Lean (Pre)Diabetes – Underestimated and Underexplored

Andreas F. H. Pfeiffer; Stefan Kabisch


J Clin Endocrinol Metab. 2021;106(8):e3278-e3280. 

The classification of type 2 diabetes (T2DM) into primarily insulin resistant or insulin deficient phenotypes is well known. Further subtyping of type 2 diabetes reflected relative differences in the risk of macrovascular and microvascular disease. However, the more extensive subclassification maintained overweight or obesity as a primary characteristic.[1] Although type 2 diabetes is generally associated with obesity and components of the metabolic syndrome, there is a significant fraction of lean people developing T2DM which is represented in the subtype proposals as the insulin deficient subtype who still had a mean body mass index (BMI) of 29 kg/m2.[1]

A study in this issue of the Journal of Clinical Endocrinology & Metabolism[2] carefully characterized women with gestational diabetes who developed prediabetes or manifest diabetes 10 months after delivery and women who had normal glucose tolerance after gestational diabetes. Remarkably, almost half of the participants with prediabetes or T2DM had a low and definitively normal BMI of around 22 kg/m2, normal triglyceride levels, low waist circumference, and a liver fat content below 1% as determined by determined by magnetic resonance imaging and spectroscopy. Despite the absence of markers of the metabolic syndrome, these individuals displayed a mixed phenotype of deficient insulin secretion but also insulin resistance and thus their conditions are not explained by insulin deficiency or beta-cell dysfunction.

Insulin secretion was characterized by the acute insulin response (AIR) in the intravenous glucose tolerance test (ivGTT), a highly sensitive method. AIR is typically lost in many cases of prediabetes.[3] However, AIR was normal in the lean women with prediabetes, suggesting a preserved beta-cell function. This was also supported by intact C-peptide responses in the oral glucose tolerance test (oGTT) in the first 30 minutes, indicating a rapid insulin response. However, the increased fasting glucose or impaired glucose tolerance documents some beta-cell dysfunction. Patterns of insulin secretion in the oGTT were not reported, which might be interesting in view of recent studies in young obese patients with prediabetes in whom the biphasic, monophasic, or incessant glucose and insulin responses were linked to different rates of progression and to reduced incretin release.[4] A determination of incretin responses would therefore be of interest, too.

The authors further examined the origins of the insulin resistance and showed components of both hepatic and muscular insulin resistance compared with the controls. The hepatic insulin resistance was calculated from oGTTs using the algorithm of Abdul-Ghani and coworkers,[5] which uses the relation of insulin to glucose values in the first 30 minutes of the oGTT when hepatic glucose disposal is assumed to predominate. Although this is an indirect assessment, it cannot be attributed to delayed insulin release in view of the normal AIR and C-peptide response. Hepatic insulin resistance is usually linked to increased liver fat in humans and the associated disturbances of fat metabolism including activation of protein kinase C enzymes by diacylglycerol (DAG) and possibly increased ceramides.[6] In this presently described sample of post-gestational diabetes mellitus (GDM) women with prediabetes or T2DM, a considerable portion of patients showed hepatic insulin resistance despite having less than 1% liver fat content.

Remarkably, hepatic insulin resistance is very rapidly lost upon caloric restriction in obese patients with fatty liver, which is linked to reduced production of DAG upon the induction of energy deficit and initiation of fat combustion instead of lipogenesis.[6] This type of energy overload–related insulin resistance would not be expected in this truly lean group. However, establishment of hepatic insulin resistance specifically in the lean subjects might require a more direct determination of the insulin sensitivity, including measurement of hepatic glucose production and lipogenesis. An interesting determination might be the fraction of palmitate in the liver-derived very-low-density lipoprotein as a marker of hepatic lipogenesis, which was recently linked to T2DM in studies of diabetes remission and relapse.[7]

In lean post-GDM women with glycemic disturbance, a potentially contributing metabolic factor despite normal triglycerides and high-density lipoprotein–cholesterol and liver fat content below 1% were markedly increased free fatty acids.[2] Since free fatty acids in lean subjects reflect peripheral adipose tissue lipolysis, this may indicate some degree of adipose tissue insulin resistance which is known to impair whole body and hepatic insulin sensitivity.

Remarkably, a similar lean phenotype without characteristics of the metabolic syndrome also represented a significant subgroup of participants with type 2 diabetes in the Whitehall II Study establishing the broader representation of this phenotype. To date, cross-sectional clustering studies still provide inconclusive data about lean subjects with glycemic impairment. In some, normal weight patients are at low risk for T2DM progression and cardiovascular disease, whereas others report a higher risk for rapid decline of beta-cell function and cardiovascular disease.[1]

In large diabetes prevention studies, subjects with combined impaired fasting glucose (IFG) and impaired glucose tolerance (IGT) were at higher likelihood of avoiding T2DM onset compared with patients with isolated IGT. Remarkably, metformin was much more effective in preventing the onset of diabetes in women with GDM than in other groups in the Diabetes Prevention Program over 15 years of follow-up, although the effectiveness was greatest in the more obese participants.[8]

The prediabetic lean women had a lower VO2 max suggesting less physical activity. Therefore, increased physical activity and dietary strategies lowering free fatty acids might be of interest. Subjects without combined glycemic impairment (ie, either isolated IFG or IGT, but not IFG-IGT) seem to respond more effectively to exercise. Overall, we lack clarity about the differential responsiveness of the identified prediabetes subtypes to specific treatments. Specifically, the lean and often young subgroup without metabolic syndrome may deserve more attention in future studies.