Non-autoimmune, lean diabetes in young people from Assam, India highlights the role of undernutrition in its aetiology – PHEnotypingNOrthEastINDianYoung type 2 diabetes (PHENOEINDY-2)
=========================================================================================================================================================================================

* Anupam Dutta
* Pranjal Kumar Dutta
* Sreemanta Madhab Baruah
* Prasanta Dihingia
* Arpita Ray
* Dattatrey S Bhat
* Sonali Wagle Patki
* Pradeep Tiwari
* Madhura Deshmukh
* Rucha Wagh
* Rubina Mulchandani
* Tanica Lyngdoh
* Sanjeeb Kakati
* Chittaranjan S Yajnik

## ABSTRACT

**Aims/hypothesis** Non-autoimmune diabetes in the young is believed to be driven by overweight-obesity and insulin resistance. However, it’s increasingly reported in undernourished people in the LMICs, including India. We hypothesized a role for undernutrition in the aetiology of non-autoimmune diabetes in impoverished young patients attending the Assam Medical College Hospital (AMCH), Dibrugarh (North-East India) for treatment.

**Methods** We measured socioeconomic status (SES), height and body mass index (BMI, kg/m2), waist and hip circumference, DXA body fat percent (adiposity), GAD antibodies, C-peptide and HOMA indices of β-cell function and insulin sensitivity in young patients (<40 years) with a presumptive diagnosis of T2D. Socio-economically matched volunteers with normal glucose tolerance (NGT, 75 gm OGTT) provided control data.

**Results** We studied 252 NGT controls (136 males, mean age 30 years, mean BMI 23.0 kg/m2) and 240 GAD-negative cases with diabetes (155 males, 36 years, 23.5 kg/m2). The majority of cases came from a relatively impoverished population of tribal people and tea garden employees. Twenty-eight percent of patients were stunted (male <161.2 cm m, female <149.8 cm), 14% were underweight (BMI <18.5 kg/m2) and 32% were overweight-obese (BMI >30 kg/m2). Height and BMI were positively associated with socio-economic status. On DXA, 61% of controls and 53% of cases were adipose (male >25%, female >35% body fat percent), conforming to the thin-fat Indian phenotype. Controls were relatively β-cell deficient (median HOMA-β 68.3) and insulin sensitive (HOMA-S 182.0) while cases were severely β-cell deficient (HOMA-β 25.7) and only moderately insulin resistant (HOMA-S 103). Cases with diabetes showed higher triglyceride and lower HDL-cholesterol concentrations. Thinner cases (lower BMI) were more hyperglycemic (FPG, HbA1c), insulin sensitive but B-cell deficient. As previously reported, two-thirds of these patients belonged to the severely insulin-deficient diabetes (SIDD) cluster of the Swedish diabetes subgroup classification.

**Conclusions/interpretation** Diabetes in the young in this impoverished population is heterogeneous but the majority of patients are not overweight-obese nor insulin-resistant. Diabetes is predominantly driven by β-cell deficiency. Our findings suggest a strong role for life-course undernutrition in the aetiology of non-autoimmune diabetes in this population.

**What is already known**

*   Non-autoimmune diabetes in young adult westerners is usually linked to obesity and insulin resistance, associated with ample lifelong nutrition.

*   In poorer parts of the world, young adult diabetes often occurs in lean individuals with evidence of past undernutrition.

*   There is sparse information on this condition

**What is the key question?**

*   Young adult, non-autoimmune diabetes is prevalent in growth-stunted people from a historically impoverished population from Assam, India. Can a detailed study of these patients improve our understanding of this disorder?

**New Findings**

*   Young lean adults had ‘type 2’ diabetes in association with evidence of stunted growth and the ‘thin-fat’, centrally obese phenotype, known to be linked with early life undernutrition.

*   Two-thirds belonged to the severely insulin-deficient diabetes (SIDD) subgroup of the Swedish classification, twice as common as in young patients with diabetes from Sweden.

**Impact on clinical practice**

*   Poverty and early-life undernutrition could play an important role in the aetiology of diabetes in many parts of the world.

Keywords
*   Young diabetes
*   lean
*   β-cell deficiency
*   Assam
*   India

## Introduction

Non-autoimmune diabetes is portrayed as a metabolic-endocrine disorder associated with aging, obesity, and a combination of insulin resistance and reduced insulin secretion. [1]. It is usually called type 2 diabetes (T2D). In developed countries, obesity and insulin resistance are considered drivers of this condition. The diabetes epidemic has rapidly grown in the world, especially in the low- and middle-income countries (LMICs), including in India which is now almost considered the world capital of diabetes [2, 3]. T2D is diagnosed at a much younger age and at a lower BMI in Indians compared to Europeans [4]. This may be partly genetically determined [5] but more likely due to their thin-fat (thrifty) phenotype which originates *in utero* due to the influence of maternal undernutrition and tracks through childhood [6, 7] and puberty into adult life [8, 9, 10, 11]. Most of the papers describing characteristics of Indian patients with diabetes are written from metropolitan cities and stress the role of obesogenic lifestyle and obesity-related insulin resistance [12]. The INDIAB study described heterogeneity of the phenotype of T2D across India and the GBD-India study highlighted that the most rapid rise in the prevalence of diabetes in the last 25 years has occurred in the relatively less developed parts of the country (Uttar Pradesh, Chhattisgarh, Odisha) [13, 14]. There are reports of lean T2D patients from the rural impoverished populations from the states of Odisha [15], Assam [16] and rural Tamil Nadu [17]. Such patients are also reported from African countries (Uganda and Ethiopia) [18, 19]. Increasing food insecurity in many parts of the world due to climate change, famines, floods, wars, and migrations will probably contribute to a higher prevalence of this phenotype across the world.

The North-East (NE) of India is unique in its ethnic diversity and a history of hardship due to ecological stresses dating back many hundreds of years [20]. The state of Assam has the lowest life expectancy at birth in the country (63.9y vs Indian average 67.9y), a low per capita income (US$ 1368 vs Indian average US$ 2191 per annum), and a low BMI (second only to the state of Jharkhand) [14] but showed an increase in the prevalence of diabetes in adults from 5.5% in 1990 to 7.7% in 2016 [13]. The Comprehensive National Nutrition Survey (CNNS) showed that the prevalence of prediabetes in children and adolescents is unexpectedly high in the relatively undernourished populations, including in some states from the North-East of India [21], and the Young Diabetes Registry (YDR) of India (diagnosed before 25 years of age) reported a relatively higher proportion of young onset T2D in Assam compared to other states [22].

We studied young patients with non-autoimmune diabetes (< 40 years of age) attending the outpatient clinic of the Assam Medical College, Dibrugarh. We reasoned that studying their physical and biochemical-endocrine characteristics will provide insight into pathophysiology and aetiology and guide appropriate anti-diabetic treatment for these patients who may otherwise be treated using Western treatment protocols.

## Materials and methods

### Study design

This is a hospital-based case-control study.

### Study setting

Patients were recruited from outpatients of the Department of Medicine, AMCH, Dibrugarh. AMCH is the first medical college in North-East India and was established in 1947 (formerly Berry White Medical School from 1901 to 1947). It is a tertiary medical referral centre for upper Assam and areas in neighboring states. It is the life-line for the poor and lower socioeconomic strata including indigenous, tribal, and tea garden workers. Consenting participants were recruited serially between May 2017 and May 2020.

### Sampling method and sample size

The sample size of 500 (cases + controls), to be recruited over a period of 3 years, was based on the average annual footfall of ∼1000 in the diabetes outpatient department of the AMCH over last five years and an estimated number of ∼10% of these patients being diagnosed before 40 years of age.

### Study population

Cases included people with diabetes attending the outpatients of the AMCH, diagnosed below the age of 40 years and clinically classified as T2D. Diabetes was diagnosed by the WHO/IDF criteria: fasting plasma glucose (FPG)> 126 mg/dl, post-prandial > 200 mg/dl, HbA1c >6.5%. They were enrolled in the PHENOIENDY-2 study if willing to be investigated for physical measurements and blood tests. We excluded patients clinically diagnosed with type 1 diabetes (diagnosed before 20 years of age, history of diabetic ketoacidosis, lifelong insulin treatment), fibro-calculous-pancreatic diabetes (clinical symptoms, and X-ray or sonographic evidence of chronic pancreatitis), clinically suspected to be MODY (Maturity Onset Diabetes in Young). We also excluded those with recent hospital admission for treatment of a medical condition. Controls included acquaintances, friends or colleagues of the patients from similar geographic and socio-economic background and if 75g OGTT showed normal glucose tolerance (NGT, ADA/IDF criteria).

### Study procedures and measurements

Participants were instructed to eat their normal dinner previous evening and visit the AMC Diabetes Research Unit in the morning after an overnight fast. Investigations included a health questionnaire, physical examination, anthropometric and body composition measurements, and blood tests. All measurements were made by trained and certified staff. The interviewer-assisted questionnaire collected information on socio-demographics, ethnicity, substance use and family history of diabetes. Smoking was defined as present if the participant reported to be current smoker and alcohol consumption if consumed three or more times a week. Socio-economic status was calculated using Standard of Living Index (SLI) score from the National Family Health Survey 2 (NFHS) [23].

Weight was measured using a calibrated digital scale (Euro Model EUP, Delhi, India) and height with a fixed stadiometer (SECA Model 755 1021994, Hamburg, Germany). Waist, hip and mid-arm circumferences were measured using a fiberglass measuring tape (SECA203, Model no. 2031717009, SECA GmbH & Co. KG, Hamburg, Germany). Skin-fold thickness (biceps, triceps, sub-scapular and supra-iliac) was measured using skin calipers (Harpenden, CMS Instruments Ltd, London). Mean of three readings at each site was used to calculate sum of the four skin folds. Body mass index (BMI) was calculated as weight divided by height in meter squared (kg/m2). Body composition was measured by Dual emission X-ray Absorptiometry (DXA) using Prodigy model (Lunar, Madison, WI, USA), following manufacturer’s guidelines. The software reports body composition as the mass and proportion (%) of fat, lean (muscle and organs), bone, and water.

BMI was categorized as underweight (<18.5 kg/m2), normal (18.5 – 24.9 kg/m2), overweight (25.0 – 29.9 kg/m2) and obese (>30.0 kg/m2); central obesity as waist-hip ratio (WHR) >0.90 in men and > 0.85 in women, and stunting as height <161.2 cm for men and <149.8 cm for women (corresponding to −2SD of height for adults in WHO standards) [24]. Adiposity was defined according to the American Association of Clinical Cardiology as body fat% > 25% in men and > 35% in women (DXA) [25].

Fasting venous blood sample was collected. In controls a 75g anhydrous oral glucose tolerance test (OGTT) was performed, and glucose tolerance classified as per WHO/IDF criteria into normal, prediabetes and diabetes mellitus [26]. Control participants with NGT provided data for comparison with patients with diabetes (cases). Clinical laboratory assays were performed at the Central Laboratory of AMCH. Plasma glucose, total cholesterol, high-density lipoprotein (HDL) cholesterol and triglycerides were measured using standard enzyme assays (Siemens Dimension RxL Auto analyzer (Vitros 5600, New Jersey, USA). HbA1c was measured by HPLC (Variant Hb typing system, BioRad, CA, USA).

Following measurements were made at the Diabetes Unit, KEM Hospital, Pune. Plasma C-peptide was measured by ELISA kit (DBC-Diagnostics, Biochem Canada Inc, Canada) calibrated against WHO-IS 84/510. GAD antibody concentration was measured in blood plasma of patients with diabetes by direct electrochemiluminescence immunoassay (Cobas C-peptide kit, Roche Diagnostics GmBH, Germany; lower detection limit: 3.3pmol/l with CV of 0.6% at 33 pmol/L)) those above 5 u/l were categorized as autoimmune diabetes. Plasma Immunoreactive Trypsinogen (IRT) concentrations were measured by ELISA (TRY ELISA kit, MyBioSource Inc., San Diego, CA, USA). Normal range for this measurement, as advised by kit manufacturers is <5 µ/ml (>=5 GAD positive). HOMA-β and HOMA-S were calculated using fasting glucose and C-peptide concentrations by Homeostasis Model Assessment (HOMA2) calculator ([https://www.dtu.ox.ac.uk/homacalculator/download.php](https://www.dtu.ox.ac.uk/homacalculator/download.php), last accessed in September 2021) [27]. This calculator excludes those with fasting glucose concentration of more than 25 mmol/L.

### Statistical Analysis

Our primary analysis was to describe physical, biochemical and endocrine characteristics of young patients with presumptive diagnosis of T2D attending AMCH and compare them with NGT controls. Adiposity and some biochemical markers differ by sex; therefore, males and females were analyzed separately. Continuous variables are summarized as median and 25th-75th centiles and categorical variables as number and percentages. Wilcoxon rank-sum test was used to compare continuous variables and Fisher’s exact test for discrete variables, by cases and controls. A p-value of 5% was considered statistically significant. The analysis included outcome variables that were continuous (BMI, WHR, Mid-Arm circumference, Skinfolds, DXA body fat %, Cholesterol and Triglycerides) and binary (Stunting, Central Obesity, Body fat %, and BMI categories). Univariable and Multivariable linear and logistic regression, respectively were used to assess independent effect of differences between cases and controls. Confounders included in the multivariable models were based on prior knowledge and scientific relevance. Results are reported as overall and stratified by sex. Modeling results were summarized using a Forest plot. Statistical Analysis was done using R 4.1.3.

### Ethical Considerations

This study was approved by the Institutional Ethics Committee, Assam Medical College (AMC/EC/932, 21-Mar-2017). A written, informed consent was obtained from all the participants before data collection. Patients were provided clinically relevant results and appropriately advised.

## Results

Between May 2017 and May 2020, we approached 300 people with diabetes attending the outpatients of the AMCH to request participation in the PHENOEINDY-2 study, based on the inclusion and exclusion criteria. Two hundred and seventy-nine patients with clinical diagnosis of T2D before 40 years of age agreed and 246 participated. Of these, 6 patients turned out to be positive for GAD antibodies and were diagnosed as autoimmune diabetes and excluded from further analysis. Two hundred and fifty-two controls participated and were investigated. The selection and flow of the study participants is presented in the flowchart (ESM fig 1). Missing data ranged from 1-3% for demographic and anthropometric variables, 5-9% for biochemical variables, and 12% for body composition (DXA) variables. Four patients with diabetes having fasting plasma glucose concentration > 25 mmol/L were excluded from HOMA calculation because of the limit imposed by the HOMA2 calculator.

![Figure 1:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/02/05/2024.02.02.24302249/F1.medium.gif)

[Figure 1:](http://medrxiv.org/content/early/2024/02/05/2024.02.02.24302249/F1)

Figure 1: 
Comparison of body size (BMI), adiposity (DXA fat %), beta cell function (HOMA-B) and insulin sensitivity (HOMA-S) between controls (NGT) and patients with type 2 diabetes (T2DM) from Assam Medical College Hospital, Dibrugarh, India. Significance of difference between groups is shown as: ns (not significant),

*p<0.05, **p<0.01 and \***|p<0.001.

### Socio-demographic and clinical

ESM table 1 describes socio-demographic profile of the participants. Most were Assamese speaking rural folk and there was a balanced representation of religion (Hindu, Muslim or other) and ethnic groups (Caucasoid, Mongoloid or Australoid). Both cases and controls came from relatively low socio-economic group (low SLI), many were tribal people and tea garden workers. Cases were less educated, and consumed alcohol more frequently but had a similar frequency of smoking compared to controls.

View this table:
[Table 1:](http://medrxiv.org/content/early/2024/02/05/2024.02.02.24302249/T1)

Table 1: Summary of baseline characteristics (Demographic, Anthropometric, and biochemical) and comparison between cases and controls stratified by sex

The average age at diagnosis of diabetes was 36 years, and cases were older than controls at the time of study (Table 1). The average duration of diabetes since diagnosis was 1.8 years for men (min < 1.0 – max 5.0 years) and 4.0 years for women (1.1-7.0 years). Cases had a higher prevalence of family history of diabetes in 1st degree relatives compared to that in controls [55% vs 29%]. Twenty (8%) cases received insulin treatment and 149 (62%) oral anti-diabetic agents; 5 (2%) received anti-hypertensive and 2 (1%) cholesterol-lowering drugs.

### Anthropometry and body composition

As a group, participants in our study (cases + controls) were short, and 110 (22%) were stunted by WHO criteria. Average BMI was 23 kg/m2: 56 (11%) were underweight, 276 (56%) normal weight, and 157 (32%) overweight/obese (27% overweight and 5% obese). On the other hand, DXA measurements showed that 64% men and 66 % women were adipose. (Table 1).

Cases were shorter, difference was significant for males (median height: 165 vs 167 cm, p=0.005) and had smaller mid-arm circumference compared to controls, there was no difference in the BMI (Table 1). Male cases had similar DXA-measured total body fat percent compared to controls (including regional measurements) while female cases had a significantly lower body fat percent compared to the controls (contributed by lower fat in arms, legs, and gynoid region) (ESM table 4, Fig. 1). Notwithstanding these findings, both male and female cases had a significantly higher waist-hip ratio compared to controls.

### Metabolic-endocrine measurements

Most patients with diabetes had unsatisfactory glucose control reflected in high glucose and HbA1c concentrations (Table 1); only 40 (17%) had HbA1c <7.0%. Cases had higher triglyceride and lower HDL-cholesterol concentrations compared to the controls. Sixty-four (27%) of cases had hypercholesterolemia (>5.2 mmol/L), 139 (58%) hypertriglyceridemia (>1.7 mmol/L), low HDL cholesterol concentrations were observed in 96 (62%) male (<1.0 mmol/L), and 42 (76%) female (<1.3 mmol/L) patients.

HOMA2 estimates of B-cell function and insulin sensitivity are in reference to those in healthy British individuals who are referenced at 100 for both measures. The young Assamese healthy NGT controls were more insulin sensitive (HOMA-S approximately twice the reference) and had noticeably lower B-cell function (HOMA-β approximately two-thirds of the reference).

Fasting plasma C-peptide concentrations were low in cases, 69 (29%) cases had concentration <0.5 nmol/L, though none had a history of diabetic ketoacidosis. Compared to NGT Assamese controls, cases had lower HOMA-β and HOMA-S (Fig. 1). The deficit in HOMA-β was larger (on average ∼70% lower than that in Assamese controls) compared to that in HOMA-S (∼ 50% lower). Figure 2 shows the hyperbolic plot of HOMA-S and HOMA-β which highlights the predominant reduction of HOMA-β in cases compared to controls.

![Figure 2:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/02/05/2024.02.02.24302249/F2.medium.gif)

[Figure 2:](http://medrxiv.org/content/early/2024/02/05/2024.02.02.24302249/F2)

Figure 2: 
Comparison of the hyperbolic relationship between HOMA-S and HOMA-B in controls (NGT)and patients with type 2 diabetes (T2DM). The plot shows lower HOMA-B as the predominant defect in patients with type 2 diabetes compared to NGT, and relatively less contribution of HOMA-S.

### Association of BMI with metabolic-endocrine parameters (ESM Table 2, 3 and Fig 3)

Thirty-four (14%) cases were underweight, 127 (53%) normal weight, and 76 (32%) were overweight/obese. Even in this population with relatively low socio-economic status, there was a significant positive association between BMI and socioeconomic status (r= ∼0.35, p<0.001), both in controls and cases. There was a graded association across BMI categories (underweight, normal weight, overweight/obese) of metabolic and endocrine measurements. Underweight cases had the lowest adiposity (skinfolds and DXA fat%), WHR and plasma triglyceride concentration, however, they had the highest glycemia (fasting glucose and HbA1c) and HDL-cholesterol concentrations (ESM tables 2a and 2b, Fig. 3). Underweight cases also had the lowest HOMA-β and highest HOMA-S (Fig. 3). A third of underweight patients were receiving insulin treatment compared to 10% of the overweight-obese group.

![Figure 3:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/02/05/2024.02.02.24302249/F3.medium.gif)

[Figure 3:](http://medrxiv.org/content/early/2024/02/05/2024.02.02.24302249/F3)

Figure 3: 
Comparison of body size (Height), SLI score, glycemia (FPG and HbA1c), beta cell function (HOMA-B) and insulin sensitivity (HOMA-S) by BMI categories [Underweight (BMI<18.5), Normal (BMI: 18.5 – 24.9) and Overweight/Obese (BMI>=25)] in patients with type 2 diabetes (T2DM) and controls (NGT) from Assam Medical College Hospital, Dibrugarh, India. Significance of trend across BMI categories is shown as: ns (not significant), *p<0.05, **p<0.01 and

\***|p<0.001, adjusted for age, duration of diabetes in case, and age in controls.

View this table:
[Table 2:](http://medrxiv.org/content/early/2024/02/05/2024.02.02.24302249/T2)

Table 2: ORs/Mean difference of characteristics in cases compared to controls stratified by sex

### MLRA (multiple linear regression analysis) (Table 2, Fig 4)

![](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/02/05/2024.02.02.24302249/F4/graphic-8.medium.gif)

[](http://medrxiv.org/content/early/2024/02/05/2024.02.02.24302249/F4/graphic-8)

![](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/02/05/2024.02.02.24302249/F4/graphic-9.medium.gif)

[](http://medrxiv.org/content/early/2024/02/05/2024.02.02.24302249/F4/graphic-9)

Figure 4: 
A: Categorical Variable Forest Plot

B: Continuous Variable Forest Plot

We performed a ‘multivariable’ analysis to investigate differences between cases and controls in SES, anthropometric and metabolic-endocrine characteristics, adjusting for confounding factors like age, sex, locality and SES. This analysis revealed that compared to the controls, the cases were deprived (lower SES). After adjusting for SES, they were stunted, had lower mid-arm circumference, were centrally obese (WHR), and had higher circulating triglycerides; female patients also had higher circulating cholesterol concentrations. In addition, cases had a lower HOMA-β and HOMA-S compared to the controls.

## Discussion

We describe spectrum of socio-demographic, physical and metabolic-endocrine characteristics of GAD-antibody negative young patients with diabetes amongst tribal people and tea garden workers in Assam, India. Our comprehensive description of body size measurements, body composition using DXA, detailed metabolic measurements, and comparison with local controls, allows appreciation of a unique phenotype of people with diabetes in a multi-generationally undernourished population. These cases with diabetes were shorter and had lower lean mass in the limbs compared to controls who themselves were short and thin. These findings suggest linear growth restriction (both intra-uterine and post-natal), presumably due to undernutrition (in addition to genetic factors). Underweight in a sizeable proportion (14% cases and 8.7 % controls) is also supportive of undernutrition in this community. Despite these features, both cases and controls had a high prevalence of adiposity (DXA-measured body fat percent) and a very high prevalence of central obesity (WHR), conforming to the characteristic thin-fat Indian phenotype described by us [4, 9, 28]. Our previous research has shown that these characteristics originate during intrauterine life (‘thin-fat’ baby) and track into childhood, puberty and adult life [8, 9, 10, 11]. Diabetes in these cases was predominantly ascribable to reduced beta cell function with a relatively preserved insulin sensitivity. We ruled out autoimmunity to pancreatic β-cells and exocrine pancreatic affection by specific measurements. Though relatively thin and β-cell deficient these cases had higher triglyceride and lower HDL concentrations compared to controls, dyslipidemia characteristic of diabetes which is usually called T2D. Thus, physical and biochemical-endocrine characteristics of these patients substantially differ from the conventional descriptions of young obese and insulin resistant T2D patients from popular textbooks and guidelines (29), and also those of urban Indians from tertiary centers (12), and aboriginal populations from different parts of the world [30, 31, 32]. Comparison with Swedish young onset T2D cases using the Swedish T2D subgroup clustering algorithm revealed that the 67% of cases from Assam belonged to the Severely Insulin Deficient Diabetes (SIDD) cluster and 23% to the Mild Obesity related Diabetes (MOD) cluster; opposite of the proportion in Swedish cases (26% SIDD and 68% MOD) [33]. In that study, Assamese patients were the youngest, thinnest, and most β-cell deficient compared to cases from the relatively more affluent cities of Pune (Maharashtra) and Ahmedabad (Gujrat), thus representing a polar undernourished phenotype of young T2D. It is possible that a proportion of underweight and insulin deficiency in these patients could be attributed to effects of uncontrolled long-standing metabolic disturbances (’glucotoxicity’ and ‘lipotoxicity’). However, the fact that similar patterns of weight, body composition and B-cell function were present even in NGT controls from this population suggests that glucotoxicity cannot be the major driver.

A bit unexpectedly, a third of these patients were overweight-obese. This was partly attributable to higher socio-economic status (as measured by SLI, see Fig 3) but it is important to remember that intrauterine undernutrition could also promote overweight-obesity in later life [34]. This was described in Dutch Winter Hunger studies, if exposure to undernutrition is in the first two trimesters of pregnancy, presumably through programming of the neuro-endocrine mechanisms that promote hyperphagia in later life [35, 36]. A multigenerationally undernourished rat becomes obese within two generations of ‘normal’ chow feeding, and its beta cells continued to show the epigenetic marks of multigenerational undernutrition [37]. Intriguingly, in the contemporary NHANES cohorts in the USA effect of early life undernutrition is evident in 1) increased risk of prediabetes in adolescents who had lower birth weight [38], and 2) increased risk of diabetes in adults who were shorter and had shorter upper leg length (reflecting early life undernutrition) [39, 40]. Despite such evidence, attention is usually focused on later-life obesity as the sole risk factor for diabetes rather than intergenerational and life-course undernutrition which will inform strategies for ‘primordial’ prevention. In this context, diabetes may be looked upon as a result of interaction between reduced ‘capacity’ due to developmental undernutrition causing lower β-cell mass and lean mass, and the ‘load’ of insulin resistance due to developmental adiposity (‘thin-fat’ phenotype) aggravated by later-life overnutrition related to socio-economic development [41]. The Pune Maternal Nutrition Study traced the life-course evolution of glucose intolerance at 18 years of age in children born in the cohort [42]. There was a progressive decrease in insulin sensitivity and an increase in compensatory β-cell function from childhood to young adult age. Prediabetes was predicted by intrauterine growth restriction (stunting at birth and smaller head circumference) and reduced compensatory insulin secretion from early childhood (6 years of age). Role of β cell deficiency in the etiology of non-autoimmune diabetes in Indians has been stressed and supported by recent studies in thin adult T2D patients in India [33, 43], in migrant Indians [44]. Lean T2D patients in Uganda and Ethiopia show a similar pathophysiology [18, 19]. These findings drive home the important contribution of social determinants of health (SDOH) in the heterogeneity of T2D phenotypes across the world and stress the need to address relevant SDG goals (1 poverty, 2 hunger, and 10 equality) to prevent and manage diabetes in these patients. These issues are increasingly relevant in the modern world torn by conflict, displacement and climate change.

Lack of excess adiposity in young Assamese cases with diabetes compared to NGT controls is a novel finding and is over and above the fact that Assamese participants, as a group, were thin but adipose. This lack of extra adiposity might be partly attributed to fat loss due to uncontrolled diabetes and suggests to us a need to look for inapparent ectopic fat (in liver and pancreas) which could contribute to the pathophysiology of diabetes. Intriguingly, potbelly, fatty liver, and shrunken pancreas were described in undernourished children in India [45] and other countries [46] and support the thesis that early life undernutrition could contribute to this phenotype. Developmental Origins of Health and Disease (DOHaD) paradigm proposes that lifestyle-related diseases are ‘programmed’ epigenetically at the time of conception by the interaction between genes and the intra-uterine environment [47]. Assamese mothers had a higher prevalence of underweight (25.9%) compared to the Indian national average (22.9%) and a high prevalence of low birth weight and under-five undernutrition [48], supporting the ‘DOHaD’ explanation for aetiology of diabetes in our cases.

To the best of our knowledge, ours is the largest report of young onset non-autoimmune diabetes in Northeast India. Strengths of this study include the availability of detailed anthropometric and DXA measurements of body composition, and epidemiological indices of β-cell function and insulin sensitivity (HOMA). Patients in our hospital-based study may not be representative of the general population, however, our study planned to describe the unique phenotype of diabetes in this population and search for mechanistic explanations. The cross-sectional design precludes causality attribution to undernutrition; however, it is supported by birth cohort research in India and stresses the need for longitudinal studies. A study of body composition and glucose-insulin metabolism in children born in AMCH is in progress and aspires to develop into a longitudinal cohort (ICMR Grant - No.55/3/2/Brain Storming Sessions/diab/2017-NCD-II).

In conclusion, we report characteristics of young onset, non-autoimmune, lean cases of diabetes in upper Assam. These patients have a thin physique due to poor lean mass, central obesity, and excess adiposity for their thinness but lack excess adiposity compared to controls. Their metabolic profile includes severe hyperglycemia, a characteristic T2D dyslipidemia, dominant insulin secretory failure, and moderate insulin insensitivity. Such a metabolic profile would predispose these patients to micro- and macro-vascular complications at the peak of their productive lives. The implications are far-reaching, both for the individual patients and for the resource-constrained health system. Our data suggest a need to revise population health strategies that might help prevent diabetes (primordial) and reduce morbidity and mortality. Genetic and early life nutrition and growth studies will help understand the etiological factors of diabetes in this population to guide prevention.

## Supporting information

This is a Electronic Supplemental File with figures and tables cited in the main text [[supplements/302249_file02.pdf]](pending:yes)

## Funding

This study was funded by Indian Council of Medical Research (ICMR study No. 55/3/1/Brain Storming Session/Diab./2017-NCD-II, 29-Mar-2017).

## Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

## Author Contribution

AD, CSY and SK contributed to the conception of the work. AD, PKD, PD, SMB and AR contributed to data collection in Dibrugarh, Assam, India. DSB, SWP, MD, RM, TL, and RW contributed to the data analysis. CSY, AD, RM, TL and RW drafted the article. All authors contributed to the interpretation of data and critical revision of the article. All authors gave final approval of the version to be published. AD, CSY and SK are the guarantors of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

## Acknowledgements

We thank all patients and healthcare providers for their support and willingness to participate. We thank H. Kashyap, H. Kaur, A. Gogoi, A. Boruah, D. Boruah, S. Sonowal, P. Boruah, R. Nath, I. Gogoi, P. Gogoi, D. Ahmed (Assam Medical College and Hospital), P. Yajnik, R. Ladkat (Diabetes Unit, Kamalnayan Bajaj Diabetology Research Centre, King Edward Memorial Hospital and Research Centre, Pune, India) and Dr L. Saikia, Dr R. Nath (Multidisciplinary Research Unit, Assam Medical College and Hospital) for their invaluable contribution to this study. We thank Nikhil Gupte, PhD, for advice on statistical analysis.

## Footnotes

*   No role of funding agencies in writing of the manuscript. The authors declare that there are no relationships or activities that might bias, or be perceived to bias, their work.

*   The author list has been reordered.

*   Received February 2, 2024.
*   Revision received February 4, 2024.
*   Accepted February 5, 2024.


*   © 2024, Posted by Cold Spring Harbor Laboratory

The copyright holder for this pre-print is the author. All rights reserved. The material may not be redistributed, re-used or adapted without the author's permission.

## References

1.  1.American Diabetes Association. (2023). 2. Classification and diagnosis of diabetes: standards of care in diabetes—2023. Diabetes care, 46, S19–S40. doi:10.2337/dc23-S002
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.2337/dc23-S002&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=36507649&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 

2.  2.International Diabetes Federation (IDF) Atlas 2021. [www.diabetesatlas,org](http://www.diabetesatlas,org).
    
    

3.  3.Anjana, R. M., Unnikrishnan, R., Deepa, M., et al., ICMR-INDIAB Collaborative Study Group (2023). Metabolic non-communicable disease health report of India: the ICMR-INDIAB national cross-sectional study (ICMR-INDIAB-17). The lancet. Diabetes & endocrinology, 11(7), 474–489. doi:10.1016/S2213-8587(23)00119-5.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S2213-8587(23)00119-5&link_type=DOI) 

4.  4.Yajnik C. S. (2004). Early life origins of insulin resistance and type 2 diabetes in India and other Asian countries. The Journal of nutrition, 134(1), 205–210. doi:10.1093/jn/134.1.205.
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6OToibnV0cml0aW9uIjtzOjU6InJlc2lkIjtzOjk6IjEzNC8xLzIwNSI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDI0LzAyLzA1LzIwMjQuMDIuMDIuMjQzMDIyNDkuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

5.  5.Loh, M., Zhang, W., Ng, H. K., et al. (2022). Identification of genetic effects underlying type 2 diabetes in South Asian and European populations. Communications biology, 5(1), 329. doi:10.1038/s42003-022-03248-5.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/s42003-022-03248-5&link_type=DOI) 

6.  6.Yajnik C. S. (2001). The insulin resistance epidemic in India: fetal origins, later lifestyle, or both? Nutr rev, 59(1 Pt 1), 1–9. doi:10.1111/j.1753-4887.2001.tb01898.x
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1111/j.1753-4887.2001.tb01898.x&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=11281246&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 

7.  7.Yajnik, C. (2019). The story of the hungry Indian foetus. Bulletin of the Nutrition Foundation of India, 40(3).
    
    

8.  8.Yajnik, C. S., Fall, C. H., Coyaji, et al. (2003) Neonatal anthropometry: the thin-fat Indian baby. The Pune Maternal Nutrition Study. Int J Obes (Lond), 27(2), 173–180. doi:10.1038/sj.ijo.802219
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/sj.ijo.802219&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=12586996&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000181032700003&link_type=ISI) 

9.  9.Yajnik, C. S., & Yudkin, J. S. (2004). The Y-Y paradox. Lancet (London, England), 363(9403), 163. doi:10.1016/S0140-6736(03)15269-5.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S0140-6736(03)15269-5&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=14726172&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000187939200022&link_type=ISI) 

10. 10.Lakshmi S, Yajnik C, Metcalf B, et al. (2008). How far does body composition explain the difference in insulin resistance between the ‘Thin-Fat’ Indian and ‘Fat-Fat’ Caucasian child? Diabetes, Vol. 57, Pp. A745-A746. 1701 N Beauregard St, Alexandria, Va 22311-1717 USA: Amer Diabetes Assoc.
    
    

11. 11.D’Angelo, S., Yajnik, C. S., Kumaran, K., et al. & SWS Study Group and the PMNS Study Group (2015). Body size and body composition: a comparison of children in India and the UK through infancy and early childhood. Journal of epidemiology and community health, 69(12), 1147–1153. doi:10.1136/jech-2014-204998.
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoiamVjaCI7czo1OiJyZXNpZCI7czoxMDoiNjkvMTIvMTE0NyI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDI0LzAyLzA1LzIwMjQuMDIuMDIuMjQzMDIyNDkuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

12. 12.Misra A, Bhardwaj S. Obesity and the metabolic syndrome in developing countries: focus on South Asians. International nutrition: achieving millennium goals and beyond 2014 (Vol. 78, pp. 133–140). Karger Publishers.
    
    

13. 13.India State-Level Disease Burden Initiative Diabetes Collaborators (2018). The increasing burden of diabetes and variations among the states of India: The Global Burden of Disease Study 1990-2016. Lancet Glob Health, 6(12), e1352–e1362. doi:10.1016/S2214-109X(18)30387-5
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S2214-109X(18)30387-5&link_type=DOI) 

14. 14.Anjana, R. M., Deepa, M., Pradeepa, R., et al. ICMR–INDIAB Collaborative Study Group (2017). Prevalence of diabetes and prediabetes in 15 states of India: results from the ICMR-INDIAB population-based cross-sectional study. Lancet Diabetes Endocrinol, 5(8), 585–596. doi:10.1016/S2213-8587(17)30174-2
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S2213-8587(17)30174-2&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 

15. 15.Das, S., Samal, K. C., Baliarsinha, A. K., & Tripathy, B. B. (1995). Lean (underweight) NIDDM - peculiarities and differences in metabolic and hormonal status - a pilot study. The Journal of the Association of Physicians of India, 43(5), 339–342.
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=9081965&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 

16. 16.Praveen, P. A., Madhu, S. V., Viswanathan, et al. (2021). Demographic and clinical profile of youth onset diabetes patients in India-Results from the baseline data of a clinic-based registry of people with diabetes in India with young age at onset-[YDR-02]. Pediatr Diabetes, 22(1), 15–21. doi:10.1111/pedi.12973
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1111/pedi.12973&link_type=DOI) 

17. 17.Lontchi-Yimagou, E., Dasgupta, R., Anoop, S., et al. (2022). An Atypical Form of Diabetes Among Individuals with Low BMI. Diabetes care, 45(6), 1428–1437. doi:10.2337/dc21-1957.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.2337/dc21-1957&link_type=DOI) 

18. 18.Kibirige, D., Sekitoleko, I., Lumu, et. al. (2022). Understanding the pathogenesis of lean non-autoimmune diabetes in an African population with newly diagnosed diabetes. Diabetologia, 65(4), 675–683. doi:10.1007/s00125-021-05644-8
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/s00125-021-05644-8&link_type=DOI) 

19. 19.Alemu, S., Dessie, A., Seid, E., et al. (2009). Insulin-requiring diabetes in rural Ethiopia: should we reopen the case for malnutrition-related diabetes? Diabetologia, 52(9), 1842–1845. doi:10.1007/s00125-009-1433-5.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/s00125-009-1433-5&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=19565213&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 

20. 20.Guite J. (2018) Against State, against History: Freedom, Resistance, and Statelessness in Upland Northeast India. Oxford University Press.
    
    

21. 21.Ministry of Health and Family Welfare (MoHFW), Government of India, UNICEF and Population Council. (2019) Comprehensive National Nutrition Survey (CNNS) National Report 2016-2018. New Delhi.
    
    

22. 22.Praveen, P. A., Madhu, S. V., Mohan, V., et al. (2016) Registry of Youth Onset Diabetes in India (YDR): Rationale, Recruitment, and Current Status. J Diabetes Sci Technol, 10(5), 1034–1041. doi:10.1177/1932296816645121
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1177/1932296816645121&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=27179010&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 

23. 23.National Family Health Survey (NFHS-2) 1998. [http://rchiips.org/nfhs/nfhs2.shtml](http://rchiips.org/nfhs/nfhs2.shtml) (accessed January 5, 2022).
    
    

24. 24.World Health Organization. Growth reference data for 5-19 years. Accessed 1 January 2022. Available from [https://www.who.int/growthref/who2007\_height\_for\_age/en/](https://www.who.int/growthref/who2007_height_for_age/en/).
    
    

25. 25.Dickey RA, Bartuska DG, Bray GW, et al. (1998) AACE/ACE Position Statement on The Prevention, Diagnosis, And Treatment of Obesity (1998 Revision).
    
    

26. 26.World Health Organization, International Diabetes Federation. (2006) Definition and diagnosis of diabetes mellitus and intermediate hyperglycaemia: report of a WHO/IDF consultation.
    
    

27. 27.The HOMA2 Calculator ©The University of Oxford 2004. [https://www.dtu.ox.ac.uk/homacalculator/download.php](https://www.dtu.ox.ac.uk/homacalculator/download.php) (accessed January 10, 20221).
    
    

28. 28.Shelgikar, K. M., Hockaday, T. D., & Yajnik, C. S. (1991). Central rather than generalized obesity is related to hyperglycaemia in Asian Indian subjects. Diabet Med, 8(8), 712–717. doi:10.1111/j.1464-5491.1991.tb01689.x
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1111/j.1464-5491.1991.tb01689.x&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=1838061&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1991GH94900003&link_type=ISI) 

29. 29.American Diabetes Association Professional Practice Committee; 14. Children and Adolescents: Standards of Care in Diabetes—2024. Diabetes Care 1 January 2024; 47 (Supplement_1): S258–S281. doi:10.2337/dc24-S014
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.2337/dc24-S014&link_type=DOI) 

30. 30.Wedekind, L. E., Mitchell, C. M., Andersen, C. C., Knowler, W. C., & Hanson, R. L. (2021). Epidemiology of Type 2 Diabetes in Indigenous Communities in the United States. Current diabetes reports, 21(11), 47. doi:10.1007/s11892-021-01406-3.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/s11892-021-01406-3&link_type=DOI) 

31. 31.Hare, M. J. L., Zhao, Y., Guthridge, S., et al. (2022). Prevalence and incidence of diabetes among Aboriginal people in remote communities of the Northern Territory, Australia: a retrospective, longitudinal data-linkage study. BMJ open, 12(5), e059716. doi:10.1136/bmjopen-2021-059716.
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NzoiYm1qb3BlbiI7czo1OiJyZXNpZCI7czoxMjoiMTIvNS9lMDU5NzE2IjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjQvMDIvMDUvMjAyNC4wMi4wMi4yNDMwMjI0OS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

32. 32.Trewin D. National Aboriginal and Torres Strait Islander Health Survey, Australia, 2004-05. Canberra: Australian Bureau of Statistics; 2006.
    
    

33. 33.Prasad, R. B., Asplund, O., Shukla, S. R., et al. (2022). Subgroups of patients with young-onset type 2 diabetes in India reveal insulin deficiency as a major driver. Diabetologia, 65(1), 65–78. doi:10.1007/s00125-021-05543-y
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/s00125-021-05543-y&link_type=DOI) 

34. 34.Stein, A. D., Obrutu, O. E., Behere, R. V., & Yajnik, C. S. (2019). Developmental undernutrition, offspring obesity and type 2 diabetes. Diabetologia, 62(10), 1773–1778. doi:10.1007/s00125-019-4930-1
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/s00125-019-4930-1&link_type=DOI) 

35. 35.Ravelli, G. P., Stein, Z. A., & Susser, M. W. (1976). Obesity in young men after famine exposure in utero and early infancy. The New England journal of medicine, 295(7), 349–353. doi:10.1056/NEJM197608122950701
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1056/NEJM197608122950701&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=934222&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1976CA08900001&link_type=ISI) 

36. 36.Skowronski, A. A., Leibel, R. L., & LeDuc, C. A. (2023). Neurodevelopmental programming of adiposity: contributions to obesity risk. Endocrine reviews, bnad031. Advance online publication. doi:10.1210/endrev/bnad031
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1210/endrev/bnad031&link_type=DOI) 

37. 37.Hardikar, A. A., Satoor, S. N., Karandikar, M. S., et al. (2015). Multigenerational Undernutrition Increases Susceptibility to Obesity and Diabetes that Is Not Reversed after Dietary Recuperation. Cell metabolism, 22(2), 312–319. doi:10.1016/j.cmet.2015.06.008
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.cmet.2015.06.008&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=26166746&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 

38. 38.Sanjeevi, N., & Freeland-Graves, J. H. (2022). Birth weight and prediabetes in a nationally representative sample of US adolescents. Clinical obesity, 12(2), e12504. doi:10.1111/cob.12504
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1111/cob.12504&link_type=DOI) 

39. 39.Asao, K., Kao, W. H., Baptiste-Roberts, K., Bandeen-Roche, K., Erlinger, T. P., & Brancati, F. L. (2006). Short stature and the risk of adiposity, insulin resistance, and type 2 diabetes in middle age: the Third National Health and Nutrition Examination Survey (NHANES III), 1988-1994. Diabetes care, 29(7), 1632–1637. doi:10.2337/dc05-1997
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NzoiZGlhY2FyZSI7czo1OiJyZXNpZCI7czo5OiIyOS83LzE2MzIiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyNC8wMi8wNS8yMDI0LjAyLjAyLjI0MzAyMjQ5LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 

40. 40.Weitzman, S., Wang, C. H., Pankow, J. S., Schmidt, M. I., & Brancati, F. L. (2010). Are measures of height and leg length related to incident diabetes mellitus? The ARIC (Atherosclerosis Risk in Communities) study. Acta diabetologica, 47(3), 237–242. doi:10.1007/s00592-009-0145-0
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/s00592-009-0145-0&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=19774332&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000280372600008&link_type=ISI) 

41. 41.Wells, J. C., Pomeroy, E., Walimbe, S. R., Popkin, B. M., & Yajnik, C. S. (2016). The Elevated Susceptibility to Diabetes in India: An Evolutionary Perspective. Frontiers in public health, 4, 145. doi:10.3389/fpubh.2016.00145
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.3389/fpubh.2016.00145&link_type=DOI) 

42. 42.Yajnik, C. S., Bandopadhyay, S., Bhalerao, A., et al. (2021). Poor In Utero Growth, and Reduced β-Cell Compensation and High Fasting Glucose from Childhood, Are Harbingers of Glucose Intolerance in Young Indians. Diabetes care, 44(12), 2747–2757. doi:10.2337/dc20-3026
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NzoiZGlhY2FyZSI7czo1OiJyZXNpZCI7czoxMDoiNDQvMTIvMjc0NyI7czo0OiJhdG9tIjtzOjUwOiIvbWVkcnhpdi9lYXJseS8yMDI0LzAyLzA1LzIwMjQuMDIuMDIuMjQzMDIyNDkuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

43. 43.Siddiqui, M. K., Anjana, R. M., Dawed, A. Y., et al. (2022). Young-onset diabetes in Asian Indians is associated with lower measured and genetically determined beta cell function. Diabetologia, 10.1007/s00125-022-05671-z. Advance online publication. doi:10.1007/s00125-022-05671-z
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/s00125-022-05671-z&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=35247066&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 

44. 44.Gujral, U. P., & Kanaya, A. M. (2021). Epidemiology of diabetes among South Asians in the United States: lessons from the MASALA study. Annals of the New York Academy of Sciences, 1495(1), 24–39. doi:10.1111/nyas.14530
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1111/nyas.14530&link_type=DOI) 

45. 45.Ramalingaswami V. (1964). Perspectives in protein malnutrition. Nature, 201, 546–551. doi:10.1038/201546a0
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1038/201546a0&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=14160638&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 

46. 46.Truswell AS, Hansen JD. (1969) Fatty liver in protein-calorie malnutrition. S Afr Med J,43(10):280–283.
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=5786639&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 

47. 47.Fleming, T. P., Watkins, A. J., Velazquez, M. A., et al. (2018). Origins of lifetime health around the time of conception: causes and consequences. Lancet (London, England), 391(10132), 1842–1852. doi:10.1016/S0140-6736(18)30312-X
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/S0140-6736(18)30312-X&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=29673874&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F02%2F05%2F2024.02.02.24302249.atom) 

48. 48.International Institute for Population Sciences (IIPS) and ICF. (2017) National Family Health Survey (NFHS-4), 2015-16: India. Mumbai: IIPS