronu-
trient absorption from the gut. However, it is not soluble DF, but
mainly the consumption of insoluble cereal DF and whole
grains, that is consistently associated with reduced risk of type 2
diabetes in large prospective cohort studies (2,3). A number of
recent studies give novel insights that might help establish a
metabolic link between insoluble DF consumption and reduced
diabetes risk. Potential candidates are improved insulin sensi-
tivity and the modulation of inflammatory markers, as well as
direct and indirect influences on the gut microbiota (Fig. 1).
1
Author disclosures: M. O. Weickert and A. F. H. Pfeiffer, no conflicts of interest.
2
Abbreviations used: DF, dietary fiber; GIP, glucose dependent insulinotropic
polypeptide; GPR, G protein-coupled receptors; LPS, lipopolysaccharide; RR,
relative risk for extreme quintiles; SCFA, short-chain fatty acids.
Manuscript received 15 October 2007.
* To whom correspondence should be addressed. E-mail: m.weickert@dife.de.
0022-3166/08 $8.00 ª 2008 American Society for Nutrition. J. Nutr. 138: 439–442, 2008.
Downloaded from https://academic.oup.com/jn/article-abstract/138/3/439/4670214
by guest
on 08 May 2018
Definition and types of DF
DF are highly complex substances that can be described as any
nondigestible carbohydrates and lignins not degraded in the
upper gut (4). Commonly, DF are classified according to their
solubility in water, even though grading according to viscosity,
gel-forming capabilities, or fermentation rate by the gut micro-
biota might be physiologically more relevant. Most DF is fer-
mented to some degree. However, fermentation rates of DF
widely vary, with soluble DF (i.e., pectin, inulin, and b-glucans)
and insoluble resistant starch and oligosaccharides tending to be
more readily fermented than cereal DF (i.e., cellulose and
hemicelluloses) (5). Foods are assumed to be whole grains if all
components of the kernel (i.e., bran, germ, and endosperm) are
present in their natural proportions. Whole grain food products
generally contain some 12% of total (mainly insoluble cereal)
DF, and there is a strong correlation between whole grain and
cereal DF intake (2). Some bran derived food products contain
up to 25% of DF. Importantly, definition of whole grain both in
cohort studies and on labels of commercially available cereal
products typically does not require an intact kernel (3). In U.S.
cohorts, whole grain and bran products from corn and wheat are
the main sources of cereal DF. Main sources of soluble DF are
fruits and vegetables (6) and, to a smaller extent, products from
oat and barley that are rich in both insoluble DF and soluble
b-glucans. It is, however, important to state that most naturally
available high-fiber foods contain both soluble and insoluble
fiber in varying amounts (6).
DF and risk of diabetes in prospective cohort studies
It is commonly assumed that beneficial effects of high fiber diets
[fiber intake .25 g/d in women and .38 g/d in men (7)] can
mainly be attributed to viscous and/or gel-forming properties of
soluble DF (1). Therefore, analyzing results from prospective
cohort studies separately for foods high in ‘‘active’’ soluble DF
and ‘‘inactive’’ insoluble DF should be expected to show stronger
protective associations for soluble DF. However, this hypothesis
is not supported by the available data. A recent meta-analysis
that included 328,212 subjects showed no association with
reduced diabetes risk both for fruit [relative risk for extreme
quintiles (RR) 0.96; 95% CI 0.88–1.04] and vegetable (RR 1.04;
95% CI 0.94–1.15) DF intake (2). In contrast, a high intake of
cereal DF was significantly associated with markedly reduced
diabetes risk in most studies (RR 0.67; 95% CI 0.62–0.72) (2).
Pooled data for 6 prospective cohort studies including 286,125
subjects indicate that a 2-serving-per-day increment in whole
grain consumption might remarkably reduce diabetes risk by
21% (3). Interestingly, associations for consumption of the outer
bran portion of the kernel, but not germ intake, were compa-
rable to those of whole grain intake (3).
439

FIGURE 1 Potential effects of DF consumption. Colonic fermenta-
tion with the production of SCFA can be observed with most types of
DF to some extent, but it tends to be more pronounced with soluble
DF in naturally available foods.
A causal relationship can not be stated, and estimation of
food intake on the basis of semiquantitative FFQ is a known
limitation (2). Moreover, even after statistical adjustment for
potentially confounding factors, residual confounding by addi-
tional unmeasured or inaccurately measured factors can not be
excluded. However, the consistency of the results clearly indi-
cates that the consumption of insoluble cereal DF could play an
important role in the prevention of diabetes.
Potential mechanisms
DF, satiety, and body weight. Potential mechanisms that
indicate DF consumption might alter diabetes risk include effects
on satiety and body weight. A number of studies showed in-
creased postmeal satiety or decreased subsequent hunger when
subjects consumed high DF diets, both under conditions of fixed
energy intake and when energy intake was consumed ad libitum
(7). However, other studies reported no significant effects (7–
10). Not all studies (11) showed an inverse association between
postprandial glucose and insulin responses and satiety, and no
clear conclusion can be drawn that low vs. high glycemic index
meals are a key factor promoting satiety (12).
Most (7), but not all (13), observational studies showed an
inverse, sometimes dose-dependent (14) relationship between
DF consumption and body weight. Effects were moderate, with
individuals in the highest quintile vs. lowest quintile of DF
consumption gaining 3.6 kg less over a 10-y period (14). Several
(7), but not all (9,15), relatively short-term interventions further
indicate that moderate reductions of body weight can be
achieved with high DF diets.
Generally, in human studies, no clear difference regarding
weight gain has been shown between soluble and insoluble DF
and fermentable and nonfermentable DF, or between foods nat-
440 Weickert and Pfeiffer
Downloaded from https://academic.oup.com/jn/article-abstract/138/3/439/4670214
by guest
on 08 May 2018
urally high in DF and fiber supplements (7). Therefore, reduced
body weight in subjects consuming high DF diets is likely to
contribute to reduced diabetes risk, but it cannot explain the
observed stronger associations for insoluble DF.
DF and hormonal responses. DF consumption affects the
secretion of various gut hormones that may act as satiety factors
(8,16–19). However, in many of the experiments in humans, hor-
monal changes were not linked to acute feelings of satiety. A guar
gum DF supplement produced a heightened postprandial chole-
cystokinin response, but did not alter satiety ratings (16). Insoluble
wheat DF-induced changes of orexigenic ghrelin and anorexigenic
peptide YY (PYY) were not associated with acute changes in
hunger and satiety (8) but may have affected satiety upon the
consumption of a subsequent meal (8,20). Highly fermentable DF
increase glucagon-like peptide 1 levels and may play a role in
the regulation of postprandial satiety in diverse animal species.
However, in humans, no fiber-induced changes of circulating
glucagon-like peptide 1 levels have been observed (17,18,21), and
ingestion of a fermentable (pectin, b-glucan) vs. a nonfermentable
(methylcellulose) DF supplement was found to be less, rather than
more, satiating (9). Glucose dependent insulinotropic polypeptide
(GIP) is another incretin hormone that appears to be involved in
the regulation of fat metabolism. Effects of DF consumption on
circulating GIP yielded mixed results. Soluble DF reduced circu-
lating GIP in diabetic humans (19), probably because of reduced
carbohydrate absorption, whereas insoluble cereal DF consump-
tion yielded accelerated responses of both biologically active GIP
(17) and insulin (17,22) in healthy subjects. In a cross-sectional
analysis, high intakes of cereal DF were positively associated with
plasma adiponectin after adjusting for lifestyle factors and dietary
glycemic load (23). Adiponectin may act as a peripheral starva-
tion signal promoting the storage of triglycerides preferentially in
adipose tissue (24). As a consequence, reduced triglyceride accu-
mulation in the liver and in the skeletal muscle might convey
improved systemic insulin sensitivity (25). In summary, various
changes in circulating concentrations of gut and adipocyte
derived hormones can be observed in humans that ingest a high
DF diet. However, no obvious mechanism can be derived ex-
plaining stronger associations for insoluble DF with diabetes risk.
DF and insulin sensitivity. An increased intake of total DF was
inversely associated with markers of insulin resistance in several
studies (26). Investigating different sorts of soluble and insoluble
DF in randomized controlled intervention studies yielded mixed
results. Consumption of wheat bran for 3 mo did not change
fasting glucose and glycated hemoglobin (HbA1c) levels in
diabetic subjects (27). High DF rye bread enhanced insulin secre-
tion but did not appear to improve insulin sensitivity in post-
menopausal women, as estimated with the frequently sampled
intravenous-glucose-tolerance test (28). However, using a sec-
ond meal test design, improved markers of insulin resistance
have been reported after consumption of various other sorts of
insoluble DF (17,18,20,29). When measuring insulin sensitivity
using euglycemic-hyperinsulinemic clamps, consumption of in-
soluble DF increased whole body glucose disposal independent
of changes in body weight in both short-term and more pro-
longed studies (21,30,31). Insulin resistant subjects are more
likely to eventually develop diabetes. Therefore, improved insu-
lin sensitivity could be a relevant factor contributing to reduced
diabetes risk in subjects consuming diets high in insoluble DF.
DF and inflammation. Recent studies show reductions of inflam-
matory markers in subjects consuming high DF diets (32). Inter-
 estingly, some sorts of DF have been reported to bind to specific
receptors on immune cells, suggesting a direct immune–modulatory
effect (5). Other potentially involved mechanisms might include
weight loss when consuming high DF diets, as well as antihyper-
glycemic effects and effects on lipid oxidation (32). Both a diet high
in total DF and consumption of a soluble DF supplement signif-
icantly decreased levels of the inflammatory marker C-reactive pro-
tein (33). Fermentation of soluble DF may also play a role due to
potential antiinflammatory properties of butyrate (32). However,
reductions in inflammatory markers have been reported to be
comparable with both insoluble DF and more readily fermentable
soluble DF (34).
Colonic fermentation and gut bacteria. Short-chain fatty
acids (SCFA) such as acetate, propionate, and butyrate are
produced by bacterial fermentation of indigestible DF polysac-
charides in the colon (1), with the proportion of different SCFA
not being fixed, depending on the substrate and eventually on
the gut microflora. Commonly, increased production of SCFA is
assumed to be beneficial by reducing hepatic glucose output and
improving lipid homeostasis (32). Further, G protein-coupled
receptors (GPR)-41 and GPR43 function as direct targets of
SCFA. An interesting finding was that oral administration of the
GPR41 ligand propionate almost doubled plasma concentra-
tions of anorexigenic leptin in mice, even though no effect on
food intake was detected (35).
However, data not unequivocally indicate that fermentability of
DF per se is a key factor contributing to reduced diabetes risk.
Consumption of low fermentable cereal DF (corn and wheat) show
stronger associations with reduced diabetes risk than more readily
fermentable soluble DF from fruit and vegetables (2,3). When
investigating effects of weakly fermentable insoluble cereal DF
and highly fermentable resistant starch in a randomized controlled
Downloaded from https://academic.oup.com/jn/article-abstract/138/3/439/4670214
by guest
on 08 May 2018
cross-over study in healthy women (17), markers of insulin
sensitivity in a second meal test were improved to a similar extent
with all DF, independent of the rate of colonic fermentation (Fig. 2).
These experiments indicate that a dose-dependent relation between
fermentability of DF and improved markers of insulin sensitivity
was unlikely, even though the limited accuracy of the available
methods to estimate colonic fermentation rates in humans must
be considered (4). Further, SCFA might stimulate adipogenesis
through GPR43 (35), and colonization of germ-free (gnotobiotic)
mice with a prominent saccharolytic member of the normal human
gut microbiota, together with the dominant human methane
producing germ, resulted in markedly improved efficacy of colonic
fermentation, associated with increased de novo lipogenesis and
obesity in the host (36). However, it needs to be emphasized that
nutritional habits and physiological effects of DF intake in rodents
and humans largely differ, with rodents recovering energy from
coprophagia (4). In humans, SCFA account for ,10% of daily
energy intake (37).
DF consumption might affect further factors linking the gut
microbiota with obesity and insulin resistance. Obese subjects
have a different composition of the gut microbiota than lean
subjects, and changes toward the ‘‘lean microbiota’’ can be
observed in obese subjects that lose weight [referenced in (36)].
When transplanting the gut microbiota from obese mice or from
lean mice to germ-free mice, the recipients of the ‘‘obese microbes’’
showed an increased gain of fat mass, even though energy intake
was comparable (36). Interestingly, a high DF diet (oligofructose)
reduced gram-negative bacterial content and body weight, whereas
a high fat diet increased the proportion of a gram-negative
bacterial lipopolysaccharides (LPS) containing microbiota in
humans [referenced in (38)]. Continuous subcutaneous infusion
of LPS for 4 wk increased weight gain, liver fat, inflammatory
markers, and markers of insulin resistance to a similar extent than
FIGURE 2 Effects of insoluble fiber on
postprandial glucose handling in a second
meal test. Postprandial serum insulin and
plasma glucose responses upon the inges-
tion of control white bread in a second meal
test, following the ingestion of 3 portions of
control white bread the previous day, or
white bread enriched with 31.5 g of the
insoluble fraction of nonfermentable wheat
DF, or white bread enriched with 31.8 g of
the insoluble fraction of moderately ferment-
able oat DF (n ¼ 14/group), or white bread
enriched with 31.2 g of highly fermentable
resistant starch (RS) (substudy, n ¼ 9/group).
Test breads were isoenergetic and macronu-
trient matched. Adapted from reference (17),
with permission.
Dietary fiber and risk of diabetes 441
 a high-fat diet (38). These studies indicate that DF consumption
could have the potential to influence the proportions of certain
members of the gut flora, thus linking the concepts of bacterial
colonization, inflammation, and insulin resistance.
Prospective cohort studies indicate that diets high in insoluble
cereal DF and whole grains might reduce diabetes risk. In contrast,
there is no strong support that soluble DF from fruits and vegetables
play a key role in this context. Many of the proposed protective
mechanisms of DF consumption are either shared by soluble and
insoluble DF (moderate weight loss, low energy density, increased
satiety, effects on inflammatory markers and gut hormones), or they
are more likely to be relevant with soluble viscous DF consumption
(hindering of macronutrient absorption, slowing of gastric empty-
ing, reduced postprandial glucose responses, reduced total and LDL
cholesterol, and colonic fermentation). Therefore, other unknown
mechanisms appear to be involved in conveying reduced diabetes
risk in subjects consuming diets high in insoluble cereal DF. A
promising factor contributing to beneficial effects of insoluble DF
consumption could be increased insulin sensitivity, even though
mechanisms leading to this phenomenon need to be defined. Hypo-
thetical mechanisms include a shift in the relation of gut microbiotic
communities, as well as direct and indirect influences on yet
unknown hormonal and molecular factors in the host that may be
altered in subjects consuming high DF diets.
Literature Cited
1. Jenkins DJ, Kendall CW, Axelsen M, Augustin LS, Vuksan V. Viscous
and nonviscous fibres, nonabsorbable and low glycaemic index carbo-
hydrates, blood lipids and coronary heart disease. Curr Opin Lipidol.
2000;11:49–56.
2. Schulze MB, Schulz M, Heidemann C, Schienkiewitz A, Hoffmann K,
Boeing H. Fiber and Magnesium Intake and Incidence of Type 2 Diabetes: A
Prospective Study and Meta-analysis. Arch Intern Med. 2007;167:956–65.
3. de Munter JS, Hu FB, Spiegelman D, Franz M, van Dam RM. Whole
grain, bran, and germ intake and risk of type 2 diabetes: a prospective
cohort study and systematic review. PLoS Med. 2007;4:e261.
4. Topping DL, Clifton PM. Short-chain fatty acids and human colonic
function: roles of resistant starch and nonstarch polysaccharides.
Physiol Rev. 2001;81:1031–64.
5. Vos A, M’Rabet L, Stahl B, Boehm G, Garssen J. Immune-modulatory
effects and potential working mechanisms of orally applied non-
digestible carbohydrates. Crit Rev Immunol. 2007;27:97–140.
6. McKee LH, Latner TA. Underutilized sources of dietary fiber: a review.
Plant Foods Hum Nutr. 2000;55:285–304.
7. Howarth NC, Saltzman E, Roberts SB. Dietary fiber and weight regu-
lation. Nutr Rev. 2001;59:129–39.
8. Weickert MO, Spranger J, Holst JJ, Otto B, Koebnick C, Mohlig M,
Pfeiffer AF. Wheat-fibre-induced changes of postprandial peptide YY
and ghrelin responses are not associated with acute alterations of
satiety. Br J Nutr. 2006;96:795–8.
9. Howarth NC, Saltzman E, McCrory MA, Greenberg AS, Dwyer J, Ausman
L, Kramer DG, Roberts SB. Fermentable and nonfermentable fiber sup-
plements did not alter hunger, satiety or body weight in a pilot study of
men and women consuming self-selected diets. J Nutr. 2003;133:3141–4.
10. Pereira MA, Ludwig DS. Dietary fiber and body-weight regulation.
Observations and mechanisms. Pediatr Clin North Am. 2001;48:969–80.
11. Keogh JB, Lau CW, Noakes M, Bowen J, Clifton PM. Effects of meals
with high soluble fibre, high amylose barley variant on glucose, insulin,
satiety and thermic effect of food in healthy lean women. Eur J Clin
Nutr. 2007;61:597–604.
12. Raben A. Should obese patients be counselled to follow a low-glycaemic
index diet? No. Obes Rev. 2002;3:245–56.
13. Iqbal SI, Helge JW, Heitmann BL. Do energy density and dietary fiber
influence subsequent 5-year weight changes in adult men and women?
Obesity (Silver Spring). 2006;14:106–14.
14. Ludwig DS, Pereira MA, Kroenke CH, Hilner JE, Van Horn L, Slattery
ML, Jacobs DR, Jr. Dietary fiber, weight gain, and cardiovascular
disease risk factors in young adults. JAMA. 1999;282:1539–46.
442 Weickert and Pfeiffer
Downloaded from https://academic.oup.com/jn/article-abstract/138/3/439/4670214
by guest
on 08 May 2018
15. Pittler MH, Ernst E. Guar gum for body weight reduction: meta-
analysis of randomized trials. Am J Med. 2001;110:724–30.
16. Heini AF, Lara-Castro C, Schneider H, Kirk KA, Considine RV, Weinsier
RL. Effect of hydrolyzed guar fiber on fasting and postprandial satiety and
satiety hormones: a double-blind, placebo-controlled trial during controlled
weight loss. Int J Obes Relat Metab Disord. 1998;22:906–9.
17. Weickert MO, Mohlig M, Koebnick C, Holst JJ, Namsolleck P, Ristow
M, Osterhoff M, Rochlitz H, Rudovich N, et al. Impact of cereal fibre
on glucose-regulating factors. Diabetologia. 2005;48:2343–53.
18. Robertson MD, Currie JM, Morgan LM, Jewell DP, Frayn KN. Prior
short-term consumption of resistant starch enhances postprandial
insulin sensitivity in healthy subjects. Diabetologia. 2003;46:659–65.
19. Gatenby SJ, Ellis PR, Morgan LM, Judd PA. Effect of partially depo-
lymerized guar gum on acute metabolic variables in patients with non-
insulin-dependent diabetes. Diabet Med. 1996;13:358–64.
20. Samra RA, Anderson GH. Insoluble cereal fiber reduces appetite and
short-term food intake and glycemic responses to food consumed 75
min later by healthy men. Am J Clin Nutr. 2007;86:972–9.
21. Robertson MD, Bickerton AS, Dennis AL, Vidal H, Frayn KN. Insulin-
sensitizing effects of dietary resistant starch and effects on skeletal muscle
and adipose tissue metabolism. Am J Clin Nutr. 2005;82:559–67.
22. Schenk S, Davidson CJ, Zderic TW, Byerley LO, Coyle EF. Different
glycemic indexes of breakfast cereals are not due to glucose entry
into blood but to glucose removal by tissue. Am J Clin Nutr. 2003;78:
742–8.
23. Qi L, Rimm E, Liu S, Rifai N, Hu FB. Dietary glycemic index, glycemic
load, cereal fiber, and plasma adiponectin concentration in diabetic
men. Diabetes Care. 2005;28:1022–8.
24. Kim JY, van de Wall E, Laplante M, Azzara A, Trujillo ME, Hofmann
SM, Schraw T, Durand JL, Li H, et al. Obesity-associated improvements
in metabolic profile through expansion of adipose tissue. J Clin Invest.
2007;117:2621–37.
25. Weickert MO, Pfeiffer AF. Signalling mechanisms linking hepatic
glucose and lipid metabolism. Diabetologia. 2006;49:1732–41.
26. Pi-Sunyer X. Do glycemic index, glycemic load, and fiber play a role in
insulin sensitivity, disposition index, and type 2 diabetes? Diabetes
Care. 2005;28:2978–9.
27. Jenkins DJ, Kendall CW, Augustin LS, Martini MC, Axelsen M,
Faulkner D, Vidgen E, Parker T, Lau H, et al. Effect of wheat bran on
glycemic control and risk factors for cardiovascular disease in type 2
diabetes. Diabetes Care. 2002;25:1522–8.
28. Juntunen KS, Laaksonen DE, Poutanen KS, Niskanen LK, Mykkanen
HM. High-fiber rye bread and insulin secretion and sensitivity in
healthy postmenopausal women. Am J Clin Nutr. 2003;77:385–91.
29. Granfeldt Y, Wu X, Bjorck I. Determination of glycaemic index; some meth-
odological aspects related to the analysis of carbohydrate load and char-
acteristics of the previous evening meal. Eur J Clin Nutr. 2006;60:104–12.
30. Pereira MA, Jacobs DR, Jr., Pins JJ, Raatz SK, Gross MD, Slavin JL,
Seaquist ER. Effect of whole grains on insulin sensitivity in overweight
hyperinsulinemic adults. Am J Clin Nutr. 2002;75:848–55.
31. Weickert MO, Mohlig M, Schofl C, Arafat AM, Otto B, Viehoff H,
Koebnick C, Kohl A, Spranger J, Pfeiffer AF. Cereal fiber improves
whole-body insulin sensitivity in overweight and obese women. Diabe-
tes Care. 2006;29:775–80.
32. Galisteo M, Duarte J, Zarzuelo A. Effects of dietary fibers on
disturbances clustered in the metabolic syndrome. J Nutr Biochem.
33. King DE, Egan BM, Woolson RF, Mainous, 3rd AG, Al-Solaiman Y,
Jesri A. Effect of a high-fiber diet vs a fiber-supplemented diet on
C-reactive protein level. Arch Intern Med. 2007;167:502–6.
34. Ma Y, Griffith JA, Chasan-Taber L, Olendzki BC, Jackson E, Stanek, 3rd
EJ, Li W, Pagoto SL, Hafner AR, Ockene IS. Association between dietary
fiber and serum C-reactive protein. Am J Clin Nutr. 2006;83:760–6.
35. Covington DK, Briscoe CA, Brown AJ, Jayawickreme CK. The
G-protein-coupled receptor 40 family (GPR40–GPR43) and its role in
nutrient sensing. Biochem Soc Trans. 2006;34:770–3.
36. Backhed F, Manchester JK, Semenkovich CF, Gordon JI. Mechanisms
underlying the resistance to diet-induced obesity in germ-free mice. Proc
Natl Acad Sci USA. 2007;104:979–84.
37. McNeil NI. The contribution of the large intestine to energy supplies in
man. Am J Clin Nutr. 1984;39:338–42.
38. Cani PD, Amar J, Iglesias MA, Poggi M, Knauf C, Bastelica D,
Neyrinck AM, Fava F, Tuohy KM, et al. Metabolic endotoxemia ini-
tiates obesity and insulin resistance. Diabetes. 2007;56:1761–72.