912

REVIEW / SYNTHE`SE

Non-obese patients with type 2 diabetes and prediabetic subjects: distinct phenotypes requiring special diabetes treatment and (or) prevention? Allan Vaag and Søren S. Lund

Abstract: A major reason for the increased incidence of type 2 diabetes mellitus (T2DM) across the world is the so-called obesity epidemic, which occurs both in developed and developing countries. However, a large proportion of patients with T2DM in European and, in particular, Asian countries are non-obese. The non-obese T2DM phenotype is characterized by disproportionally reduced insulin secretion and less insulin resistance, as compared with obese patients with T2DM. Importantly, non-obese patients with T2DM have a similar increased risk of cardiovascular disease as obese T2DM patients. The risk of T2DM in non-obese patients is influenced by genetics as well as factors operating in utero indicated by low birth weight. Furthermore, this phenotype is slightly more prevalent among patients with latent autoimmune diabetes in adults, characterized by positive anti-GAD antibodies. The recently identified TCF7L2 gene polymorphism resulting in low insulin secretion influences the risk of T2DM in both obese and non-obese subjects, but is relatively more prevalent among non-obese patients with T2DM. Furthermore, the Pro12Ala polymorphism of the PPAR gene influencing insulin action increases the risk of T2DM in non-obese subjects. Despite a ‘‘normal’’ body mass index, non-obese patients with T2DM are generally characterized by a higher degree of both abdominal and total fat masses (adiposity). Prevention of T2DM with lifestyle intervention is at least as effective in non-obese as in obese prediabetic subjects, and recent data suggest that metformin treatment targeting insulin resistance and non-glycemic cardiovascular disease risk factors is as beneficial in non-obese as in obese patients with T2DM. Nevertheless, non-obese patients with T2DM may progress to insulin treatment more rapidly as compared with obese patients with T2DM. Key words: type 2 diabetes mellitus, non-obese, pathophysiology, prevention, treatment. Re´sume´ : La cause principale de la plus grande incidence de diabe`te de type 2 (T2DM) dans le monde entier est ladite e´pide´mie d’obe´site´ observe´e tant dans les pays de´veloppe´s que dans les pays en voie de de´veloppement. Par contre, on ne note pas d’obe´site´ chez une bonne proportion de patients souffrant du T2DM en Europe et notamment en Asie. Comparativement aux patients obe`ses souffrant de diabe`te de type 2, le phe´notype T2DM de poids normal est caracte´rise´ par une diminution disproportionne´e de se´cre´tion d’insuline et de la re´sistance a` l’insuline. Soulignons-le, les patients de poids normal souffrant de T2DM ont le meˆme risque de maladie cardiovasculaire (CVD) que les patients obe`ses souffrant de T2DM. Le risque de T2DM des patients de poids normal de´pend de facteurs ge´ne´tiques et de facteurs agissant dans l’ute´rus comme le re´ve`le le faible poids a` la naissance. De plus, on retrouve davantage ce phe´notype chez les patients souffrant du diabe`te de type 1 a` marche lente (« latent autoimmune diabetes » ou LADA) caracte´rise´ par la pre´sence d’anticorps anti-GAD (GAD, glutamate de´carboxylase). Le polymorphisme du ge`ne TCF7/L2, re´cemment identifie´ et caracte´rise´ par une faible se´cre´tion d’insuline, est associe´ au risque de T2DM chez les individus de poids normal et les obe`ses, mais on le retrouve davantage chez des patients de poids normal souffrant de T2DM. En outre, le polymorphisme PPAR Pro12Ala qui agit sur l’insuline augmente le risque de T2DM chez les sujets de poids normal. Malgre´ un IMC « normal », les patients non obe`ses souffrant de T2DM ont une forte teneur de gras abdominal et de gras total (adiposite´). La pre´vention du T2DM par l’intervention sur le mode de vie est tout aussi efficace chez les sujets de poids normal et chez les sujets obe`ses avant la manifestation du diabe`te; des e´tudes re´centes indiquent que le traitement des facteurs de risque non glyce´miques et de l’insulinore´sistance par la metformine re´ussit bien aux deux groupes, les obe`ses et ceux de poids norReceived 15 December 2006. Accepted 18 March 2007. Published on the NRC Research Press Web site at apnm.nrc.ca on 20 August 2007. A. Vaag.1 Steno Diabetes Center, Niels Steensens Vej 2, 2820 Gentofte, Denmark; Department of Endocrinology, Lund University, University Hospital Malmo¨, S-20502, Malmo¨, Sweden. S.S. Lund. Steno Diabetes Center, Niels Steensens Vej 2, 2820 Gentofte, Denmark. 1Corresponding

author (e-mail: [email protected]).

Appl. Physiol. Nutr. Metab. 32: 912–920 (2007)

doi:10.1139/H07-100

#

2007 NRC Canada

Vaag and Lund

913 mal. Quoi qu’il en soit, les patients de poids normal souffrant de T2DM re´agissent plus rapidement que les patients obe`ses souffrant aussi de T2DM au traitement de l’insuline. Mots-cle´s : diabe`te de type 2, poids normal, pathophysiologie, pre´vention, traitement. [Traduit par la Re´daction]

______________________________________________________________________________________ Introduction The prevalence and incidence of type 2 diabetes mellitus (T2DM) is increasing throughout the world, and it has been forecasted by the World Health Organization that the total numbers of patients with T2DM will be close to 300 million by the year of 2025 (King et al. 1998). Notably, the prevalence of T2DM in developing countries including China and India is expected to rise from around 100 million in 2006 to around 228 million by the year of 2025 (King et al. 1998). Undoubtedly, the global so-called ‘‘diabetes epidemic’’ is partly due to a more sedentary lifestyle and increased intake of energy-dense foods across the world. Nevertheless, it is often neglected that around 20% of patients with T2DM, at least in the northern European countries, are non-obese (Dalton et al. 2003; Garancini et al. 1995; Skarfors et al. 1991). Furthermore, it is well known that the Asian T2DM phenotype is commonly less obese when defined by body mass index (BMI) (Mohan et al. 1997; Mohan and Deepa 2006; Nakagami et al. 2003) (Fig. 1). However, as will be discussed later in this paper, non-obese patients with T2DM commonly have a higher total fat mass and abdominal obesity compared with a BMI-matched normal control group. Patients with T2DM in general have a 2- to 4-fold increased risk of cardiovascular disease (CVD) (Manson et al. 1991), and obesity is considered a risk factor of both T2DM and CVD per se (Willett et al. 1999). Importantly, studies have shown that non-obese patients with T2DM have a risk of CVD similar to that of obese patients with T2DM (Adlerberth et al. 1998; Manson et al. 1991) (Fig. 2). Accordingly, the association between the degree of obesity and the risk of CVD in patients with overt T2DM disappears after corrections for other confounding risk factors, including hypertension, dyslipidemia, family history of CVD, smoking, etc (Adlerberth et al. 1998). Our general and specific knowledge about the etiology, pathophysiology, and prevention of T2DM as well as the treatment of patients with T2DM has increased substantially throughout the last decade. The development of T2DM is a result of a complicated interplay between several primary, secondary, and tertiary etiological predisposing factors on one side, and, on the other side, a number of distinct organ defects of glucose homeostasis including muscle insulin resistance, elevated hepatic glucose production, defective insulin secretion, defective gut incretin hormone secretion, and altered adipocyte fatty acid as well as adipokine metabolism and secretion (Vaag 1999). The primary predisposing factors include an adverse intrauterine environment and genetics, and the secondary precipitation factors are obesity, low physical activity, and age. Tertiary accelerating factors are represented by glucose and lipid toxicities. The aim of this review is to focus explicitly on the etiology, pathophysiol-

Fig. 1. Relationship between obesity (BMI) and risk of type 2 diabetes (T2DM) in 60 year-old European and Asian men (A) and women (B). The impact of obesity has reached its maximal potential on risk of T2DM at a BMI of only around 25 kg/m2 in Indians. Markers: Europeans (~), Maltese (~), Indian (&), Chinese (*), and Japanese (*). (From Nakagami et al. for the DECODEDECODA Study Group 2003, reproduced with permission from Diabetologia, Vol. 46, p. 1066. # 2003 Springer Verlag.)

ogy, and prevention of T2DM, and treatment in non-obese patients with T2DM.

Obesity versus adiposity Obesity is defined as BMI above 30 kg/m2, whereas overweight is defined as BMI above 25 kg/m2 (World Health Organization 2003). Thus, obesity is defined solely by BMI as calculated from current weight and height, and it may be questioned whether such a simplistic classification is sufficient to categorize patients with T2DM into obese and nonobese subjects. It has been known for many years that the expanded adipose tissue mass plays a key role in the pathophysiology of the adverse metabolic and cardiovascular outcomes in obesity. However, our knowledge about the metabolic and hormonal effects of the adipose tissue per se has expanded significantly within recent years. From previ#

2007 NRC Canada

914 Fig. 2. The risk of coronary heart disease (CHD) is increased to the same magnitude in both non-obese (Quetelet index (i.e., body mass index) < 29) and obese (Quetelet index ‡ 29) patients with type 2 diabetes (open bars) compared with nondiabetic subjects (closed bars). (From Manson et al. 1991, reproduced with permission from Arch. Intern. Med. Vol. 151, p. 1144. # 1991 The American Medical Association.)

ously being considered as a relatively inert energy depot, we know today that the adipose tissue is indeed an important metabolic and, more specifically, endocrine organ, releasing a number of key substances associated to (or involved in) the development of subclinical inflammation, insulin resistance and defective insulin secretion, increased hepatic glucose production, increased coagulation, and altered appetite regulation in obesity and T2DM. The most interesting hormones released are leptin, adiponectin, plasminogen activator inhibitor-1, tumor necrosis factor alpha, interleukin 6, and, most recently, the retinoid binding protein 4 (Graham et al. 2006). The localization of the adipose tissue seems also to be relevant to the metabolic and hormonal effects. In particular, abdominal obesity is, independently of degree of total obesity, associated with insulin resistance and increased risk of CVD (Dalton et al. 2003; Yusuf et al. 2005). This may be partially because of distinct metabolic features of the intraabdominal visceral fat tissue. However, increased subcutaneous abdominal fat may also contribute to the metabolic risks of abdominal obesity, and recent studies have indicated that adiposity localized around the buttocks and thighs may even be associated with increased insulin action (Buemann et al. 2005). Concerning the focus of the present review being non-obese prediabetic and T2DM phenotypes, studies have shown that different distinct prediabetic and T2DM non-obese subjects exhibit expanded total fat mass, increased abdominal obesity, as well as relatively less adipose tissue in the lower region (Rasmussen et al. 2005; Yajnik et al. 2002; Yajnik and Yudkin 2004). This may, in particular, be the case for patients with T2DM of Asian origin (Mohan and Deepa 2006; Yajnik et al. 2002; Yajnik and Yudkin 2004). Accordingly, it would be more relevant to describe prediabetic and T2DM phenotypes into degrees of adiposity, describing the percentage of body mass accounted for by adipose tissue, and it may be questioned whether non-obese patients with T2DM are truly lean. This reservation may also be valid for some distinct prediabetic groups including those born with low birth weight

Appl. Physiol. Nutr. Metab. Vol. 32, 2007

(Rasmussen et al. 2005). Large-scale dual X-ray absorptiometry scan studies of prediabetic and diabetic populations are needed to gain more insights into that question.

Pathophysiology of type 2 diabetes in nonobese subjects A current issue for debate has been the extent to which impaired insulin secretion versus insulin resistance is the key defect responsible for the development of hyperglycemia in T2DM. Nevertheless, today there is general agreement that all patients with T2DM exhibit different degrees of defective insulin secretion, making them unable to compensate for the ambient degree of peripheral insulin resistance (Vaag 1999). In addition, patients with overt T2DM have an elevated rate of hepatic glucose production, and altered adipose tissue glucose and free fatty acid metabolism, as well as adipokine secretion, gut incretin hormone secretion and action, and, on top of that, possibly significant altered metabolism and hormone sensing in the central nervous system. Thus, T2DM represent a true ‘‘multiple organ disease,’’ and it is likely that different distinct T2DM subgroups are characterized by a predominance of different metabolic defects. Concerning the non-obese patients with T2DM, the predominant defect of metabolism is impaired insulin secretion (Vaag 1999). Peripheral and hepatic insulin resistance is also present in non-obese patients with T2DM, but not to the same extent as in obese patients (Hollenbeck et al. 1984; Vaag 1999). Studies of Asian (Korean) subjects have showed that, even within the non-obese range, BMI is correlated to the degree of insulin resistance in a linear manner (Chang et al. 2004). The key defect responsible for the elevation of fasting plasma glucose is an elevated rate of hepatic glucose production owing to enhanced gluconeogenesis (Vaag 1999). The disproportionately elevated basal hepatic glucose production is caused by a combination of metabolic defects intrinsic and extrinsic to the liver, including hepatic steathosis, elevated plasma free fatty acid levels, and Cori cycling (formation of glucose from excess tissue release of the 3-carbon substances lactate, pyruvate, and alanine), as well as impaired insulin and enhanced glucagon secretion (Vaag 1999). Non-obese patients with T2DM have the same degree of elevation of basal hepatic glucose production as seen in obese patients with T2DM, for the same fasting plasma glucose level (Vaag 1999). A relatively novel identified defect contributing to impaired insulin secretion in patients with T2DM is decreased secretion of the gut incretin hormone glucagon-like peptide-1 (GLP-1) (Vaag et al. 1996). Interestingly, the waist-to-hip ratio and BMI have been identified as independent (but oppositely directed) covariants for the incremental GLP-1 responses, with waist-tohip ratio being positively (and BMI inversely) related to GLP-1 secretion (Vaag et al. 1996). This means that nonobese patients with T2DM, as defined by BMI alone, may exhibit relatively well-conserved secretion of GLP-1 in response to both oral glucose and meals (Vaag et al. 1996). Concerning abnomalities of adipose tissue metabolism, obese patients with T2DM have an elevated rate of lipolysis in the absolute sense because of expanded fat mass (Vaag 1999). For the same reason, non-obese patients with T2DM are characterized by lower plasma leptin and elevated #

2007 NRC Canada

Vaag and Lund

plasma adiponectin levels compared with obese patients (Mojiminiyi et al. 2007). The elevated adiponectin level may partly explain the relatively well-conserved peripheral insulin action in the non-obese T2DM patients. Owing to the concept of leptin resistance, the role of leptin in both obese and non-obese patients with T2DM is incompletely understood. Defects of metabolism in other organs, including in particular appetite regulation in the central nervous system, is likely to play a significant role in the pathophysiology of T2DM, but unfortunately little is known about this in both obese and non-obese patients. As mentioned earlier in the text, the key defect responsible for the development of hyperglycemia in non-obese patients with T2DM is impaired pancreatic insulin secretion (Holman. 2006; Turner et al. 1995; Vaag 1999). The landmark United Kingdom Prospective Diabetes Study (UKPDS) documented that T2DM is characterized by a progressive impairment of insulin secretion, even preceding the development of overt hyperglycemia, and unfortunately it seems impossible to stop this progression with any of the current available glucose lowering drugs including metformin, insulin secretagogues, or insulin (Holman 2006; Turner et al. 1995) (Fig. 3). Recent autopsy studies of patients with T2DM reported that the defective insulin secretion, at least partly, may be due to a reduced beta-cell mass (Butler et al. 2003; Ritzel et al. 2006). Interestingly, the beta-cell mass was equally reduced in lean and obese patients with T2DM (Fig. 4). Nevertheless, there clearly appears to be an additional functional defect of pancreatic insulin secretion in patients with T2DM (Vaag 1999). This dysfunctional element is best described as a relative blindness towards glucose (as opposed to non-glucose) secretagogue-stimulated insulin secretion. The earliest and most consistent defect of insulin secretion in T2DM is a loss of first-phase insulin secretion in response to intravenous glucose (Vaag 1999). The extent to which the functional defect(s) of insulin secretion are secondary to glucose toxicity is currently unknown. Recent studies have indicated that the progressive decline of beta-cell mass in T2DM is caused by increased beta-cell apoptosis (Butler et al. 2003; Ritzel et al. 2006). The finding of similar beta-cell mass in obese compared with non-obese patients with T2DM suggests that the more severe beta-cell dysfunction in non-obese patients may be functional rather than structural. Nevertheless, caution is warranted in comparing insulin secretion in the absolute sense without correction for an ambient degree of insulin action in both nondiabetic and diabetic subjects (Vaag 1999). Given the significant contribution of peripheral insulin resistance to hyperglycemia in especially obese patients with T2DM and the apparently inevitable loss of beta-cell function with time in all patients with T2DM, it is likely that the notion of a more severe betacell dysfunction in the non-obese phenotype is simply an expression of a relatively postponed onset of T2DM for the degree of age-dependent loss of beta-cell function. In other words, the most significant pathophysiological difference between obese and non-obese patients with T2DM may be that the obese subjects have developed the disease at an earlier time point owing to the precipita-

915 Fig. 3. The natural history of type 2 diabetes (T2D) in both obese and non-obese patients with T2D involves an age-dependent decline of insulin secretion (homeostasis model assessment estimates), regardless of treatment with diet only (conventional), metformin, or sulfonylurea. (From Holman 2006, reproduced with permission from Metab. Clin. Exp., Vol. 55(Suppl. 1), p. S3. # 2006 W.B. Saunders.)

Fig. 4. The mean relative beta-cell mass, as determined in autopsies in obese (nondiabetic (ND), impaired fasting glucose (IFG), and type 2 diabetic (T2D) subjects) and lean cases (nondiabetic and T2D subjects). *p < 0.05 versus nondiabetic obese cases. {p < 0.01 versus nondiabetic lean cases. (Adapted from Butler et al. 2003, reproduced with permission from Diabetes, Vol. 52, p. 105. # 2003 The American Diabetes Association.)

tion by obesity. In support of this, twins who have developed overt T2DM are significantly more obese compared with their genetically identical co-twins who have escaped the disease (Vaag et al. 1995).

Genetics The notion of a genetic component in T2DM has been supported by a number of epidemiological and metabolic family studies, including studies in twins as well as studies in first-degree relatives (FDR) of patients with T2DM (Fernandez-Castaner et al. 1996; Poulsen et al. 1999; Vaag 1999; Vaag et al. 1992, 1996). Given the multifactorial etiology of T2DM, the extent to which genetics per se explains the various different defects of metabolism in T2DM is #

2007 NRC Canada

916

partly unknown. However, the role of genetics in non-obese compared with obese patients with T2DM may at least theoretically be more ‘‘straightforward.’’ Thus, total as well as abdominal obesity are strongly regulated by genetic factors, and it is therefore difficult to distinguish between ‘‘genetics of obesity’’ versus ‘‘genetics of hyperglycemia’’ in the obese T2DM phenotype. Detailed metabolic studies of subjects with genetic predisposition to T2DM, including both FDR and nondiabetic monozygotic co-twins of patients with T2DM, have shown defects of both insulin secretion and insulin action to be present several decades before overt onset of T2DM (Fernandez-Castaner et al. 1996; Vaag 1999; Vaag et al. 1992, 1995, 1996). In general, peripheral insulin resistance has been demonstrated in nondiabetic FDR of patients with T2DM in all age groups from the early 20s (Vaag et al. 1992; Vaag 1999), whereas defects of insulin secretion primarily has been reported in somewhat elderly FDR aged 30 years and upwards (Fernandez-Castaner et al. 1996; Vaag 1999; Vaag et al. 1995). Although some investigators have failed to document defects of insulin secretion in FDR, and therefore claimed that the predominant defect is at the site of insulin action, our studies of monozygotic twin pairs discordant for T2DM provided strong support in favor of a genetic defect of insulin secretion in T2DM (Vaag et al. 1995). A Spanish study divided FDR of patients with T2DM into 3 different age groups and showed that beta-cell dysfunction in the absolute sense was not detectable in FDR below the age of 28 years, whereas FDR above the age of 28 years showed signs of a subtle but significant impairment of insulin secretion (Fernandez-Castaner et al. 1996). The defective insulin secretion was even more pronounced in FDR above the age of 38 years (Fernandez-Castaner et al. 1996). Altogether, the majority of studies of genetically predisposed individuals support the concept of an accelerated decline of beta-cell function initiated more than a decade before overt onset of T2DM, subsequently continuing with unaltered or even more accelerated rates of decline, regardless of treatment after the onset of overt T2DM. There is little doubt that the primary (and, at least partly, genetically determined) defect of metabolism in non-obese patients with T2DM is at the site of insulin secretion. Presently, only a few genetic polymorphisms associated with defective insulin secretion and increased risk of T2DM have been documented; namely, a polymorphism of the transcription factor-7L2 gene (TCF7L2) (Saxena et al. 2006) and a genetic defect of the ATP-sensitive potassium canal Kir6.2 (or KCNJ11) in the pancreatic beta-cell. Interestingly, a recent study showed that carriers of the risk allele of the TCF7L2 gene polymorphism, besides impaired insulin secretion, were leaner and more insulin sensitive compared with the general T2DM phenotype (Saxena et al. 2006). Although a number of positive association studies have been published claiming novel T2DM susceptibility genes, it is important to emphasize that in fact only a few consistent but relatively minor T2DM susceptibility gene polymorphisms are known. Besides the TCF7L2 and KCNJ11 polymorphisms, 2 gene polymorphisms associated with insulin resistance are known. These are the peroxisome-proliferator-activated receptor-gamma 2 Pro12Ala variant and a variant of its cotranscriptional factor PGC1-. The extent to which these 2 gene polymorphisms

Appl. Physiol. Nutr. Metab. Vol. 32, 2007

contribute to the genetics of the non-obese T2DM phenotype is currently unknown. Large-scale genome-wide association studies applying several hundred thousands of single nucleotide polymorphisms are ongoing, and the initial results from these analyses have recently been released, confirming a role of the TCF7L2 genotype in T2DM as well as reporting suggestive evidence for a role of some other genes, including a pancreatic zinc transporter, in T2DM (Sladek et al. 2007). Regardless, given the notion of a more simplistic or pure genetics of hyperglycemia in the non-obese patient with T2DM, it is somewhat surprising that geneticists have not focused more explicitly on association and candidate gene studies in nonobese populations with T2DM.

The prenatal environment The position of genetics being the major primary predisposing factor in T2DM has been seriously challenged by the concept of the so-called ‘‘thrifty phenotype hypothesis,’’ proposing that altered programming in utero associated with low birth weight plays a significant role in predisposing to T2DM (Hermann et al. 2003; Jensen et al. 2002; Ozanne et al. 2005; Poulsen et al. 1997; Vaag et al. 2006). It is likely (but still unproven) that this hypothesis may play a significant and predominant role in the diabetes epidemics in the third world. The association between low birth weight and T2DM may explain and (or) contribute to the slightly lower adult height as well as increased abdominal obesity in patients with T2DM (Vaag 1999). The extent to which the intrauterine component contributes to diabetes in non-obese as compared with obese patients with T2DM is unknown, but studies in both Caucasian and Asian populations suggest that an adverse intrauterine environment linked with low birth weight contributes to increased adiposity and abdominal obesity, regardless of current BMI (Rasmussen et al. 2005; Yajnik et al. 2002). Besides adiposity and abdominal obesity, low birth weight has been linked to other key components of the metabolic syndrome, including hypertension, dyslipidemia, low grade inflammation, and, most importantly, premature risk from CVD. Animal studies, as well as studies of Caucasian young, lean, and otherwise healthy subjects, have documented the presence of many of the key defects in T2DM to be present in subjects born with low birth weight several decades before the subjects are supposed to be at risk for this disease (Vaag 1999; Vaag et al. 2006). These defects include an age-dependent decline of insulin secretion, reduced muscle glucose uptake and insulin stimulated glycolysis (present before the onset of overt wholebody insulin resistance), slightly elevated fasting plasma glucose concentrations within the normal range, abdominal obesity, and lower fasting plasma glycerol levels indicating reduced lipolysis and increased risk of fat accumulation (adiposity) (Hermann et al. 2003; Jensen et al. 2002; Ozanne et al. 2005; Vaag 1999; Vaag et al. 2006). Recent analysis of skeletal muscle as well as adipose tissue biopsies showed reduced expression of several key proteins involved in insulin signaling and glucose transport, including protein kinase C zeta, the 2 subunits of phosphoinositol-3 kinase (P-85 and P-110 ), and insulin-sensitive glucose transporter 4, in young lean men born with low birth weight (Ozanne et al. #

2007 NRC Canada

Vaag and Lund

2005). The striking similarities of the full-blown range of metabolic abnormalities in this distinct young and lean prediabetic phenotype on one side, and in patients with overt T2DM on the other side, strongly support a significant role of the intrauterine environment in the etiology and pathophysiology of T2DM. The fact that these abnormalities may be programmed by an adverse intrauterine environment is supported by animal studies showing that protein undernourishment in utero results in virtually similar quantitative and qualitative in vivo metabolic defects, as well as similar reduced tissue protein expression levels in the rat offspring (Ozanne et al. 2005). The finding of muscle insulin resistance and defective insulin signaling in this lean prediabetic phenotype indicates that peripheral insulin resistance, to some unknown extent, may also contribute to the pathophysiology of apparently non-obese patients with T2DM, at least when the eliciting factor is fetal programming.

Autoimmunity A distinct subpopulation of patients with T2DM is characterized by the presence of glutamic acid decarboxylase (GAD) antibodies, indicating autoimmunity resembling type 1 diabetes (T1DM) (Tuomi 2005). GAD antibodies were present in 15%–35% of patients with diagnosed T2DM at a younger age than 45 years and only in 7%–9% of older patients in the UKPDS (Tuomi 2005). It has been proposed to call this subgroup of the T2DM population as latent autoimmune diabetes in adults (LADA). Although the phenotype of patients with LADA at the onset of diabetes is very difficult or impossible to distinguish from the common T2DM phenotype, others have proposed to include these patients among T1DM patients. This idea is supported by the World Health Organization diabetes classification (World Health Organization 1999). The LADA diabetes phenotype is characterized by slightly earlier onset of disease, slightly less obesity, and a need for earlier initiation of insulin therapy compared with the general T2DM population (Tuomi 2005). Nevertheless, concerning degree of obesity in particular, there is substantial overlap with the common T2DM phenotype, and the extent to which autoimmunity and LADA contribute to the pathophysiology of non-obese patients with T2DM remains unknown. The prevalence of positive GAD antibodies among Danish Caucasian nonobese patients with T2DM is around 5%–10% (Lund et al. 2007). Furthermore, the prevalence of GAD antibodypositive patients among relatively non-obese, and even extremely lean (BMI < 18.5), Asian patients with T2DM has been reported to be lower than the prevalence in the more obese Caucasian T2DM population (Mohan et al. 1997; Turner et al. 1997). It is currently unknown whether the presence of GAD antibodies is a primary feature responsible for autoimmune destruction of the beta-cells, or whether these antibodies are a secondary phenomenon occurring from loss of beta-cell function by other autoimmune or non-autoimmune mechanisms. Altogether, the extent to which autoimmunity is a more significant player in nonobese as compared with obese patients with T2DM remains unknown.

917

Prevention of type 2 diabetes in non-obese subjects A number of studies published over the last decade have documented that T2DM, if not prevented, at least can be postponed by intensive lifestyle interventions with increased physical activity and low-calorie diet. In fact, the first study showing this was the Chinese Da Qing study, including relatively non-obese patients (Pan et al. 1997). Importantly, the study showed that lifestyle intervention was equally effective in subjects with a BMI below or above the median of this population (Pan et al. 1997). The large US Diabetes Prevention Program similarly showed that lifestyle intervention was equally effective in preventing T2DM in nonobese as compared with obese subjects (Ratner 2006). Delayed onset of T2DM can also be obtained with different antidiabetic medications, including metformin, glitazones, and acarbose. Interestingly, metformin treatment targeting hepatic and peripheral insulin resistance postponed or delayed the onset of overt T2DM only in obese subjects in the US Diabetes Prevention Program study (Ratner 2006).

Treatment of non-obese patients with type 2 diabetes The finding in the US Diabetes Prevention Program that metformin did not prevent T2DM in non-obese impaired glucose tolerance subjects was, to some extent, consistent with the idea that metformin treatment may be less effective in reducing hyperglycemia in non-obese patients. Thus, it has been clinical practice for many years to use insulin secretagogues as the ‘‘drug-of-first-choice’’ to treat non-obese patients with T2DM in the early phases of the disease (Inzucchi 2002). However, the UKPDS trial showed that metformin treatment was in fact the only glucose-reducing therapy associated with reduced incidence and mortality from CVD in obese patients with T2DM (UK Prospective Diabetes Study Group 1998). Given the excess morbidity and mortality from CVD in all patients with T2DM (Fig. 2), this was indeed an important finding. A number of studies have indicated that metformin exerts a number of beneficial effects on various non-glycemic mechanisms involved in the development of arteriosclerosis, including endothelial function, subclinical inflammation, coagulation, and foam cell formation (Kirpichnikov et al. 2002; Mamputu et al. 2003). Thus, metformin may exhibit protective CVD effects beyond those achieved through glucose reduction. Two observational studies suggested that metformin may be equally effective in lowering plasma glucose in nonobese as compared with obese patients with T2DM (Donnelly et al. 2006; Ong et al. 2006), and recent guidelines recommend metformin to be the drug-of-first-choice even in non-obese patients with T2DM (Nathan et al. 2006). We recently performed a randomized, double masked, prospective, crossover study comparing the efficacy of metformin against the insulin secretagogue repaglinide on glycemic control and a number of key non-glycemic CVD risk factors involved in coagulation, inflammation, and endothelial function in 96 non-obese Caucasian patients with T2DM (Lund et al. 2007). Consistent with the results from the observational data, we did not find any clinical significant differences between the effects of metformin versus re#

2007 NRC Canada

918

paglinide treatment on glycemic control (HbA1c), plasma adiponectin, or C-reactive protein concentrations in this trial (Lund et al. 2007). The additional results from this trial including a number of CVD risk factors related to postprandial glucose and lipid metabolism, endothelial function, inflammation, and coaguation are currently being analyzed, and this information is expected to provide important new knowledge of whether metformin may exert beneficial effects on a range of CVD risk factors, above and beyond those of insulin secretagogues, in non-obese patients with T2DM. Nevertheless, there is no doubt that non-obese patients with T2DM progress earlier to initiation of insulin therapy compared with obese patients with T2DM. In that respect, combination therapy of insulin with oral antidiabetic drugs is commonly used, and data have indicated a beneficial effect of combining metformin with insulin in predominantly obese patients with T2DM (Yki-Jarvinen et al. 1999). The potential beneficial effects include lower HbA1c, lower body mass, and reduced risk of hypoglycemia, as compared with insulin without metformin (Yki-Jarvinen et al. 1999). Although most patients in this trial were obese, the mean and range of BMI among study participants indicated that several patients were non-obese (Yki-Jarvinen et al. 1999). The extent to which metformin maintains its potential cardiovascular protective effects when combined with insulin in both obese and non-obese patients is unknown. To gain further insights into this important clinical question, we recently finished a 12 month study comparing the combination of insulin plus metformin versus insulin plus repaglinide on glycemic control and a number of nonglycemic CVD risk factors and end points in approximately 100 non-obese patients with T2DM (the ReForm trial). Until the results of this trial are calculated and published, the recently updated joint European Association for the Study of Diabetes and American Diabetes Association guidelines recommending metformin to be the drug-of-first-choice in mono as well as combination therapies with insulin in nonobese patients with T2DM should be followed (Nathan et al. 2006).

Conclusions Besides lower insulin secretion as well as relatively preserved in vivo insulin action, the etiology and pathophysiology, as well as the current state-of-the-art prevention and treatment regiments, are similar in non-obese compared with obese patients with T2DM. However, the criteria of BMI may be inadequate to classifying distinct subgroups of patients with T2DM according to ‘‘leanness.’’ Instead, more attention should be paid to classifying patients according to the degree of adiposity, reflecting the amount of adipose tissue in proportion to total body mass, as well as to the degree of abdominal obesity. The more pronounced impairment of beta-cell function in non-obese patients may indicate or reflect that the onset of overt T2DM has been postponed until the age-dependent decline of beta-cell function has progressed to a more severe degree compared with beta-cell function at the time of diabetes onset in obese patients. The impaired insulin secretion in both obese and non-obese patients with T2DM is owing to a combination of reduced beta-cell mass as well as an intrinsic dysfunction characterized by ‘‘glucose blindness.’’ Furthermore, the primary etiol-

Appl. Physiol. Nutr. Metab. Vol. 32, 2007

ogy of the beta-cell dysfunction in both obese and non-obese subjects is the result of a combination of genetic factors, intrauterine programming, and, perhaps to some extent, autoimmune operating factors. Prevention of T2DM can be obtained with lifestyle intervention, equally effective in both obese and non-obese patients, and results from recent clinical trials suggest that metformin treatment, targeting insulin resistance, is the oral antidiabetic drug-of-first-choice in monotherapy as well as in combination with insulin, even in non-obese patients. Nevertheless, the true potential of metformin versus insulin secretagogues on glycemic control and CVD risk reduction in non-obese patients with T2DM remains unknown, until the results of currently ongoing prospective and randomized trials are known.

Acknowledgements Allan Vaag is receiving research support from the European Union, 6th frame work, EXGENES consortium, grant 005272. Allan Vaag and Søren Lund have received research grants from Novo Nordisk A/S.

References Adlerberth, A.M., Rosengren, A., and Wilhelmsen, L. 1998. Diabetes and long-term risk of mortality from coronary and other causes in middle-aged Swedish men. A general population study. Diabetes Care, 21: 539–545. doi:10.2337/diacare.21.4. 539. PMID:9571339. Buemann, B., Sorensen, T.I., Pedersen, O., Black, E., Holst, C., Toubro, S., et al. 2005. Lower-body fat mass as an independent marker of insulin sensitivity – the role of adiponectin. Int. J. Obes. (Lond), 29: 624–631. doi:10.1038/sj.ijo.0802929. PMID: 15824752. Butler, A.E., Janson, J., Soeller, W.C., and Butler, P.C. 2003. Increased beta-cell apoptosis prevents adaptive increase in beta-cell mass in mouse model of type 2 diabetes: evidence for role of islet amyloid formation rather than direct action of amyloid. Diabetes, 52: 2304–2314. doi:10.2337/diabetes.52.9.2304. PMID:12941770. Chang, S.A., Kim, H.S., Yoon, K.H., Ko, S.H., Kwon, H.S., Kim, S.R., et al. 2004. Body mass index is the most important determining factor for the degree of insulin resistance in non-obese type 2 diabetic patients in Korea. Metabolism, 53: 142–146. doi:10.1016/S0026-0495(03)00314-7. PMID:14767863. Dalton, M., Cameron, A.J., Zimmet, P.Z., Shaw, J.E., Jolley, D., Dunstan, D.W., et al. 2003. Waist circumference, waist–hip ratio and body mass index and their correlation with cardiovascular disease risk factors in Australian adults. J. Intern. Med. 254: 555–563. doi:10.1111/j.1365-2796.2003.01229.x. PMID:14641796. Donnelly, L.A., Doney, A.S., Hattersley, A.T., Morris, A.D., and Pearson, E.R. 2006. The effect of obesity on glycaemic response to metformin or sulphonylureas in type 2 diabetes. Diabet. Med. 23: 128–133. doi:10.1111/j.1464-5491.2005.01755.x. PMID: 16433709. Fernandez-Castaner, M., Biarnes, J., Camps, I., Ripolles, J., Gomez, N., and Soler, J. 1996. Beta-cell dysfunction in first-degree relatives of patients with non-insulin-dependent diabetes mellitus. Diabet. Med. 13: 953–959. doi:10.1002/(SICI)1096-9136(199611) 13:113.0.CO;2-V. PMID:8946153. Garancini, M.P., Calori, G., Ruotolo, G., Manara, E., Izzo, A., Ebbli, E., et al. 1995. Prevalence of NIDDM and impaired glucose tolerance in Italy: an OGTT-based population study. Diabetologia, 38: 306–313. PMID:7758877. Graham, T.E., Yang, Q., Bluher, M., Hammarstedt, A., Ciaraldi, #

2007 NRC Canada

Vaag and Lund T.P., Henry, R.R., et al. 2006. Retinol-binding protein 4 and insulin resistance in lean, obese, and diabetic subjects. N. Engl. J. Med. 354: 2552–2563. doi:10.1056/NEJMoa054862. PMID: 16775236. Hermann, T.S., Rask-Madsen, C., Ihlemann, N., Dominguez, H., Jensen, C.B., Storgaard, H., et al. 2003. Normal insulinstimulated endothelial function and impaired insulin-stimulated muscle glucose uptake in young adults with low birth weight. J. Clin. Endocrinol. Metab. 88: 1252–1257. doi:10.1210/jc. 2002-021550. PMID:12629115. Hollenbeck, C.B., Chen, Y.D., and Reaven, G.M. 1984. A comparison of the relative effects of obesity and non-insulin-dependent diabetes mellitus on in vivo insulin-stimulated glucose utilization. Diabetes, 33: 622–626. doi:10.2337/diabetes.33.7.622. PMID:6376218. Holman, R.R. 2006. Long-term efficacy of sulfonylureas: a United Kingdom Prospective Diabetes Study perspective. Metab. Clin. Exp. 55(Suppl. 1): S2–S5. doi:10.1016/j.metabol.2006.02.006. PMID:16631806. Inzucchi, S.E. 2002. Oral antihyperglycemic therapy for type 2 diabetes: scientific review. JAMA, 287: 360–372. doi:10.1001/ jama.287.3.360. PMID:11790216. Jensen, C.B., Storgaard, H., Dela, F., Holst, J.J., Madsbad, S., and Vaag, A.A. 2002. Early differential defects of insulin secretion and action in 19-year-old Caucasian men who had low birth weight. Diabetes, 51: 1271–1280. doi:10.2337/diabetes.51.4. 1271. PMID:11916955. King, H., Aubert, R.E., and Herman, W.H. 1998. Global burden of diabetes, 1995–2025: prevalence, numerical estimates, and projections. Diabetes Care, 21: 1414–1431. doi:10.2337/diacare.21. 9.1414. PMID:9727886. Kirpichnikov, D., McFarlane, S.I., and Sowers, J.R. 2002. Metformin: an update. Ann. Intern. Med. 137: 25–33. PMID:12093242. Lund, S.S., Tarnow, L., Stehouwer, C.D., Schalkwijk, C.G., Frandsen, M., Smidt, U.M., et al. 2007. Targeting hyperglycaemia with either metformin or repaglinide in non-obese patients with type 2 diabetes: results from a randomized crossover trial. Diabetes Obes. Metab. 9: 394–407. doi:10.1111/j.1463-1326. 2007.00713.x. PMID:17391168. Mamputu, J.C., Wiernsperger, N.F., and Renier, G. 2003. Antiatherogenic properties of metformin: the experimental evidence. Diabetes Metab. 29: S671–S676. Manson, J.E., Colditz, G.A., Stampfer, M.J., Willett, W.C., Krolewski, A., Rosner, B., et al. 1991. A prospective study of maturity-onset diabetes mellitus and risk of coronary heart disease and stroke in women. Arch. Intern. Med. 151: 1141–1148. doi:10.1001/archinte.151.6.1141. PMID:2043016. Mohan, V., and Deepa, R. 2006. Adipocytokines and the expanding ‘Asian Indian Phenotype’. J. Assoc. Physicians India, 54: 685–686. PMID:17212014. Mohan, V., Vijayaprabha, R., Rema, M., Premalatha, G., Poongothai, S., Deepa, R., et al. 1997. Clinical profile of lean NIDDM in South India. Diabetes Res. Clin. Pract. 38: 101–108. doi:10.1016/S0168-8227(97)00088-0. PMID:9483373. Mojiminiyi, O.A., Abdella, N.A., Al, A.M., and Ben, N.A. 2007. Adiponectin, insulin resistance and clinical expression of the metabolic syndrome in patients with type 2 diabetes. Int. J. Obes. (Lond), 31: 213–220. doi:10.1038/sj.ijo.0803355. PMID: 16755284. Nakagami, T., Qiao, Q., Carstensen, B., Nøhr-Hansen, C., Hu, G., Tuomilehto, J., et al. 2003. Age, body mass index and type 2 diabetes-associations modified by ethnicity. Diabetologia, 46: 1063–1070. PMID:12827246. Nathan, D.M., Buse, J.B., Davidson, M.B., Heine, R.J., Holman,

919 R.R., Sherwin, R., et al. 2006. Management of hyperglycaemia in type 2 diabetes: a consensus algorithm for the initiation and adjustment of therapy. A consensus statement from the American Diabetes Association and the European Association for the Study of Diabetes. Diabetologia, 49: 1711–1721. doi:10.1007/ s00125-006-0316-2. PMID:16802130. Ong, C.R., Molyneaux, L.M., Constantino, M.I., Twigg, S.M., and Yue, D.K. 2006. Long-term efficacy of metformin therapy in nonobese individuals with type 2 diabetes. Diabetes Care, 29: 2361–2364. doi:10.2337/dc06-0827. PMID:17065668. Ozanne, S.E., Jensen, C.B., Tingey, K.J., Storgaard, H., Madsbad, S., and Vaag, A.A. 2005. Low birthweight is associated with specific changes in muscle insulin-signalling protein expression. Diabetologia, 48: 547–552. doi:10.1007/s00125-005-1669-7. PMID:15729577. Pan, X.R., Li, G.W., Hu, Y.H., Wang, J.X., Yang, W.Y., An, Z.X., et al. 1997. Effects of diet and exercise in preventing NIDDM in people with impaired glucose tolerance. The Da Qing IGT and Diabetes Study. Diabetes Care, 20: 537–544. doi:10.2337/ diacare.20.4.537. PMID:9096977. Poulsen, P., Vaag, A.A., Kyvik, K.O., Moller, J.D., and BeckNielsen, H. 1997. Low birth weight is associated with NIDDM in discordant monozygotic and dizygotic twin pairs. Diabetologia, 40: 439–446. doi:10.1007/s001250050698. PMID:9112021. Poulsen, P., Kyvik, K.O., Vaag, A., and Beck-Nielsen, H. 1999. Heritability of type II (non-insulin-dependent) diabetes mellitus and abnormal glucose tolerance – a population-based twin study. Diabetologia, 42: 139–145. doi:10.1007/s001250051131. PMID: 10064092. Rasmussen, E.L., Malis, C., Jensen, C.B., Jensen, J.E., Storgaard, H., Poulsen, P., et al. 2005. Altered fat tissue distribution in young adult men who had low birth weight. Diabetes Care, 28: 151–153. doi:10.2337/diacare.28.1.151. PMID:15616250. Ratner, R.E. 2006. An update on the Diabetes Prevention Program. Endocr. Pract. 12(Suppl. 1): 20–24. PMID:16627375. Ritzel, R.A., Butler, A.E., Rizza, R.A., Veldhuis, J.D., and Butler, P.C. 2006. Relationship between beta-cell mass and fasting blood glucose concentration in humans. Diabetes Care, 29: 717–718. doi:10.2337/diacare.29.03.06.dc05-1538. PMID:16505537. Saxena, R., Gianniny, L., Burtt, N.P., Lyssenko, V., Giuducci, C., Sjogren, M., et al. 2006. Common single nucleotide polymorphisms in TCF7L2 are reproducibly associated with type 2 diabetes and reduce the insulin response to glucose in nondiabetic individuals. Diabetes, 55: 2890–2895. doi:10.2337/db06-0381. PMID:17003358. Skarfors, E.T., Selinus, K.I., and Lithell, H.O. 1991. Risk factors for developing non-insulin dependent diabetes: a 10 year follow up of men in Uppsala. BMJ, 303: 755–760. PMID:1932936. Sladek, R., Rocheleau, G., Rung, J., Dina, C., Shen, L., Serre, D., et al. 2007. A genome-wide association study identifies novel risk loci for type 2 diabetes. Nature, 445: 881–885. doi:10. 1038/nature05616. PMID:17293876. Tuomi, T. 2005. Type 1 and type 2 diabetes: what do they have in common? Diabetes, 54(Suppl. 2): S40–S45. doi:10.2337/ diabetes.54.suppl_2.S40. PMID:16306339. Turner, R.C., Cull, C.A., Stratton, I.M., Manley, S.E., Kohner, E.M., Matthews, D.R., et al. 1995. U.K. prospective diabetes study 16. Overview of 6 years’ therapy of type II diabetes: a progressive disease. U.K. Prospective Diabetes Study Group. Diabetes, 44: 1249–1258. PMID:7589820. Turner, R., Stratton, I., Horton, V., Manley, S., Zimmet, P., Mackay, I.R., et al. 1997. UKPDS 25: autoantibodies to isletcell cytoplasm and glutamic acid decarboxylase for prediction of insulin requirement in type 2 diabetes. UK Prospective Dia#

2007 NRC Canada

920 betes Study Group. Lancet, 350: 1288–1293. doi:10.1016/S01406736(97)03062-6. PMID:9357409. UK Prospective Diabetes Study Group. 1998. Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). Lancet, 352: 854–865. doi:10.1016/S0140-6736(98)07037-8. PMID:9742977. Vaag, A. 1999. On the pathophysiology of late onset non-insulin dependent diabetes mellitus: current controversies and new insights. Dan. Med. Bull. 46:197–234. PMID:10421979. Vaag, A., Henriksen, J.E., and Beck-Nielsen, H. 1992. Decreased insulin activation of glycogen synthase in skeletal muscles in young nonobese Caucasian first-degree relatives of patients with noninsulin-dependent diabetes mellitus. J. Clin. Invest. 89: 782–788. PMID:1541672. Vaag, A., Henriksen, J.E., Madsbad, S., Holm, N., and BeckNielsen, H. 1995. Insulin secretion, insulin action, and hepatic glucose production in identical twins discordant for non-insulindependent diabetes mellitus. J. Clin. Invest. 95: 690–698. PMID:7860750. Vaag, A.A., Holst, J.J., Volund, A., and Beck-Nielsen, H.B. 1996. Gut incretin hormones in identical twins discordant for noninsulin-dependent diabetes mellitus (NIDDM): evidence for decreased glucagon-like peptide 1 secretion during oral glucose ingestion in NIDDM twins. Eur. J. Endocrinol. 135: 425–432. PMID:8921824. Vaag, A., Jensen, C.B., Poulsen, P., Brons, C., Pilgaard, K., Grunnet,

Appl. Physiol. Nutr. Metab. Vol. 32, 2007 L., et al. 2006. Metabolic aspects of insulin resistance in individuals born small for gestational age. Horm. Res. 65(Suppl. 3): 137–143. doi:10.1159/000091519. PMID:16612127. Willett, W.C., Dietz, W.H., and Colditz, G.A. 1999. Guidelines for healthy weight. N. Engl. J. Med. 341: 427–434. doi:10.1056/ NEJM199908053410607. PMID:10432328. World Health Organization. 1999. Definition, diagnosis and classification of diabetes mellitus and its complications. Report of a WHO Consultation. Part 1: diagnosis and classification of diabetes mellitus. WHO/NCD/NCS, 99.2: 1–59. World Health Organization. 2003. Diet, nutrition and the prevention of chronic diseases. WHO Tech. Rep. Ser. 916: 1–148. Yajnik, C.S., and Yudkin, J.S. 2004. The Y-Y paradox. Lancet, 363: 163. doi:10.1016/S0140-6736(03)15269-5. PMID:14726172. Yajnik, C.S., Lubree, H.G., Rege, S.S., Naik, S.S., Deshpande, J.A., Deshpande, S.S., et al. 2002. Adiposity and hyperinsulinemia in Indians are present at birth. J. Clin. Endocrinol. Metab. 87: 5575–5580. doi:10.1210/jc.2002-020434. PMID:12466355. Yki-Jarvinen, H., Ryysy, L., Nikkila, K., Tulokas, T., Vanamo, R., and Heikkila, M. 1999. Comparison of bedtime insulin regimens in patients with type 2 diabetes mellitus. A randomized, controlled trial. Ann. Intern. Med. 130: 389–396. PMID:10068412. Yusuf, S., Hawken, S., Ounpuu, S., Bautista, L., Franzosi, M.G., Commerford, P., et al. 2005. Obesity and the risk of myocardial infarction in 27 000 participants from 52 countries: a casecontrol study. Lancet, 366: 1640–1649. doi:10.1016/S01406736(05)67663-5. PMID:16271645.

#

2007 NRC Canada