Quick Hit Summary

Advanced Glycation End Products (AGEs) are found in many food products; most notably in meat and foods with added fat that have been cooked using dry, high heat cooking techniques (grilling, roasting, broiling, frying). These compounds have beneficial effects on food from a culinary standpoint. Unfortunately, they also appear to play a significant role in the development of many cardiometabolic diseases. Part I of this article will examine the research regarding their negative effects on blood vessel dilation, kidney function, lipid profiles, diabetes and possibly even muscle function.

Advanced Glycation End Products (AGEs)

Figure 1. It’s hard to beat pizza straight out of a wood burning oven!¹³

Roasting, frying, broiling, baking… nothing gets my mouth watering quite like hearing these words. With the mere mention of them, my mind starts to drift off as I visualize a piece of meat slowly spinning in a rotisserie oven, hear the sizzles of meats/vegetables as they are being sautéed, or smell the aroma of pizza baking in an oven. If I had to guess, I’d say that I’m not the only one who conjures up pleasant memories upon hearing those various methods of cooking.

The aforementioned cooking techniques are great from a food appeal standpoint. However, these high heat cooking methods also speed up the development of Advanced Glycation End Products. Advanced-glyca-enda-what you say? Advanced glycation end products, commonly referred to as AGEs, are products that form when sugars undergo a chemical reaction with free amino groups of proteins or lipids¹. Once present within the body, they can alter protein structures or bind to cellular receptors (RAGE), kicking off inflammation and oxidative stress, leading to various cardiometabolic issues as described below²⁷.

With respect to dietary sources, AGEs are formed most readily under hot, dry heat cooking conditions. This chemical process was first noted by Dr. Louis Camille Maillard in 1912²³. Thus, it was termed the Maillard reaction and was characterized by the browning of food. [Technically speaking Maillard reaction products are intermediaries that later turn into AGEs.] Adding fat to recipes prior to cooking also enhances the formation of AGE’s. Thus, from a culinary perspective AGEs do have their benefits as they contribute to the overall aroma and flavor associated with a specific food.

As mentioned above, the formation of dietary AGEs (dAGEs) during the cooking process was first described in 1912²³. However, it wasn’t until the 1970’s before research really started to pick up on this topic. At this time, it was observed that diabetics had elevated levels of glycosylated hemoglobin (HbA1c), an intermediary product in the glycation process, in their blood²⁴. This was the first evidence that AGEs were naturally produced in-vivo in response to high blood sugar levels. Yet, even at this time, individuals had little awareness of how these compounds truly affected the body and it was still not known if dAGEs could be absorbed by one’s digestive track. Answers to these questions would have to wait a few more years.

In 1992, Helen et al. first described in detail the in-vivo physiological effects of AGEs. In their study, healthy rats and rabbits were injected with AGEs compounds for 2-4 weeks²⁵. After only a couple weeks of treatment, the animals started developing vascular defects similar to that seen in diabetics. Although this led to changes in how we looked at AGEs produced naturally by the body, it was still not believed that those which formed during the cooking process (ie – dAGEs) were absorbed by the body. This changed in 1997 when Koschinsky et al. found that 10% of the dietary AGEs one consumes are absorbed into circulation²⁶. Furthermore, of the AGEs ingested, only 1/3rd were excreted in individuals with normal renal function. In other words, 2/3rd of the absorbed dAGEs stayed in the body. In individuals with severe kidney disease, less than 1/20th of the absorbed AGEs were excreted.

With that as a backdrop, I’d like to look at…

The effect of Dietary AGEs on Cardiometabolic Health in Animal Studies

NOTE: For the remainder of this article (both animal and human based studies) I compare AGEs content of various meals. Please note that this is not the total AGEs content of the meal, but rather the total amount of AGEs most commonly associated with the development of chronic diseases.

Various animal based studies have shown the dietary AGEs contribute to the development of atherosclerosis² kidney disease³ as well as delayed wound healing¹⁶. Furthermore, they seem to contribute to other chronic diseases such as type I & II diabetes⁴⁵. They accomplish this primarily through binding to receptors (R-AGE) throughout the body. In doing so, they kick off inflammatory processes that promote “cellular dysfunction and tissue dysfunction”.²³

In an interesting study completed by Sandu et al., 40 HEALTHY mice consumed either a high fat low AGEs meal (LAGE) or a high fat high AGEs meal (HAGE) which contained 2.4x more AGEs than the other dietary option⁴. These diets were consumed for 6 months⁴. At the end of this time period, it was found that 75% of the mice assigned to the HAGE diet had developed Type II diabetes. In contrast, despite having similar gains in weight, not 1 of the 20 mice consuming the LAGE developed Type II diabetes. Therefore, it appears that differences in dietary AGEs, rather than consuming high fat diets, caused the higher incidence of diabetes in the mice on the HAGE diet.

Other studies, completed in mice genetically predisposed to develop Type I diabetes have also shown startling results. Peppa and colleagues assigned mice with these characteristics to either a LAGE or HAGE diet and followed both the parent group as well as 2 successive generations of offspring⁵. In the parent generation, 94% of them on the HAGE diet developed Type I diabetes vs. 33% of those on the LAGE diet. In the first generation of offspring, 61% of the infants fed a HAGE diet, after weaning, developed type I diabetes vs. 14% of the infants fed a LAGE diet. Similar results were seen in the 2rd generation of offspring. See Figure 1. For reference, the HAGE diet consisted of ~5x as much AGEs than the LAGE diet.

Figure 2. Percent of mice developing type I diabetes upon eating either a HAGE or LAGE diet

In addition, despite receiving insulin therapy, to normalize blood glucose levels, ALL of the founder mice in the HAGE diet were dead by week 44⁵. In contrast 76% of the mice receiving the LAGE diet were still alive be week 56.

To help put these AGEs content of these diets in perspective, a McDonalds® hamburger patty, cooked under dry high heat, has a ~2x higher AGEs content than beef which has cooked as part of a stew (ie – moist, reduced heat conditions) – 4876 AGE ku/90g vs. 2391 AGE ku/90g¹. Taking a look at common breakfast foods, cereals produced under dry heat, such as General Mills® Fiber One, have a 16x greater AGEs content than prepared instant oatmeal (421 AGE ku/30g vs. 25 AGE ku/175g). Kellogg’s ® Rice Krispies are the worst offender of the breakfast cereals (600 AGE kU/30g). Switching gears and examining snack foods, Nabisco’s® Chips Ahoy chocolate chip cookie has 130x higher concentration of AGEs vs. that of a good old Macintosh apple (505 AGEs ku/30g vs. 13 AGEs ku/100g).

Please note the serving sizes for the above foods ARE NOT EQUAL for all the comparisons. These sizes are used to reflect the more typical serving size.

The effect of Dietary AGEs on Cardiometabolic Health in Human Studies

I’m sure the aforementioned studies have some individuals relatively nervous about the way they cook their food. However, as I discussed in Making Sense of Animal Studies, results from animal based experiments do not always generalize over to humans. Thus, let’s take a peak at some human based studies…

Studies Completed in Diabetic Populations

Acute Studies

Negrean et al. investigated the effects of AGEs on vascular function and oxidative stress in 20 individuals with Type II diabetes⁶. During the study, participants consumed both a HAGE and LAGE meal, 2 days apart from one another. Meals were isocaloric and consisted of the same exact ingredients (chicken breast, potatoes, carrots, tomatoes and vegetable oil). The only difference between the meals was how the food was cooked. All items in the HAGE meal were prepared via broiling or frying at 230˚C for 20 minutes. In contrast, the LAGE meal was made by boiling or steaming the meal for 10 minutes at 100˚C.

Following the meal, markers of oxidative stress, vascular function and various lab values (blood glucose, lipids and insulin levels) were analyzed. No significant changes in lab values were found during the 6 hour observation period that followed each meal. However, AGEs content of the meals did significantly impact blood vessel function; macrovascular circulation decreased 150% more following the HAGE vs. LAGE meal (See Figure 3). A similar decrease in microvascular circulation was observed as well. The research team hypothesized that the reduced bloodflow observed following the HAGE meal was a direct result of impaired nitric oxide (NO) production by the vessels. Furthermore, the HAGE meal significantly increased oxidative stress by 21.3% (2 hours post meal) vs. baseline measurements. This was not observed following the LAGE meal.

Figure 3. Percent decrease in macrovascular function following a single HAGE or LAGE meal.

Chronic Studies

Vlassara et al. also examined the effects of AGE’s on various markers of health⁷. In their study, 13 diabetic individuals (4 type I diabetics/ 8 type II diabetics; mean age 62 years) were assigned to either LAGE or a HAGE diet for 6 weeks. The HAGE diet contained 5x as many AGEs (3.67 ×10^6 AGE units/day vs. 16.3×10^6 AGE units/day). In addition to abiding by the American Diabetes Association recommendations (50–55% carbohydrate, 20% protein, <30% fat,<10% saturated fat, and <300 mg cholesterol) each diet had the same relative kcal intakes.

Upon completion of the study Vlassera et al. found no significant differences between groups with respect to glucose control or triglyceride levels⁷. However, significant differences were found when measuring AGE-modified LDL cholesterol; those on the HAGE diet saw a 32% increase in AGE-LDL levels whereas those on the LAGE diet saw a 33% decrease. Furthermore, as seen in Table 1, significant changes in inflammatory markers (vs. baseline levels) were noted between groups.

Inflammatory Marker	LAGE Diet	HAGE Diet
C-Reactive Protein	-20%	+35%
TNFα	No Change	+86.3%
VCAM-1	-20%	No change

Table 1. Changes in inflammatory markers vs. that of baseline levels. A “+” represents a significant increase whereas a “-” represents a decrease in that variable. No significant differences were present between groups on these variables at the start of the study.

A recent study completed by Uribarri et al. examined the affects of consuming a typical western diet (which is high in AGEs ~ > 20,000 kU/day) or a reduced AGE meal for 4 months.²⁸ PLEASE NOTE – this study was looking at a TYPICAL WESTERN DIET, not one that has been purposely elevated with AGE content; the reduced group was ~50% less than normal. The study included 18 type 2 diabetics (14 women, 4 men; mean age – 61) who were free of kidney or overt cardiovascular disease.

At the end of 4 months, some very interesting results were found; reducing dAGE led to improved insulin control, decreased oxidative damage, and doubled adiponectin levels.²⁸ For those not familiar with adiponectin, this substance is produced by adipose tissue and released into one’s bloodstream. Reduced circulating levels of this hormone have been found in obese patients, type 2 diabetics and those with heart disease.²⁹ Thus, it has been hailed by some as “The missing link in insulin resistance and obesity.”³⁰ In other words, low adiponectin = bad news! However, this study clearly shows that these issue can be reversed simply by lowering daily dAGE consumption!

Glucose control remained under control for both study groups regardless of the diet they were on.⁷²⁸ Thus, taken into account the observed changes in inflammatory markers and AGE-LDL levels, Vlassara et al. hypothesized that dietary AGEs are another factor (independent of blood glucose control) that may contribute to the progression of diabetic atherogenesis.

Studies Completed on Healthy Populations

The effects of dietary AGEs have also been examined in healthy populations. Birlouez-Aragon et al. recently completed a study in which 62 individuals (mean age – 19 y; 32 women and 32 men; mean BMI: 21.8) completed 2 separate, 4 week diets which were either high or low in AGEs content⁸. Also, menu plans were isocaloric and of equal macronutrient distribution. The HAGE diet consisted of foods prepared by traditional dry high heat methods – grilling, frying, roasting, etc. In contrast, the LAGE diet had 2.5x lower AGEs and included food options that were either steamed, eaten raw or only underwent minimal baking (sponge cakes, lightly browned bread, etc). Upon conclusion of the study, researchers found that triglyceride levels were 9% lower following the LAGE vs. HAGE diet. No differences were seen in LDL cholesterol levels. As noted by the authors, insulin sensitivity was reduced…

“A strong increase (+17%) in the HOMA index after consumption of the STD [HAGE diet] compared with that of the STMD [LAGE diet] suggested that insulin was less efficient at normalizing plasma glucose in the case of a higher exposure to cooking-derived MRPs [AGEs]…⁸”

In conclusion, the researchers stated that…

“Replacing high-heat treatment techniques by mild cooking techniques may help to positively modulate biomarkers associated with an increased risk of diabetes mellitus and cardiovascular diseases.⁸”

For ethical reasons, this study could not be carried out in a human population for a long period of time. However, I’d like to remind you of the study in which HEALTHY mice were given a high or low AGE diet⁴… 6 months into it, 75% of the mice on the HAGE diet developed diabetes vs. 0% in the other group.

The Effects of AGEs on the Musculoskeletal System

Thus far we’ve talked about the effect of dietary AGEs on various disease states. Now, I’d like to shift our attention slightly and focus on an area that may be of even greater concern to the active, aging individual – muscular performance. It’s well known that muscular strength decreases whereas general joint stiffness increases with age. Similar to me, I’m sure you’ve heard individuals say, “I’m so dang stiff these days… When I was a young, my muscles and joints were so much limber. Don’t get old my friend!” Yes, I realize that the loss of muscle strength and range of motion is caused by multiple factors such as reduced hormone levels, etc. However, recent research indicates that AGEs may also contribute to the loss of these physical capabilities with age.

Muscular Function

Haus et al. completed an interesting study that examined the amount of AGEs within the connective tissue of skeletal muscle tissue of 22 older (mean age – 78 years) and 20 younger (mean age – 25 years) individuals¹⁴. Specifically, they were examining the amount of AGEs present in the connective tissue that surrounds each individual muscle fiber (ie- the endomysial collagen). Tissue samples obtained from the older individuals revealed ~ 200% higher concentration of AGEs (pentosidine) vs. that found in younger individuals. Thus, the research team stated that

“… the formation of AGEs over the lifespan … may contribute to increased muscle connective tissue protein stiffness and thus contribute to impaired muscle function in the elderly.”¹⁴

Muscle tissue may be affected by dAGEs in by a second mechanism. As pointed out to me by Dr. Helen Vlassara, one of the world’s leading experts on dAGEs, these compounds have been shown to accumulate in fat¹⁷. With the aging process, one experiences a natural increase in intramuscular fat content¹⁸. Taking into account these two principles, along with the fact that AGEs increase oxidative stress and inflammation, dAGEs may also increase scar tissue and/or cross-linking of muscle proteins¹⁹. If these theories hold true, dAGEs may also directly limit muscle flexibility and performance. See Figure 4

Figure 4. A hypothetical model as to how dAGEs may decrease muscle function beyond simply altering connective tissue within a muscle. Figure created by Sean Casey

Articular Cartilage

Articular cartilage is the main type of cartilage found in the smooth gliding joints (hips, knee, etc) of one’s body. The loss of this tissue results in the formation of osteoarthritis which is characterized by bone rubbing on bone. In an in-vitro study conducted by Verzijl and colleagues, healthy human articular cartilage was exposed to various levels of a pro-AGE forming compound.¹⁵ Final results of the study indicated that increased AGE formation decreased the pliability of articular cartilage. In other words, AGEs increased the stiffness of this connective tissue, making it more brittle. The research team speculated that this stiffness may be a contributing factor to cartilage degeneration. Thus, AGEs may accelerate the formation of osteoarthritis.

AGEs and Physical Activity

To my knowledge, there have been two studies that have examined the effects of AGEs and muscular performance. In older populations, reduced walking speed²⁰ as well as diminished grip strength²¹, have been found in those with higher circulating AGEs. If these relationships exist in younger, more active populations must still be determined.

I must note a word of caution when interpreting the results of the studies mentioned above¹⁴,¹⁵,²⁰,²¹. None of these directly measured the effects of dAGEs on muscle function. Rather, they measured the total amount of AGEs in muscle tissue samples at a given time. Remember, AGEs are from both dietary and metabolic sources. That being said, I strongly believe that dietary AGEs do influence the musculoskeletal system from a functional standpoint. However, more human based studies must be completed in order to confirm my hypothesis on dAGEs and physical activity/function.

A Vicious Cycle

As everyone knows, exercise plays a large role in maintaining sound cardiometabolic health. When one is overweight, etc, their ability to participate in physical exercise is compromised. With compromised ability to exercise, the risk of acquiring heart disease and/or diabetes is enhanced. This cycle turns nasty quick…. weight gain—> decreased cardiometabolic health—> reduced ability to exercise—-> further weight gain —->etc —->etc. Let’s take a step back, add in the potential effect of dAGEs, and we’ll see that this vicious cycle is even further enhanced, painting the picture seen in Figure 5.

Figure 5. A flow diagram of the vicious cycle involving dAGEs, ability to exercise, and the development of various cardiometabolic diseases. Figure created by Sean Casey

Bottom Line

AGEs are present in many foods, particularly meats and foods with added fat that have been cooked using dry high heating techniques. One can observe the formation of AGEs by watching the browning of meat or a pie crust as they are grilled/bake. Although they are beneficial from a culinary perspective, they appear to increase the risk of many cardiometabolic problems such as Type I and II diabetes, vascular and kidney disease. Furthermore, they also appear to negatively affect the musculoskeletal system.

Additionally, although not addressed in this article, AGEs have also been implicated in various neurological/cognitive based diseases.²²

And with that, Part I comes to a close. Be sure to check out Part II of this article as we discuss how one can still eat their favorite foods while minimizing dietary AGEs intake.

All values for specific food items (ie-not whole meals) were obtained from reference #1¹

References

¹ Uribarri J, Woodruff S, Goodman S, Cai W, Chen X, Pyzik R, Yong A, Striker GE, Vlassara H. Advanced glycation end products in foods and a practical guide to their reduction in the diet. J Am Diet Assoc. 2010 Jun;110(6):911-16.e12.

² Lin RY, Choudhury RP, Cai W, Lu M, Fallon JT, Fisher EA, Vlassara H. Dietary glycotoxins promote diabetic atherosclerosis in apolipoprotein. E-deficient mice. Atherosclerosis. 2003;168:213-220.

³ Zheng F, He C, Cai W,Hattori M, Steffes M, Vlassara H. Prevention of
diabetic nephropathy in mice by a diet low in glycoxidation products. Diabetes Metab Res Rev. 2002;18:224-237.

⁴ Sandu O, Song K, Cai W, Zheng F, Uribarri J, Vlassara H. Insulin resistance and type 2 diabetes in high-fat-fed mice are linked to high glycotoxin intake. Diabetes. 2005;54:2314-2319.

⁵ Peppa M, He C, Hattori M, McEvoy R, Zheng F, Vlassara H. Fetal or neonatal low-glycotoxin environment prevents autoimmune diabetes in NOD mice. Diabetes. 2003;52:1441-1445.

⁶ Negrean M, Stirban A, Stratmann B, Gawlowski T, Horstmann T, Götting C, Kleesiek K, Mueller-Roesel M, Koschinsky T, Uribarri J, Vlassara H, Tschoepe D. Effects of low- and high-advanced glycation endproduct meals on macro- and microvascular endothelial function and oxidative stress in patients with type 2 diabetes mellitus. Am J Clin Nutr. 2007 May;85(5):1236-43.

⁷ Vlassara H, Cai W, Crandall J, Goldberg T, Oberstein R, Dardaine V, Peppa M, Rayfield EJ. Inflammatory mediators are induced by dietary glycotoxins, a major risk factor for diabetic angiopathy. Proc Natl Acad Sci U S A. 2002 Nov 26;99(24):15596-601. Epub 2002 Nov 12.

⁸ Birlouez-Aragon I, Saavedra G, Tessier FJ, Galinier A, Ait-Ameur L, Lacoste F, Niamba CN, Alt N, Somoza V, Lecerf JM. A diet based on high-heat-treated foods promotes risk factors for diabetes mellitus and cardiovascular diseases. Am J Clin Nutr. 2010 May;91(5):1220-6. Epub 2010 Mar 24.

⁹ Burger J, Gochfeld M. Mercury in canned tuna: white versus light and temporal variation. Environ Res. 2004 Nov;96(3):239-49.

¹⁰ Gerstenberger SL, Martinson A, Kramer JL. An evaluation of mercury concentrations in three brands of canned tuna. Environ Toxicol Chem. 2010 Feb;29(2):237-42.

¹¹ Castoldi AF, Coccini T, Manzo L. Neurotoxic and molecular effects of methylmercury in humans. Rev Environ Health. 2003 Jan-Mar;18(1):19-31.

¹² ServSafe Essentials: with the Certification Exam Answer Sheet. NRA Educational Foundation. Wiley; 4 ed. April 21, 2006.

¹³ Image taken by Claus Ableiter. Accessed on October 1st, 2010 from: http://simple.wikipedia.org/wiki/File:Pizza_im_Pizzaofen_von_Maurizio.jpg

¹⁴ Haus JM, Carrithers JA, Trappe SW, Trappe TA. Collagen, cross-linking, and advanced glycation end products in aging human skeletal muscle. J Appl Physiol. 2007 Dec;103(6):2068-76. Epub 2007 Sep 27.

¹⁵ Verzijl N, DeGroot J, Ben ZC, Brau-Benjamin O, Maroudas A, Bank RA, Mizrahi J, Schalkwijk CG, Thorpe SR, Baynes JW, Bijlsma JW, Lafeber FP, TeKoppele JM. Crosslinking by advanced glycation end products increases the stiffness of the collagen network in human articular cartilage: a possible mechanism through which age is a risk factor for osteoarthritis. Arthritis Rheum. 2002 Jan;46(1):114-23.

¹⁶ Peppa M, Brem H, Ehrlich P, Zhang JG, Cai W, Li Z, Croitoru A, Thung S, Vlassara H.Adverse effects of dietary glycotoxins on wound healing in genetically diabetic mice. Diabetes. 2003 Nov;52(11):2805-13.

¹⁷ Vlassara, Helen.“Dietary AGE’s and decrements in strength with aging”. Message to Sean Casey. October 25, 2010. email

¹⁸ Lang T, Streeper T, Cawthon P, Baldwin K, Taaffe DR, Harris TB.
Sarcopenia: etiology, clinical consequences, intervention, and assessment. Osteoporos Int. 2010 Apr;21(4):543-59. Epub 2009 Sep 25.

¹⁹ Semba RD, Nicklett EJ, Ferrucci L. Does accumulation of advanced glycation end products contribute to the aging phenotype? J Gerontol A Biol Sci Med Sci. 2010 Sep;65(9):963-75. Epub 2010 May 17.

²⁰ Semba RD, Bandinelli S, Sun K, Guralnik JM, Ferrucci L.Eur J Appl Physiol. Relationship of an advanced glycation end product, plasma carboxymethyl-lysine, with slow walking speed in older adults: the InCHIANTI study. 2010 Jan;108(1):191-5. Epub 2009 Sep 16.

²¹ Dalal M, Ferrucci L, Sun K, Beck J, Fried LP, Semba RD. Elevated serum advanced glycation end products and poor grip strength in older community-dwelling women. J Gerontol A Biol Sci Med Sci. 2009 Jan;64(1):132-7. Epub 2009 Jan 31.

²² Seneff S, Wainwright G, Mascitelli L. Nutrition and Alzheimer’s disease: The detrimental role of a high carbohydrate diet. Eur J Intern Med. 2011 Apr;22(2):134-40. Epub 2011 Jan 26.

²³ Semba RD, Nicklett EJ, Ferrucci L. Does accumulation of advanced glycation end products contribute to the aging phenotype? J Gerontol A Biol Sci Med Sci. 2010 Sep;65(9):963-75. Epub 2010 May 17.

²⁴ Bunn HF, Gabbay KH, Gallop PM. The glycosylation of hemoglobin: relevance to diabetes mellitus. 1978 Apr 7;200(4337):21-7

²⁵ Vlassara H, Fuh H, Makita Z, Krungkrai S, Cerami A, Bucala R. Exogenous advanced glycosylation end products induce complex vascular dysfunction in normal animals: a model for diabetic and aging complications. Proc Natl Acad Sci U S A. 1992;89:12043–12047.

²⁶ Koschinsky T, He CJ, Mitsuhashi T, et al. Orally absorbed reactive glycation products (glycotoxins): an environmental risk factor in diabetic nephropathy. Proc Natl Acad Sci USA 1997;94:6474–9.

²⁷ Barlovic DP, Soro-Paavonen A, Jandeleit-Dahm KA. RAGE biology, atherosclerosis and diabetes. Clin Sci (Lond). 2011 Jul 1;121(2):43-55.

²⁸ Uribarri J, Cai W, Ramdas M, Goodman S, Pyzik R, Chen X, Zhu L, Striker GE, Vlassara H. Restriction of advanced glycation end products improves insulin resistance in human type 2 diabetes: potential role of AGER1 and SIRT1.Diabetes Care. 2011 Jul;34(7):1610-6.

²⁹ Antuna-Puente B, Feve B, Fellahi S, Bastard JP. Adipokines: the missing link between insulin resistance and obesity.Diabetes Metab. 2008 Feb;34(1):2-11.

³⁰ Gil-Campos M, Cañete RR, Gil A. Adiponectin, the missing link in insulin resistance and obesity. Clin Nutr. 2004 Oct;23(5):963-74.