{"paper_id":"297e1ee8-4d7b-4e84-bb36-336e4cba82ce","body_text":"Partitioned polygenic scores highlight role of beta-cell function and unfavourable fat distribution patterns in young onset type 2 diabetes in south Asians | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Partitioned polygenic scores highlight role of beta-cell function and unfavourable fat distribution patterns in young onset type 2 diabetes in south Asians Moneeza Siddiqui, Sam Hodgson, Alice Williamson, Daniel Stow, and 10 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4027509/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 26 Nov, 2024 Read the published version in Nature Medicine → Version 1 posted You are reading this latest preprint version Abstract South Asians experience a higher risk of early-onset Type 2 diabetes (T2D) with normal BMI. However, since genetic research is largely focussed on white Europeans, the reasons for this are poorly understood. We used 12 recently derived multi-ancestry partitioned polygenic risk scores (pPS) to identify the aetiological pathways underlying T2D, gestational diabetes mellitus (GDM), earlier onset, progression to complications and insulin dependence, and treatment response in a south Asian cohort. Using electronic health record and genetic data from 51,108 British Pakistani and Bangladeshi individuals with T2D (n = 11,673) and GDM (n = 1,965) in the Genes & Health study, we explored associations between pPS, T2D, GDM, diabetes complications, and treatment response using sex- and ancestry-adjusted multivariable regression and Cox proportional-hazards models. A pPS representing insulin deficiency was most strongly associated with T2D per standard deviation (OR: odds ratio):1.46, 95%CI:1.42–1.50), GDM (OR:1.27, 95%CI: 1.20–1.34) and age at T2D diagnosis (beta = -1.7 years, 95%CI: -1.5 to -1.9), followed by a pPS representing an unfavourable fat distribution (lipodystrophy). Individuals at high genetic risk of both insulin deficiency and lipodystrophy were diagnosed with T2D 8.2 years earlier with BMI 3 kg/m 2 lower compared to those at low genetic risk. The insulin deficiency pPS was associated with poorer response to metformin, thiazolidinediones, and SGLT2 inhibitors (post-treatment HbA1c increased from baseline by 0.51%, 1.83%, and 1.13% respectively). Higher Insulin deficiency and lipodystrophy pPS were also associated with faster progression to insulin dependence and microvascular complications. Using UK Biobank, we found that south Asians had a greater genetic burden of both these pPS compared to white Europeans. In British Pakistani and Bangladeshi individuals, genetic predisposition to insulin deficiency and lipodystrophy helps identify individuals at risk of earlier onset of type 2 diabetes, who progress faster to complications and insulin dependence, and are less likely to respond to standard diabetes management pathways. Health sciences/Medical research/Genetics research Health sciences/Endocrinology/Endocrine system and metabolic diseases/Diabetes/Type 2 diabetes Health sciences/Pathogenesis/Clinical genetics/Disease genetics Health sciences/Diseases/Endocrine system and metabolic diseases/Diabetes/Type 2 diabetes Health sciences/Risk factors Figures Figure 1 Figure 2 Figure 3 Figure 4 Background Type 2 diabetes is common, particularly in South Asian individuals, among whom the prevalence is estimated to be 12.7% globally, and as high as 30% in Pakistan 1 , 2 . Compared to white Europeans, South Asians tend to be diagnosed with type 2 diabetes at younger ages and with lower Body Mass Index (BMI) 2 , 3 . This phenomenon exists even when the environment isn't shared: both migrant and non-migrant Asian Indians are up to 4 times more likely to be diagnosed with type 2 diabetes young (< 40 years) and lean compared to white Europeans 4 , and up to 3 times more likely to develop gestational diabetes mellitus (GDM) 5 . Despite this increased risk, South Asian individuals have historically been poorly represented in genetic studies, including those relating to diabetes. 6 , 7 . This is of particular relevance as in recent years there has been increased understanding of type 2 diabetes as a heterogeneous condition comprising multiple subgroups or clusters (typically referred to as endotypes), determined on the basis of phenotypic and genetic characteristics, such as endotypes related to obesity or lipodystrophy 8 . Phenotypic endotyping approaches have shown a preponderance of Severe Insulin Deficient Diabetes (SIDD) in south Asians observed in two independent cohorts in India: 35% and 52% respectively 9,10 . This contrasts with white Europeans, among whom insulin resistance plays a more prominent role 10 . Research defining genetically determined endotypes has been largely restricted to White European participants, with poor representation of south Asian populations in more recent multi-ancestry efforts 11 – 14 . The latest multi-ancestry meta-GWA study has specifically highlighted the utility of partitioned polygenic scores relative to polygenic scores in identifying individuals at risk of diabetes complications. However, only 2% of the 2.5 million individuals in the study were of south Asian ancestry, limiting the amount of representation of these individuals in pooled multi-ancestral analyses. 15 Unlike polygenic scores that predict overall risk of an outcome (like type 2 diabetes) by summing the genetic burden of risk from genetic variants identified in genome-wide association studies, partitioned polygenic scores (pPS) offer insight into the mechanistic and aetiological factors driving this risk by leveraging clustering approaches to estimate genetic propensity to underlying mechanistic pathways like beta cell dysfunction, obesity, lipid metabolism or lipodystrophy 16 . They have been shown to be associated with risk of diabetes complications, e.g a pPS for Beta-cell function, predictive of insulin deficiency, was associated impaired renal function 17 . Populations like South Asians, who are not well represented in studies elucidating the complex architecture of type 2 diabetes, may not benefit from advances in understanding the condition. Genetically-determined diabetes endotypes relating to distinct aetiological mechanisms have been well-described and widely validated in white European populations 11 , 18 , 19 However, as many aspects of genetic architecture and clinical phenotypes of type 2 diabetes differ between South Asians and White Europeans 209 , it remains unclear to what extent these pPS might transfer to understanding type 2 diabetes risk and progression in individuals of South Asian ancestry. Furthermore, the phenotypic traits and clinically relevant characteristics associated with an individual having high genetic risk across multiple distinct aetiological pathways, which we term “pPS extremes”, has not been explored, but could offer opportunities to identify individuals at high risk of clinically relevant outcomes. We used Genes & Health, a long-term community-based study with genetics and linked electronic health and prescribing data for over 51,000 British Pakistani and Bangladeshi individuals living in the UK 21 to investigate whether pPS can help unravel the aetiological factors driving young onset type 2 diabetes and clinically relevant related outcomes. Methods Cohort profile Genes & Health is a long-term, community-based study of British Pakistani and British Bangladesh individuals aged 16 years and above living in the United Kingdom 21 . At recruitment, participants provide a saliva sample for genotyping, complete a short questionnaire on basic demographic information, and consent to linkage at primary and secondary care electronic health records. Since recruitment began in 2015 over 50,000 participants have been recruited, with linked genetic and electronic health record information available for 44,189 as of July 2023 (number of type 2 diabetes cases = 9771), and 51,170 as of February 2024. We used the additional data release of 7383 newly genotyped individuals, of whom 1907 had T2D, to replicate key findings. Inclusion and exclusion criteria We used no specific inclusion criteria. We excluded individuals with clinical codes consistent with type 1 diabetes, MODY diabetes, or causes of secondary diabetes such as cystic fibrosis and pancreatectomy. Genetic data processing and curation Genotyping was performed on Illumina Infinium Global Screening Array v3 with additional multi-disease variants. Quality control was performed in line with previous descriptions 22 , 23 . In brief, variants with call rates < 0.99 and /or MAF < 1% were excluded. We excluded individuals unlikely to have genetically-inferred Pakistani or Bangladeshi ancestry. Imputation was performed using TopMED R2 panel. We excluded SNPs with low imputation scores for INFO (< 0.3) or MAF < 0.1%. Further details on genetic data processing, curation, and ancestry inference are presented in the supplementary information. Electronic Health Record data processing and curation We curated routine UK electronic health record (EHR) data from primary care (SNOMED coded) and secondary care (ICD10 coded) sources. Data were combined without mapping between coding formats. For each clinical code, we took the earliest ever measure recorded in a participant’s medical records, excluding erroneous code dates preceding the participants’s recorded date of birth. Exposures: Partitioned polygenic risk score (pPS) construction and ancestry correction We calculated pPS for twelve diabetes-associated genetically determined endotypes described by Smith et al derived from high-throughput genetic clustering techniques in White European, East Asian and African ancestry individuals, using only SNPs above the authors’ specified inclusion threshold (cluster weight > 0.78) 18 . These comprise three endotypes related to glucose sensing, insulin secretion and insulin production (Beta cell 1, Beta cell 2, Proinsulin respectively), three clusters related to insulin resistance and unfavourable adiposity (Obesity, Lipodystrophy 1, Lipodystrophy 2) and 6 clusters with unclear effects on insulin resistance and deficiency (Liver/lipid, ALP [alkaline phosphatase] negative, Hyper insulin secretion, Cholesterol, Sex Hormone Binding Globulin / Lipoprotein A (SHBG/LpA), and bilirubin). We additionally calculated scores for a global T2D polygenic risk score (T2D PRS), using a previously-published score 24 with the highest beta and area under the receiver-operator curve for predicting incident T2D in the Genes & Health cohort. Principal components analysis of genetic data shows distinct population structure for people of Bangladeshi and Pakistani ancestries 25 , and we observed differences in pPS distribution between these groups. Therefore, to maximise power and facilitate combined analyses of all individuals we regressed out the effect of the first 5 genetic principal components from each pPS, using an approach described in 25 (further details presented in Supplementary methods and Fig S1 ). pPS extremes We defined pPS “extremes” as scores in the top or bottom 10% of each residualized pPS distribution, and “combined extremes” as individuals with scores in the top or bottom 10% of multiple pPS distributions. UKBiobank - cross-ancestry differences in genetic burden We used UK Biobank to compare the distribution of pPS between individuals with T2D of European (EUR) and South Asian (SAS) ancestry. Differences in distribution of pPS across ancestry groups were assessed using ANOVA for normally distributed and Kruskall-Wallis testing for all other pPS. All analyses in UK Biobank were conducted under application ID 44448. Medications Diabetes-controlling medication classes were defined according to method of action: insulin secretagogues (sulfonylureas and meglitinide), incretin mimetics (GLP1 receptor agonists and DDP4 inhibitors), insulin sensitizers (pioglitazone and thiazolidinediones), renal tubular glucose reabsorption modifiers (SGLT2 inhibitors), in addition to metformin and insulin. Initiation of medication was defined as the first instance of medication prescription in the EHR. For treatment response analyses (described in further detail in the “outcomes” section below), concurrently prescribed medications were defined as medications from another class prescribed within a conservative window of 6 months before or after initiation of the medication being treated as the exposure. Outcomes: Diabetes phenotypes These categorical outcome variables were defined using ICD-10 and SNOMED codelists adapted from the AI-MULTIPLY resource 26 ( Table S1 ). Diabetes phenotypes included type 2 diabetes, gestational diabetes mellitus, and incident type 2 diabetes after gestational diabetes mellitus. Diabetes-related complications were defined as microvascular (nephropathy, neuropathy, and retinopathy) and macrovascular (coronary artery disease, cerebrovascular disease, and peripheral vascular disease).( Table S1 ). Age at Diagnosis Date of diagnosis for each outcome was defined as the earliest recorded date in either primary or secondary care above the age of 18. Quantitative traits Quantitative outcomes were, unless otherwise stated, defined as the measure taken closest to the date of type 2 diabetes within 1 year (before and/or after) and included age, body mass index (BMI), waist circumference, glycated haemoglobin (HbA1c), Fasting and random blood glucose, low- and high-density lipoprotein cholesterol (LDL-C and HDL-C), serum triglycerides, alkaline phosphatase (ALP), and alanine transaminase (ALT). Because diabetes-related traits may rapidly change after diagnosis and/or initiation of treatment, for each trait, the value closest to the time of diagnosis was used. In addition to traits at diagnosis, we explored the number of medication classes (as defined above) an individual was prescribed in 5 years, and the change in HbA1c from time of diagnosis to 5 years (the HbA1c closest in time to 5 years from diagnosis date was taken, and only values between 4–6 years post-diagnosis included in analysis). Response to glucose-lowering treatment For treatment response analyses, medication data was extracted from the primary care EHR. In line with pharmacogenomic studies 27 , treatment response was defined as the percentage change between the most recent HbA1c in 6 months before medication initiation, and lowest HbA1c in the 1 year after initiation, as a proportion of pre-medication HbA1c (i.e., %change from prior to initiation). Statistical analyses Descriptive Analysis We calculated mean values for quantitative traits at diagnosis and five years post-diagnosis, stratified by ancestral group, and compared these using ANOVA. Multivariable Analysis We described the association of each pPS (the exposure) with each diabetes phenotype outcome, using multivariable logistic regression models adjusted for age, sex, and ancestry, to estimate the per-standard-deviation increase in odds of diabetes phenotype between diabetes phenotype cases and non-diabetic controls. We estimated the association of each pPS (the exposure) with diabetes-related traits at the time of diagnosis (the outcome). To allow comparison of effects of pPS between quantitative traits at the time of diagnosis, each quantitative trait was scaled to a normal distribution, and the beta per standard deviation of pPS presented for each trait, estimated from multivariable logistic regression models adjusted for age, sex and ancestry. Multivariable linear regression was used to estimate the effect of pPS on age of T2D diagnosis, adjusted for ancestry and sex; partial R 2 was calculated with the R package “partialR2”. We hypothesised a-priori that associations may differ between sexes and ancestry groups; because of this, we sex-stratified and ancestry-stratified analyses were also performed. For pharmacogenomic analysis of treatment response, association between each pPS (the exposure) and HbA1c change in response to medication (the outcome) was estimated from multivariable logistic regression models adjusted for age, sex and ancestry, as well as concurrently prescribed antidiabetic medication from all other classes within 6 months before or after initiation of each. We meta-analysed treatment response analyses from the discovery and replication samples using fixed effects models. Survival analysis We constructed survival models starting from each individual’s date of diagnosis, running until last data extraction, for two outcome categories: initiation of insulin, and progression to diabetes-related complications. We explored the association of each pPS with each complication outcome using Cox proportional hazards models adjusted for age, sex and ancestry. We calculated Schoenfeld residuals for each model to check assumptions of proportionality. Software and statistical computing Genotype curation and partitioned polygenic risk score calculation was performed using Plink v2.0 28 . Statistical analyses were performed using R V4.2.3. Role of the funding source The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. Reporting We report this study following the STREGA guidelines 29 . Results Characteristics of populations studied and differences in clinical outcomes between British Bangladeshis and Pakistanis. We analysed data from 9,771 British Pakistani and Bangladeshi individuals with a diagnosis of type 2 diabetes and 35,190 diabetes-free controls; their demographic and clinical information stratified by ancestry and sex, is shown in Table 1. We observed differences in clinical features and prevalence of diabetes-related outcomes by ancestral group: compared to Pakistani individuals, Bangladeshi individuals tended to be diagnosed with type 2 diabetes earlier in life, at lower BMI, and had higher rates of prior GDM. Consequently, Bangladeshis had a higher prevalence of type 2 diabetes. However, at diagnosis and in the 5 years following diagnosis, Pakistanis had higher HbA1c and BMI. Overall, Pakistanis were more likely to be prescribed insulin and develop nephropathy than Bangladeshis, and had higher HbA1c 5 years after diagnosis, despite being prescribed antidiabetic medications from more classes. Association of partitioned polygenic scores and quantitative traits at time of diagnosis We observed numerous expected associations between single pPS and traits at time of diagnosis ( Fig S2 , Table S2 ) , including associations between beta cell related pPSs (Beta Cell 1, Beta Cell 2, Proinsulin), and lower BMI, and greater HbA1c, fasting and random blood glucose at diagnosis. Lipodystrophy 1, was associated with lower BMI and waist circumference and raised ALT at time of diagnosis. Both lipodystrophy-related pPSs were associated with lower HDL and raised triglycerides, whereas the Cholesterol pPS was associated with raised HDL and lower triglycerides. The obesity pPS was associated with greater BMI and waist circumferences. Several pPS demonstrated reasonably strong correlations with one another ( Fig S3 , Table S3 ) , with strongest associations observed between pPS with similar proposed mechanisms of action, eg between Lipodystrophy 1 and Lipodystrophy 2 (Pearson R 2 = 0.50), and for Beta Cell 1 and Beta Cell 2 (R 2 = 0.46). Weaker correlations were observed between pPS acting via distinct pathophysiological pathways, e.g. Beta Cell 1 and Lipodystrophy 1 (R 2 = 0.35), and Beta Cell 1 and Obesity (R 2 = 0.31). These results suggest they are unlikely to be exerting independent effects. Ancestral differences in pPS Distribution Noting the variation in clinical features of diabetes between British Bangladeshi and Pakistani individuals in our cohort, we first characterised distributions of unmodified pPS (not corrected for PC-defined ancestry). We observed higher unmodified pPS scores in Bangladeshi individuals, notably for Beta Cell 2, Obesity, and Lipodystrophy 1 (Fig S1 ) (ANOVA p between Bangladeshis and Pakistanis < 0.001 for all three pPS). To maximise power in subsequent analyses, we pooled the two sub-ancestries using PC-corrected pPS. We compared distributions of pPS between individuals with type 2 diabetes of White European and South Asian ancestry in UK-Biobank, observing higher scores among South Asians for several pPS including Beta Cell 2 (p = 3e-8) and Lipodystrophy 1(p = 1.95e-157), while Obesity pPS was higher among Europeans(p = 1.09e-47) (Fig S4 , Table S4 ). Associations between Partitioned polygenic scores are associated with type 2 diabetes, GDM, and transition from GDM to T2D For all pPS other than bilirubin, and a type 2 diabetes PRS, scores were higher among T2D cases, GDM cases, and individuals developing T2D after GDM, compared to non-diabetic controls (Fig S5 ) . We observed associations between pPS and T2D risk,, after adjustment for sex and ancestry in 44,189 individuals (number of T2D cases = 9771) (Fig. 1 ). The strongest associations between pPS and T2D were for beta cell function- mediated endotypes Beta Cell 1 - related to glucose sensing (beta per standard deviation (SD) 0.39, 95% CI 0.35–0.42, p < 0.001) and Beta Cell 2 - related to insulin deficiency (beta per SD 0.32, 95% CI 0.28–0.35, p < 0.001). All pPS, other than Bilirubin were associated with both GDM and incident T2D after GDM were observed in 5430 unrelated individuals with a history of at least one pregnancy ( Fig. 1 ) . As with T2D, the strongest association between pPS and GDM was with pPSs related to Beta Cell 1 (beta per SD 0.24, 95% CI 0.18–0.29) and Beta Cell 2 (beta per SD 0.23, 95% CI 0.17–0.28). Similar patterns were observed in association of pPS with incident T2D after GDM, with strongest associations for Beta Cell 1 and 2. Genetic predisposition to insulin deficiency drives earlier age of onset in type 2 diabetes Although all pPS other than bilirubin were associated with earlier age of T2D onset (Fig S6 ) , many of these effects were non-significant in a multivariate regression model incorporating all twelve pPS ( Fig. 2 A ). The pPS most strongly associated with age of onset were Beta Cell 2 (diagnosis 1.1 years earlier in life per SD, 95% CI 0.9–1.4 years, p = 3e-16), Obesity (0.57 years per SD, 95% CI 0.30–0.83 years, p = 3e-5) and Lipodystrophy 1 (0.54 years per SD, 95% 0.25–0.82 years, p = 2e-4)(Fig. 2 A ). In this model, the pPS with the highest partial R 2 was Beta Cell 2 (0.007), followed by Obesity (R 2 = 0.002) and Lipodystrophy 1 (0.001) ( Fig. 2 B ) . In ancestry- and sex-stratified analyses we found strongest effects for both beta cell 2 and lipodystrophy 1 in Bangladeshi females (Beta per SD = 1.93, P = 7.2e-20 and 1.70, P = 9.3e-15, respectively, ( Fig S7 ). We found no significant interaction between pPS and sex in ancestry-specific analyses in individuals of either Pakistani or Bangladeshi ancestry ( Fig S7 ) . Subsequent analyses were focused on these three endotypes, as key drivers of early onset T2D risk. Association between partitioned polygenic risk scores and treatment response We observed associations between pPS and response to oral antidiabetic medications, measured as the % change in HbA1c (in mmol/mol) following medication initiation ( Fig. 3 D, Fig S8 , Table S4 ) . We replicated our initial findings in a replication sample of 7383 individuals (number of T2D cases = 1907) in Genes & Health, then meta-analysed our discovery and replication results. Higher scores across all pPS were generally associated with negative, or no, response to oral antidiabetic medication initiation ( Fig S8 A) , while across the whole sample of included participants, HbA1c typically declined after initiation of each medication class (median % change in HbA1c ranged from − 14.6% (metformin) to -20.3% (sulfolnylurea)( Fig S8 B, Table S5 ). Higher Beta Cell 2 pPS score was associated with increased HbA1c after initiation of thiazolidinediones (meta-analysed HbA1c increase 1.68% per SD, 95% CI 0.37–2.99, p = 0.01), SGLT2i (1.22% per SD, 95% CI 0.56–1.89, p = 3e-4), and metformin (0.51%, 95% CI 0.01–1.01, p = 0.047). Higher Liver-Lipid pPS was associated with decrease in HbA1c after initiation of metformin (-0.65% per SD, 95% CI -0.16 - -1.13, p = 0.008), while the T2D PRS was not associated with treatment response to metformin, thiazolidinediones, SGLT2 inhibitors, or sulfonylureas (Fig S8 A) . The Beta Cell 2 and Lipodystrophy 1 pPSs were predictive of progression to insulin treatment in adjusted survival models (HR per SD 1.08, 95% CI 1.03–1.12, p = 0.0008; HR per SD 1.11 95% CI 1.06–1.15, p = 0.0002), while the Obesity pPS was not associated with progression to insulin treatment.(Fig. 3 B ) . Results for other pPS are shown in Fig S9 . Genetic Risk Extremes Predict Lean and Early Onset Diabetes, and Progression To Complications Finally, we explored whether extremes of genetic risk, defined as being in the top versus bottom decile of a pPS distribution (high or low risk, respectively) were associated with T2D phenotype (Table 2) . Across all participants, T2D prevalence was 22.1% with mean age of onset of 46.6 years. Among individuals in the top decile of the Lipodystrophy 1 distribution, prevalence of T2D was 26.3% (mean age of onset 46.6 years) ( Fig. 4 a ) . Among individuals in the top 10% of the Beta Cell 2 distribution T2D prevalence was 29.6% and mean age of onset 43.6 years. Compared to individuals in the bottom 10% of the Beta Cell 2 pPS distribution, these individuals were more likely to develop nephropathy (HR 1.58, 95% CI 1.19–2.06, p = 0.001) ( Fig. 4 C, Fig S10 , Table S6 -7) . The obesity pPS, in contrast, was not associated with progression to complications. Individuals at combined high genetic risk for Lipodystrophy 1 and Beta Cell 2 (in the top decile of both distributions, n = 110) were on average diagnosed with diabetes 8 years earlier in life than those in the bottom decile of both distributions (n = 304), had 3kg/m 2 lower BMI at diagnosis (Fig. 4 A /B) , 7.8% higher lifetime prevalence of diabetic retinopathy, 12% higher prevalence of insulin dependence, and 3.73 mmol/mol higher 5 year HbA1c in spite of similar baseline HbA1c (Table 2 ). In survival models, we found they were more likely to progress to diabetic neuropathy (HR 2.75, 95% CI 1.10–6.9, p = 0.031)(Fig. 4 C, Fig S10 ) . Discussion Summary of findings In this study of 51,170 British Pakistani and Bangladeshi individuals, we show that multi-ancestry T2D partitioned polygenic scores (pPS) developed in White European, East Asian and African ancestry individuals are applicable to British South Asian individuals, where they are predictive of both type 2 diabetes and gestational diabetes mellitus. We show the genetic architecture of T2D and GDM, mapped through pPS, bear striking similarities. Further, we identify beta-cell 2 and lipodystrophy 1 as being key drivers for type 2 diabetes diagnosis at a younger age with a lower BMI, and more rapid progression to microvascular complications. Based on associations used for endotype development, the beta cell 2 pPS was driven by lower HOMA- B and high proinsulin levels suggesting an insulin secretion defect. Lipodystrophy 1 was associated with lower gluteofemoral adipose tissue and adiponectin levels which indicate greater risk of unfavourable fat distribution-mediated insulin resistance. Using UK Biobank we show how the genetic burden of these pPS are greater in people of south Asian ancestry compared to white Europeans. We also show how higher genetic risk of the beta cell 2 pPS is associated with worse response to commonly-prescribed and first-line oral antidiabetic medications including metformin, SGLT2 inhibitors, and thiazolidinediones. A global T2D polygenic score does not show similar associations. Partitioned polygenic scores can help identify individuals with combined extreme genetic risk in specific pathways (lipodystrophy 1 and beta-cell 2) who have 8 year earlier onset of T2D, at 3 kg/m 2 lower BMI, and rapid progression to insulin use and nephropathy, with poor response to oral antidiabetic medications. We show that pPS can be used to identify individuals at high, clinically actionable risk and highlight the specific pathophysiological pathways underlying T2D risk in south Asians. These results recapitulate findings from the largest recent genetic study of T2D and are supportive of the use of genetic information, specific partitioned polygenic scores, in guiding screening and therapeutic management for type 2 diabetes, especially in south Asians. Type 2 diabetes pPS and PRS distributions differ between British Pakistani and Bangladeshi individuals Using UK Biobank, we observed higher genetic predisposition to beta cell dysfunction and unfavourable fat distribution patterns, but lower genetic predisposition to obesity, among individuals of South Asian compared to European ancestry (Fig S5 ). In Genes & Health, using stratified analyses, we identified higher unmodified (not corrected for principal components) genetic risk of all pPS in Bangladeshi compared to Pakistani individuals ( Fig S1 ) . While we were under-powered to observe interactions between ancestry and sex we observed that Bangladeshi women showed the strongest associations between genetic risk and earlier onset T2D. These findings argue to an extent against the pooling of these distinct ancestral groups under the banner of “South Asian / SAS” in genetic epidemiological studies when eventually sample size is no longer a limitation 30 . These findings also highlight important sex-stratified effects that can explain higher observed risk of T2D and GDM in women of certain ancestries. Genetic predictors of beta cell dysfunction, obesity, and lipodystrophy drive early-onset diabetes in Genes & Health We identify genetic propensity to impaired insulin secretion as a key driver of the genetic basis of age at diagnosis of type 2 diabetes in British Pakistanis and Bangladeshis (Fig. 2 , Fig. 4 A ) . Epidemiological studies have shown that south Asians are more likely to present with lower HOMA-B (an estimate of beta cell function), BMI, and greater dyslipidemia at the time of diagnosis referred to as the Severe Insulin Deficient Diabetes (SIDD) endotype, as compared to white Europeans 31 . The greater genetic burden of insulin secretion and early beta-cell exhaustion is a likely reason. The observed effect of greater genetic burden of impaired insulin secretion recapitulates findings from Asian Indians, where a previously defined beta-cell dysfunction pPS was found to be associated with earlier onset T2D 4 . We replicate this using a newer multi-ancestry pPS and extend the finding to highlight the role of specific aetiologies in GDM, treatment response, progression to insulin dependence and nephropathy. Numerous studies have reported the clinical phenotype of low BMI but larger waist circumferences in south Asian populations 32 . For example, SABRE, a 20-year-long prospective cohort study comparing white Europeans, south Asians, and Afro-Caribbean groups living in the UK found that, compared to south Asian men developed T2D with lower BMI, higher waist-hip ratio, higher truncal skin-folds, higher insulin resistance, and increased (compensatory) insulin secretion 33 . Therefore, our observation that the aetiological pathways of unfavourable fat distribution plays a key role in early onset and rapid progression of T2D in South Asians is well-supported. There are limited studies that explore the underlying architecture of age of diagnosis and by proxy, earlier onset of T2D 20 , 34 . However, efforts to do so, agree with our findings that drivers of earlier age of diabetes onset do not overlap entirely with overall drivers of T2d risk. Our use of pPS aid the identification of specific aetiologies driving earlier onset in South Asians. A limited role for genetically-determined BMI-mediated obesity We observed an effect of BMI-mediated obesity pPS in driving early-onset type 2 diabetes, however, the pPS was not associated with greater risk of insulin dependence or diabetes-related complications. This effect is mirrored in epidemiological studies that have shown south Asians who develop T2D at younger ages have increased weight over those who do not, but that this weight gain is relatively lesser than observed in white Europeans or Black individuals 3 . This was also observed in a national survey of Asian Indians, where 45% of young diagnosed (< 40 years) T2D was in individuals with obese (45%) or overweight (15%) BMI. This is the first study using genetics that uncovers the role of these two adiposity-related aetiologies (unfavourable fat distribution) and BMI in south Asian type 2 diabetes. Genetic architecture of GDM mirrors that of Type 2 Diabetes in British Pakistanis and Bangladeshis In comparison to a rich literature for type 2 diabetes, the genetics of gestational diabetes mellitus is less well understood. In this study, we show that the genetic architecture of GDM, assessed by association with genetically determined T2D endotypes, closely resembles that of the genetic basis of Type 2 Diabetes itself in British Bangladeshis and Pakistanis ( Fig. 1 A, Fig S5 ) , in keeping with similar findings recently reported in European ancestry individuals 35 . In our study, the strongest associations with GDM were observed for Beta Cell 1 and Beta Cell 2, suggesting genetic predisposition to insulin deficiency in British South Asian women may contribute to GDM risk as well as T2D risk. The association of GDM with the Obesity pPS in our study, in contrast, was relatively weaker, despite this being highlighted as a major aetiological pathway in White European 35 and Turkish 36 mothers with GDM. Genetic testing allows prediction of clinically relevant outcomes We show application of pPS to real-world linked electronic health records allows identification of individuals at extremes of genetic risk who are at particularly high risk of developing type 2 diabetes in early adulthood (Fig. 4 A ) , responding poorly to widely-used oral antidiabetic drugs (Fig. 3 A ) , and of progressing rapidly to insulin requirement ( Fig. 3 B ) and complications ( Fig. 4 C ). However, no pPS or combination of pPSs was as predictive as previously reported hazard ratios for phenotypically-derived clusters related to insulin deficiency in the IMI-RHAPSODY study. 37 This may reflect the fact that the 12 endotypes included in this analysis were derived in a multi-ancestral study which did not train cluster allocation on insulin-secretion related traits 19 . Unlike phenotypic clustering efforts, genetic data can predict risk much earlier in the life course, rather than proximate to the time of disease onset. As the use of genetic testing in the real-world clinical setting increases 38 , 39 , genetic risk scores may offer means to stratify individuals into those at high risk of requiring specialist input (for example, those rapidly progressing to insulin requirement, who are more likely to require specialist secondary care management) 40 , or who are less likely to respond to particular classes of medication. Overall, most individuals responded to the introduction of glucose lowering medication with a reduction in HbA1c ( Fig S8 A). However, high genetic risk of insulin deficiency determined by high beta Cell 2 pPS was associated with increased HbA1c after initiation of metformin, SGLT2 inhibitors, and thiazolidinediones (Fig. 4 A ) .These findings are biologically plausible considering, for example, the mode of action of thiazolidinediones as insulin sensitizers 41 , which may offer little benefit if insulin deficiency is underpinning an individual’s hyperglycaemia. While mechanistically congruent, these treatment-response findings are novel and possible only due to linkage with primary, secondary care and prescribing records in G&H, and have not been shown in other pPS efforts. In contrast, we did not observe differential treatment response using a type 2 diabetes polygenic score, highlighting the value of pPS in dissecting clinically relevant pathophysiologies. Our study underscores the utility of pPS and extends their utility to complications such as insulin dependence and treatment response and highlights that as pPS become more robust and ancestrally diverse, their utility in guiding diabetes prevention and management strategies should increase 15 , 19 . pPS allow us to identify greater genetic burden of certain pathophysiological pathways of T2D in south Asians and the use of this under-represented ancestry allows us to uncover novel associations with beta-cell deficiency which would not be apparent in European cohorts. Strengths and limitations Strengths of this study include its exploration of an under-represented population with high burden of cardiometabolic disease, and linkage of real-world electronic health record and prescribing data, which provides a platform for real-world application and translation of clinically relevant findings, in addition to internal validation of novel findings around pharmacogenetic applications of pPS. We also demonstrate the utility of pPS derived by Smith et al 42 in a population not included in their pPS derivation. Weaknesses include the lack of external validation of results, which is limited in part by the paucity of non-European studies combining genetic data with health record data 7 ; in fact, some of the results shown such as response to medication have not even been shown in European cohorts due to rarity of required clinical and prescribing data. Due to attrition of sample size for complications these analyses are quite under-powered, however where this is the case, we have used an internal replication cohort to validate results.. Another limitation is the fact that this is not an ancestry-specific pPS, which suggests we might not be using the most optimal causal variants. However, this is likely to cause an under-estimation of true genetic effects 23 , 43 . Finally, in common with all studies using real-world electronic health record data, there is a risk of misclassification of diabetes, miscoding, and sampling bias towards individuals with chronic disease. Declarations Acknowledgment/Funding Statement : SH is funded by a Wellcome HARP Doctoral Fellowship 227532/Z/23/Z. RM and MKS are funded by Barts Charity (MGU0504). DS is funded by the Tackling Multimorbidity at Scale Strategic Priorities Fund programme [grant number MR/W014416/1] delivered by the Medical Research Council and the National Institute for Health Research in partnership with the Economic and Social Research Council and in collaboration with the Engineering and Physical Sciences Research Council. Genes & Health is/has recently been core-funded by Wellcome (WT102627, WT210561), the Medical Research Council (UK) (M009017, MR/X009777/1, MR/X009920/1), Higher Education Funding Council for England Catalyst, Barts Charity (845/1796), Health Data Research UK (for London substantive site), and research delivery support from the NHS National Institute for Health Research Clinical Research Network (North Thames). Genes & Health is/has recently been funded by Alnylam Pharmaceuticals, Genomics PLC; and a Life Sciences Industry Consortium of Astra Zeneca PLC, Bristol-Myers Squibb Company, GlaxoSmithKline Research and Development Limited, Maze Therapeutics Inc, Merck Sharp & Dohme LLC, Novo Nordisk A/S, Pfizer Inc, Takeda Development Centre Americas Inc. We thank Social Action for Health, Centre of The Cell, members of our Community Advisory Group, and staff who have recruited and collected data from volunteers. We thank the NIHR National Biosample Centre (UK Biocentre), the Social Genetic & Developmental Psychiatry Centre (King's College London), Wellcome Sanger Institute, and Broad Institute for sample processing, genotyping, sequencing and variant annotation. We thank: Barts Health NHS Trust, NHS Clinical Commissioning Groups (City and Hackney, Waltham Forest, Tower Hamlets, Newham, Redbridge, Havering, Barking and Dagenham), East London NHS Foundation Trust, Bradford Teaching Hospitals NHS Foundation Trust, Public Health England (especially David Wyllie), Discovery Data Service/Endeavour Health Charitable Trust (especially David Stables), Voror Health Technologies Ltd (especially Sophie Don), NHS England (for what was NHS Digital) - for GDPR-compliant data sharing backed by individual written informed consent. Most of all we thank all of the volunteers participating in Genes & Health. Current Genes & Health Research Team (in alphabetical order by surname): Shaheen Akhtar, Mohammad Anwar, Omar Asgar, Samina Ashraf, Saeed Bidi, Gerome Breen, James Broster, Raymond Chung, David Collier, Charles J Curtis, Shabana Chaudhary, Grainne Colligan, Panos Deloukas, Ceri Durham, Faiza Durrani, Fabiola Eto, Sarah Finer, Joseph Gafton, Ana Angel, Chris Griffiths, Joanne Harvey, Teng Heng, Sam Hodgson, Qin Qin Huang, Matt Hurles, Karen A Hunt, Shapna Hussain, Kamrul Islam, Vivek Iyer, Benjamin M Jacobs, Georgios Kalantzis, Ahsan Khan, Claudia Langenberg, Cath Lavery, Sang Hyuck Lee, Daniel MacArthur, Sidra Malik, Daniel Malawsky, Hilary Martin, Dan Mason, Rohini Mathur, Mohammed Bodrul Mazid, John McDermott, Caroline Morton, Bill Newman, Elizabeth Owor, Asma Qureshi, Shwetha Ramachandrappa, Mehru Raza, Jessry Russell, Nishat Safa, Miriam Samuel, Moneeza Siddiqui, Michael Simpson, John Solly, Marie Spreckley. Daniel Stow, Michael Taylor, Richard C Trembath, Karen Tricker, David A van Heel, Klaudia Walter, Caroline Winckley, Suzanne Wood, John Wright, Ishevanhu Zengeya, Julia Zöllner. References Magliano, D. & Boyko, E. J. IDF Diabetes Atlas . (International Diabetes Federation, 2021). Gujral, U. P., Pradeepa, R., Weber, M. B., Narayan, K. M. V. & Mohan, V. Type 2 diabetes in South Asians: similarities and differences with white Caucasian and other populations. Ann. N. Y. Acad. Sci. 1281 , 51–63 (2013). Wright, A. K. et al. Age-, sex- and ethnicity-related differences in body weight, blood pressure, HbA1c and lipid levels at the diagnosis of type 2 diabetes relative to people without diabetes. Diabetologia 63 , 1542–1553 (2020). Siddiqui, M. K. et al. Young-onset diabetes in Asian Indians is associated with lower measured and genetically determined beta cell function. Diabetologia 65 , 973–983 (2022). Farrar, D. et al. Association between hyperglycaemia and adverse perinatal outcomes in south Asian and white British women: analysis of data from the Born in Bradford cohort. Lancet Diabetes Endocrinol 3 , 795–804 (2015). Sirugo, G., Williams, S. M. & Tishkoff, S. A. The Missing Diversity in Human Genetic Studies. Cell 177 , 26–31 (2019). Fatumo, S. et al. A roadmap to increase diversity in genomic studies. Nat. Med. 28 , 243–250 (2022). Deutsch, A. J., Ahlqvist, E. & Udler, M. S. Phenotypic and genetic classification of diabetes. Diabetologia 65 , 1758–1769 (2022). Anjana, R. M. et al. Novel subgroups of type 2 diabetes and their association with microvascular outcomes in an Asian Indian population: a data-driven cluster analysis: the INSPIRED study. BMJ Open Diabetes Res Care 8 , (2020). Prasad, R. B. et al. Correction to: Subgroups of patients with young-onset type 2 diabetes in India reveal insulin deficiency as a major driver. Diabetologia 65 , 254 (2022). Mansour Aly, D. et al. Genome-wide association analyses highlight etiological differences underlying newly defined subtypes of diabetes. Nat. Genet. 53 , 1534–1542 (2021). Dennis, J. M., Shields, B. M., Henley, W. E., Jones, A. G. & Hattersley, A. T. Disease progression and treatment response in data-driven subgroups of type 2 diabetes compared with models based on simple clinical features: an analysis using clinical trial data. Lancet Diabetes Endocrinol 7 , 442–451 (2019). Ahlqvist, E. et al. Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables. Lancet Diabetes Endocrinol 6 , 361–369 (2018). Udler, M. S. et al. Type 2 diabetes genetic loci informed by multi-trait associations point to disease mechanisms and subtypes: A soft clustering analysis. PLoS Med. 15 , e1002654 (2018). Suzuki, K. et al. Genetic drivers of heterogeneity in type 2 diabetes pathophysiology. Nature (2024) doi:10.1038/s41586-024-07019-6. Udler, M. S., McCarthy, M. I., Florez, J. C. & Mahajan, A. Genetic Risk Scores for Diabetes Diagnosis and Precision Medicine. Endocr. Rev. 40 , 1500–1520 (2019). DiCorpo, D. et al. Type 2 Diabetes Partitioned Polygenic Scores Associate With Disease Outcomes in 454,193 Individuals Across 13 Cohorts. Diabetes Care 45 , 674–683 (2022). Kim, H. et al. High-throughput genetic clustering of type 2 diabetes loci reveals heterogeneous mechanistic pathways of metabolic disease. Diabetologia 66 , 495–507 (2023). Smith, K. et al. Multi-ancestry polygenic mechanisms of type 2 diabetes. Nat. Med. (2024) doi:10.1038/s41591-024-02865-3. Srinivasan, S. et al. Common and Distinct Genetic Architecture of Age at Diagnosis of Diabetes in South Indian and European Populations. Diabetes Care 46 , 1515–1523 (2023). Finer, S. et al. Cohort Profile: East London Genes & Health (ELGH), a community-based population genomics and health study in British Bangladeshi and British Pakistani people. Int. J. Epidemiol. 49 , 20–21i (2020). Huang, Q. Q. et al. Transferability of genetic loci and polygenic scores for cardiometabolic traits in British Pakistanis and Bangladeshis. bioRxiv (2021) doi:10.1101/2021.06.22.21259323. Hodgson, S. et al. Integrating polygenic risk scores in the prediction of type 2 diabetes risk and subtypes in British Pakistanis and Bangladeshis: A population-based cohort study. PLoS Med. 19 , e1003981 (2022). Mars, N. et al. Systematic comparison of family history and polygenic risk across 24 common diseases. Am. J. Hum. Genet. 109 , 2152–2162 (2022). Liu, T. et al. Investigating misclassification of type 1 diabetes in a population-based cohort of British Pakistanis and Bangladeshis using polygenic risk scores. bioRxiv (2023) doi:10.1101/2023.08.23.23294497. Creators Fabiola Eto1 Miriam Samuel finersarah Show affiliations 1. Queen Mary University of London. F-eto/MULTIPLY-Initiative: Version 1.1 . doi:10.5281/zenodo.7643566. Zhou, K. et al. Variation in the glucose transporter gene SLC2A2 is associated with glycemic response to metformin. Nat. Genet. 48 , 1055–1059 (2016). Purcell, S. et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am. J. Hum. Genet. 81 , 559–575 (2007). Little, J. et al. STrengthening the REporting of Genetic Association Studies (STREGA)--an extension of the STROBE statement. Genet. Epidemiol. 33 , 581–598 (2009). Chambers, J. C. et al. The South Asian genome. PLoS One 9 , e102645 (2014). Ke, C., Narayan, K. M. V., Chan, J. C. N., Jha, P. & Shah, B. R. Pathophysiology, phenotypes and management of type 2 diabetes mellitus in Indian and Chinese populations. Nat. Rev. Endocrinol. 18 , 413–432 (2022). Tillin, T. et al. Insulin resistance and truncal obesity as important determinants of the greater incidence of diabetes in Indian Asians and African Caribbeans compared with Europeans: The Southall and Brent Revisited (SABRE) cohort. Diabetes Care 36 , 383–393 (2013). Tillin, T., Hughes, A. D., Godsland, I. F. & Whincup, P. Insulin resistance and truncal obesity as important determinants of the greater incidence of diabetes in Indian Asians and African Caribbeans compared with …. Diabetes (2013). Kwak, S. H. et al. Genetic architecture and biology of youth-onset type 2 diabetes. Nat Metab 6 , 226–237 (2024). Elliott, A. et al. Distinct and shared genetic architectures of Gestational diabetes mellitus and Type 2 Diabetes Mellitus. medRxiv (2023) doi:10.1101/2023.02.16.23286014. Beysel, S. et al. Maternal genetic contribution to pre-pregnancy obesity, gestational weight gain, and gestational diabetes mellitus. Diabetol. Metab. Syndr. 11 , 37 (2019). Slieker, R. C. et al. Replication and cross-validation of type 2 diabetes subtypes based on clinical variables: an IMI-RHAPSODY study. Diabetologia 64 , 1982–1989 (2021). Franceschini, N., Frick, A. & Kopp, J. B. Genetic Testing in Clinical Settings. Am. J. Kidney Dis. 72 , 569–581 (2018). Yu, O. H. Y. & Shin, J.-Y. Treating type 2 diabetes: moving towards precision medicine. The Lancet. Digital health vol. 4 e851–e852 (2022). Recommendations | Type 2 diabetes in adults: management | Guidance | NICE. Hauner, H. The mode of action of thiazolidinediones. Diabetes. Metab. Res. Rev. 18 Suppl 2 , S10–5 (2002). Udler, M. et al. Multi-ancestry Polygenic Mechanisms of Type 2 Diabetes Elucidate Disease Processes and Clinical Heterogeneity. Res Sq (2023) doi:10.21203/rs.3.rs-3399145/v1. Huang, Q. Q. et al. Transferability of genetic loci and polygenic scores for cardiometabolic traits in British Pakistani and Bangladeshi individuals. Nat. Commun. 13 , 4664 (2022). Tables Tables 1 and 2 are available in the Supplementary Files section. Additional Declarations There is NO Competing Interest. Supplementary Files figs1.png figs2.png figs3.png figs4.png figs5.png figs6.png figs7.png figs8.png figs9.png figs10.png pPSsupp.docx Extended Methods nreditorialpolicychecklistSH17099089932.pdf Article File - Editorial Policy Checklist nrreportingsummarySH17099089933.pdf Article File - Reporting Summary Cite Share Download PDF Status: Published Journal Publication published 26 Nov, 2024 Read the published version in Nature Medicine → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {\"props\":{\"pageProps\":{\"initialData\":{\"identity\":\"rs-4027509\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":false,\"archivedVersions\":[],\"articleType\":\"Article\",\"associatedPublications\":[],\"authors\":[{\"id\":277049391,\"identity\":\"86061729-112c-4393-b084-1341a2c99d0e\",\"order_by\":0,\"name\":\"Moneeza Siddiqui\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABCklEQVRIiWNgGAWjYFADZgYDiQQGBh5+EOcBWIiNSC2SDUBOAlFaGIBawOQBAlp023sPPvjwx45Bvp15440HZYdljI9npwGts5NnkEhLwKbF7My5ZMMZPMkMBofZii0Szh3mMTvzdhtQS7Jhg0TaAaxabuSYSfNIAD3CzGMmkdgG1HIjF6SFOYFBIr0Bhxbz338M6hnkm6FajGeAtdTj02LGzJBwmIHhMFSLgQRYy2GgFhwOO3PGWLLnwHEeqF/SeSTOvN1skWBw3LCN5xl27x/vMfzw40+1nHz/4Y03f5RZ2/O352688aGiWp6fPc0AmxYY4IFQ4JgAmW1ATESiahkFo2AUjIJRgAAAVypZ41oLKWoAAAAASUVORK5CYII=\",\"orcid\":\"\",\"institution\":\"Queen Mary University of London\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Moneeza\",\"middleName\":\"\",\"lastName\":\"Siddiqui\",\"suffix\":\"\"},{\"id\":277049392,\"identity\":\"bcc3c261-afc9-47ec-a905-4384669ff1a8\",\"order_by\":1,\"name\":\"Sam Hodgson\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Queen Mary University of London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Sam\",\"middleName\":\"\",\"lastName\":\"Hodgson\",\"suffix\":\"\"},{\"id\":277049393,\"identity\":\"22f2c323-3678-4569-ad60-b6170a8c6593\",\"order_by\":2,\"name\":\"Alice Williamson\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Queen Mary University of London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Alice\",\"middleName\":\"\",\"lastName\":\"Williamson\",\"suffix\":\"\"},{\"id\":277049394,\"identity\":\"9156309a-bf8d-46ac-9e55-1ded747a1487\",\"order_by\":3,\"name\":\"Daniel Stow\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Queen Mary University of London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Daniel\",\"middleName\":\"\",\"lastName\":\"Stow\",\"suffix\":\"\"},{\"id\":277049395,\"identity\":\"d7a965bf-7789-4877-8e58-57c7db6c2068\",\"order_by\":4,\"name\":\"Benjamin Jacobs\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Preventive Neurology Unit, Wolfson Institute of Preventive Medicine, Queen Mary University of London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Benjamin\",\"middleName\":\"\",\"lastName\":\"Jacobs\",\"suffix\":\"\"},{\"id\":277049396,\"identity\":\"a90cbfcd-c53e-495c-85c8-7556984b06b0\",\"order_by\":5,\"name\":\"Miriam Samuel\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Queen Mary University of London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Miriam\",\"middleName\":\"\",\"lastName\":\"Samuel\",\"suffix\":\"\"},{\"id\":277049397,\"identity\":\"e264d2c9-40d3-4d7b-af6f-1b92565925e9\",\"order_by\":6,\"name\":\"Joseph Gafton\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Queen Mary University of London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Joseph\",\"middleName\":\"\",\"lastName\":\"Gafton\",\"suffix\":\"\"},{\"id\":277049398,\"identity\":\"a9858593-2b70-4d88-893e-dab9ea0e7a54\",\"order_by\":7,\"name\":\"Julia Zöllner\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"University College London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Julia\",\"middleName\":\"\",\"lastName\":\"Zöllner\",\"suffix\":\"\"},{\"id\":277049399,\"identity\":\"98d02e8f-f29d-4904-9cd6-5a915577f61c\",\"order_by\":8,\"name\":\"Marie Spreckley\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Queen Mary University of London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Marie\",\"middleName\":\"\",\"lastName\":\"Spreckley\",\"suffix\":\"\"},{\"id\":277049400,\"identity\":\"41cb4cff-fe9d-4678-ac85-dc896446f3a9\",\"order_by\":9,\"name\":\"Genes \\u0026 Health Research Team .\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Queen Mary University of London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Genes\",\"middleName\":\"\\u0026 Health Research Team\",\"lastName\":\".\",\"suffix\":\"\"},{\"id\":277049401,\"identity\":\"cd31e6bb-80d8-4dc4-900f-741a33b48a81\",\"order_by\":10,\"name\":\"Claudia Langenberg\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-5017-7344\",\"institution\":\"Queen Mary University of London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Claudia\",\"middleName\":\"\",\"lastName\":\"Langenberg\",\"suffix\":\"\"},{\"id\":277049402,\"identity\":\"946163cb-f2da-40b9-9cd6-6980c438f329\",\"order_by\":11,\"name\":\"David van Heel\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-0637-2265\",\"institution\":\"Queen Mary University of London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"David\",\"middleName\":\"van\",\"lastName\":\"Heel\",\"suffix\":\"\"},{\"id\":277049403,\"identity\":\"bda19646-4332-4f60-b1f7-9a32a97e51c5\",\"order_by\":12,\"name\":\"Rohini Mathur\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Queen Mary University of London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Rohini\",\"middleName\":\"\",\"lastName\":\"Mathur\",\"suffix\":\"\"},{\"id\":277049404,\"identity\":\"c47c5849-1836-4535-a90c-422a4c48c118\",\"order_by\":13,\"name\":\"Sarah Finer\",\"email\":\"\",\"orcid\":\"https://orcid.org/0000-0002-2684-4653\",\"institution\":\"Queen Mary University of London\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Sarah\",\"middleName\":\"\",\"lastName\":\"Finer\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2024-03-07 15:31:31\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-4027509/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-4027509/v1\",\"draftVersion\":[],\"editorialEvents\":[{\"content\":\"https://doi.org/10.1038/s41591-024-03317-8\",\"type\":\"published\",\"date\":\"2024-11-26T05:00:00+00:00\"}],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":52447499,\"identity\":\"f6011962-7068-41cd-9c30-bb9a99e612ac\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:35\",\"extension\":\"png\",\"order_by\":1,\"title\":\"Figure 1\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":194572,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eAssociation of pPS with type 2 diabetes and gestational diabetes risk\\u003c/strong\\u003e: Association of partitioned polygenic scores (pPS) with incident type 2 diabetes (T2D, n = 9771), gestational diabetes (GDM, n = 1740) and T2D after GDM (n = 960) in 44,189 unrelated individuals in the Genes \\u0026amp; health study. Results for each pPS are presented as beta per standard deviation of pPS with 95% confidence intervals after adjustment for sex, age, and ancestry.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"fig1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/5aa004b94c3eece57f802b67.png\"},{\"id\":52447498,\"identity\":\"0c8bf99b-a49b-44be-a831-7f670dcaf5ca\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:35\",\"extension\":\"png\",\"order_by\":2,\"title\":\"Figure 2\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":118059,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eAssociation of pPS with T2D age of onset. \\u003c/strong\\u003e2A: association between twelve T2D pPS and age at diagnosis of type 2 diabetes, presented as beta (in years) per standard deviation of pPS, estimated from a multivariate logistic regression model incorporating all twelve pPS and adjusted for sex and ancestry. 2B: Partial \\u003csup\\u003eR2 \\u003c/sup\\u003efor effect of twelve diabetes pPS on age at type 2 diabetes diagnosis, estimated from the same model.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"fig2.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/13a7b7db7d6171dcc344ae83.png\"},{\"id\":52447503,\"identity\":\"417a9814-ce25-43f0-8b6e-354b3d07e449\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:36\",\"extension\":\"jpg\",\"order_by\":3,\"title\":\"Figure 3\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":720572,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eAssociation of pPS with antidiabetic medication initiation and response (\\u003c/strong\\u003eA: Association of pPS with change in HBA1c in response to medication initiation, presented as beta per standard deviation, estimated from multivariate regression models adjusted for sex and ancestry. The change presented is % change in HbA1c from pre-treatment to on-treatment; HbA1c units are mmol/mol. B: Insulin-free survival from time of type 2 diabetes diagnosis in 9756 individuals for whom prescribing data was available (number of cases = 1756), presented as hazard ratios estimated from Cox proportional hazards survival models adjusted for sex and genetically-determined ancestry. Figure 3B:\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"3.jpg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/fcfded4104d351b4556556a3.jpg\"},{\"id\":52448116,\"identity\":\"c016d214-fc0d-49a5-9ad6-659e7e200704\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:35:36\",\"extension\":\"jpg\",\"order_by\":4,\"title\":\"Figure 4\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":622529,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eExtremes of genetic risk association with age (Panel A) and BMI (panel B) at diagnosis, and progression to microvascular complications (Panel C). For panels A and B, box plots are presented contrasting individuals in the top and bottom 10% of the genetic risk distributions for three key pPS (Obesity, beta Cell 2, and Lipodystrophy 1), and for individuals in the top and bottom 10% of both the Beta Cell 2 and Lipodystrophy distributions (rightmost panel). Distributions for all individuals with T2D are presented in the left-most panel for comparison. Distributions were compared using ANOVA. For panel C, hazard ratios are presented for each genetic risk extreme comparison, comparing complication-free survival from diagnosis between the bottom 10% of each pPS distribution (reference) and the top 10%. Hazard ratios were estimated from Cox proportional hazards models adjusted for sex and ancestry. Further data are presented in Table S7 (Schoenfeld residuals for survival models) and Fig S10 (illustrative Kaplan-Meier survival plots for positive results).\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"4.jpg\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/55cfbb6b19462823e59b6ff9.jpg\"},{\"id\":69981988,\"identity\":\"f2927b8c-017f-4df0-9c7c-d23d13b934a6\",\"added_by\":\"auto\",\"created_at\":\"2024-11-27 08:07:47\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":2597313,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"manuscript.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/7f09a4be-df86-4435-b911-bbe6be80c38b.pdf\"},{\"id\":52447502,\"identity\":\"f0e353f8-b924-479d-82e8-243bc5ccd42b\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:36\",\"extension\":\"png\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":1041515,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cbr\\u003e\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"figs1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/2536beb2b80b55d005b3f9d2.png\"},{\"id\":52447500,\"identity\":\"5320a89d-21fe-4be7-be21-9fa39f26e1ab\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:35\",\"extension\":\"png\",\"order_by\":2,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":390421,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"figs2.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/90e1b09619602131ea75b951.png\"},{\"id\":52447512,\"identity\":\"0f34568b-2224-4b99-a78b-507d1109a6f0\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:37\",\"extension\":\"png\",\"order_by\":3,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":364790,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"figs3.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/8d15e482e146296a28074433.png\"},{\"id\":52447505,\"identity\":\"a04ca5cd-9f97-4616-af4b-28cd6820708d\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:36\",\"extension\":\"png\",\"order_by\":4,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":241669,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"figs4.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/aefc4876c3ec80f3e4fee9fd.png\"},{\"id\":52448117,\"identity\":\"51573cc6-429c-42a1-86c4-dc2f1cd375af\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:35:36\",\"extension\":\"png\",\"order_by\":5,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":321801,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"figs5.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/6b55da628e2e3dd6f81b2e0f.png\"},{\"id\":52447510,\"identity\":\"c4d8569f-3056-4c68-9730-3ab668eda838\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:36\",\"extension\":\"png\",\"order_by\":6,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":135535,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"figs6.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/df7f8565d84a62ac5eaa5b2b.png\"},{\"id\":52447511,\"identity\":\"91856e1d-b4bd-4989-9638-c34ac4c402df\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:36\",\"extension\":\"png\",\"order_by\":7,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":2160579,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"figs7.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/38629284868abd68f42f494f.png\"},{\"id\":52447513,\"identity\":\"07b41480-9b21-4f8f-8f27-c7f31d3f67c6\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:37\",\"extension\":\"png\",\"order_by\":8,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":514654,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"figs8.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/831c4e3df7fb735862c5e83a.png\"},{\"id\":52447506,\"identity\":\"f362fa39-44a7-4e84-a08a-2e07da9785bf\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:36\",\"extension\":\"png\",\"order_by\":9,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":92628,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"figs9.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/4a9f7222f95391246ddf08c7.png\"},{\"id\":52447507,\"identity\":\"0920f419-08ea-46aa-b9b7-64c5d6543c11\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:36\",\"extension\":\"png\",\"order_by\":10,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":137926,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"figs10.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/9cda4f50f0d03c5a9491a988.png\"},{\"id\":52447501,\"identity\":\"e3e0f412-0529-44f1-8c74-9dfc615ce826\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:36\",\"extension\":\"docx\",\"order_by\":11,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":15191,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eExtended Methods\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"pPSsupp.docx\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/ed771b4db88ffa9af785cf51.docx\"},{\"id\":52447514,\"identity\":\"79a67e26-4432-4328-b12b-790a733dfefe\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:37\",\"extension\":\"pdf\",\"order_by\":13,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":1666915,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eArticle File - Editorial Policy Checklist\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"nreditorialpolicychecklistSH17099089932.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/c2b0a5e4d7856a1b1d159e03.pdf\"},{\"id\":52447515,\"identity\":\"11a15e60-16d0-4208-b602-e69149e7a8db\",\"added_by\":\"auto\",\"created_at\":\"2024-03-11 18:27:37\",\"extension\":\"pdf\",\"order_by\":14,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":1666028,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eArticle File - Reporting Summary\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"nrreportingsummarySH17099089933.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4027509/v1/9de94081d36cd0f08d8023db.pdf\"}],\"financialInterests\":\"There is \\u003cb\\u003eNO\\u003c/b\\u003e Competing Interest.\",\"formattedTitle\":\"Partitioned polygenic scores highlight role of beta-cell function and unfavourable fat distribution patterns in young onset type 2 diabetes in south Asians\",\"fulltext\":[{\"header\":\"Background\",\"content\":\"\\u003cp\\u003eType 2 diabetes is common, particularly in South Asian individuals, among whom the prevalence is estimated to be 12.7% globally, and as high as 30% in Pakistan\\u003csup\\u003e\\u003cspan citationid=\\\"CR1\\\" class=\\\"CitationRef\\\"\\u003e1\\u003c/span\\u003e\\u003c/sup\\u003e, \\u003csup\\u003e\\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e\\u003c/sup\\u003e. Compared to white Europeans, South Asians tend to be diagnosed with type 2 diabetes at younger ages and with lower Body Mass Index (BMI)\\u003csup\\u003e\\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e\\u003c/sup\\u003e. This phenomenon exists even when the environment isn't shared: both migrant and non-migrant Asian Indians are up to 4 times more likely to be diagnosed with type 2 diabetes young (\\u0026lt;\\u0026thinsp;40 years) and lean compared to white Europeans\\u003csup\\u003e\\u003cspan citationid=\\\"CR4\\\" class=\\\"CitationRef\\\"\\u003e4\\u003c/span\\u003e\\u003c/sup\\u003e, and up to 3 times more likely to develop gestational diabetes mellitus (GDM)\\u003csup\\u003e\\u003cspan citationid=\\\"CR5\\\" class=\\\"CitationRef\\\"\\u003e5\\u003c/span\\u003e\\u003c/sup\\u003e.\\u003c/p\\u003e \\u003cp\\u003eDespite this increased risk, South Asian individuals have historically been poorly represented in genetic studies, including those relating to diabetes.\\u003csup\\u003e\\u003cspan citationid=\\\"CR6\\\" class=\\\"CitationRef\\\"\\u003e6\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e\\u003c/sup\\u003e. This is of particular relevance as in recent years there has been increased understanding of type 2 diabetes as a heterogeneous condition comprising multiple subgroups or clusters (typically referred to as endotypes), determined on the basis of phenotypic and genetic characteristics, such as endotypes related to obesity or lipodystrophy\\u003csup\\u003e\\u003cspan citationid=\\\"CR8\\\" class=\\\"CitationRef\\\"\\u003e8\\u003c/span\\u003e\\u003c/sup\\u003e. Phenotypic endotyping approaches have shown a preponderance of Severe Insulin Deficient Diabetes (SIDD) in south Asians observed in two independent cohorts in India: 35% and 52% respectively\\u003csup\\u003e9,10\\u003c/sup\\u003e. This contrasts with white Europeans, among whom insulin resistance plays a more prominent role\\u003csup\\u003e\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e\\u003c/sup\\u003e. Research defining genetically determined endotypes has been largely restricted to White European participants, with poor representation of south Asian populations in more recent multi-ancestry efforts \\u003csup\\u003e\\u003cspan additionalcitationids=\\\"CR12 CR13\\\" citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR14\\\" class=\\\"CitationRef\\\"\\u003e14\\u003c/span\\u003e\\u003c/sup\\u003e. The latest multi-ancestry meta-GWA study has specifically highlighted the utility of partitioned polygenic scores relative to polygenic scores in identifying individuals at risk of diabetes complications. However, only 2% of the 2.5\\u0026nbsp;million individuals in the study were of south Asian ancestry, limiting the amount of representation of these individuals in pooled multi-ancestral analyses.\\u003csup\\u003e\\u003cspan citationid=\\\"CR15\\\" class=\\\"CitationRef\\\"\\u003e15\\u003c/span\\u003e\\u003c/sup\\u003e\\u003c/p\\u003e \\u003cp\\u003eUnlike polygenic scores that predict overall risk of an outcome (like type 2 diabetes) by summing the genetic burden of risk from genetic variants identified in genome-wide association studies, partitioned polygenic scores (pPS) offer insight into the mechanistic and aetiological factors driving this risk by leveraging clustering approaches to estimate genetic propensity to underlying mechanistic pathways like beta cell dysfunction, obesity, lipid metabolism or lipodystrophy\\u003csup\\u003e\\u003cspan citationid=\\\"CR16\\\" class=\\\"CitationRef\\\"\\u003e16\\u003c/span\\u003e\\u003c/sup\\u003e. They have been shown to be associated with risk of diabetes complications, e.g a pPS for Beta-cell function, predictive of insulin deficiency, was associated impaired renal function\\u003csup\\u003e\\u003cspan citationid=\\\"CR17\\\" class=\\\"CitationRef\\\"\\u003e17\\u003c/span\\u003e\\u003c/sup\\u003e.\\u003c/p\\u003e \\u003cp\\u003ePopulations like South Asians, who are not well represented in studies elucidating the complex architecture of type 2 diabetes, may not benefit from advances in understanding the condition. Genetically-determined diabetes endotypes relating to distinct aetiological mechanisms have been well-described and widely validated in white European populations \\u003csup\\u003e\\u003cspan citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR18\\\" class=\\\"CitationRef\\\"\\u003e18\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR19\\\" class=\\\"CitationRef\\\"\\u003e19\\u003c/span\\u003e\\u003c/sup\\u003e However, as many aspects of genetic architecture and clinical phenotypes of type 2 diabetes differ between South Asians and White Europeans\\u003csup\\u003e209\\u003c/sup\\u003e, it remains unclear to what extent these pPS might transfer to understanding type 2 diabetes risk and progression in individuals of South Asian ancestry. Furthermore, the phenotypic traits and clinically relevant characteristics associated with an individual having high genetic risk across multiple distinct aetiological pathways, which we term \\u0026ldquo;pPS extremes\\u0026rdquo;, has not been explored, but could offer opportunities to identify individuals at high risk of clinically relevant outcomes.\\u003c/p\\u003e \\u003cp\\u003eWe used Genes \\u0026amp; Health, a long-term community-based study with genetics and linked electronic health and prescribing data for over 51,000 British Pakistani and Bangladeshi individuals living in the UK \\u003csup\\u003e\\u003cspan citationid=\\\"CR21\\\" class=\\\"CitationRef\\\"\\u003e21\\u003c/span\\u003e\\u003c/sup\\u003eto investigate whether pPS can help unravel the aetiological factors driving young onset type 2 diabetes and clinically relevant related outcomes.\\u003c/p\\u003e\"},{\"header\":\"Methods\",\"content\":\"\\u003cdiv id=\\\"Sec3\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eCohort profile\\u003c/h2\\u003e \\u003cp\\u003eGenes \\u0026amp; Health is a long-term, community-based study of British Pakistani and British Bangladesh individuals aged 16 years and above living in the United Kingdom\\u003csup\\u003e\\u003cspan citationid=\\\"CR21\\\" class=\\\"CitationRef\\\"\\u003e21\\u003c/span\\u003e\\u003c/sup\\u003e. At recruitment, participants provide a saliva sample for genotyping, complete a short questionnaire on basic demographic information, and consent to linkage at primary and secondary care electronic health records. Since recruitment began in 2015 over 50,000 participants have been recruited, with linked genetic and electronic health record information available for 44,189 as of July 2023 (number of type 2 diabetes cases\\u0026thinsp;=\\u0026thinsp;9771), and 51,170 as of February 2024. We used the additional data release of 7383 newly genotyped individuals, of whom 1907 had T2D, to replicate key findings.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec4\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eInclusion and exclusion criteria\\u003c/h2\\u003e \\u003cp\\u003eWe used no specific inclusion criteria. We excluded individuals with clinical codes consistent with type 1 diabetes, MODY diabetes, or causes of secondary diabetes such as cystic fibrosis and pancreatectomy.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec5\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eGenetic data processing and curation\\u003c/h2\\u003e \\u003cp\\u003eGenotyping was performed on Illumina Infinium Global Screening Array v3 with additional multi-disease variants. Quality control was performed in line with previous descriptions \\u003csup\\u003e\\u003cspan citationid=\\\"CR22\\\" class=\\\"CitationRef\\\"\\u003e22\\u003c/span\\u003e,\\u003cspan citationid=\\\"CR23\\\" class=\\\"CitationRef\\\"\\u003e23\\u003c/span\\u003e\\u003c/sup\\u003e. In brief, variants with call rates\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.99 and /or MAF\\u0026thinsp;\\u0026lt;\\u0026thinsp;1% were excluded. We excluded individuals unlikely to have genetically-inferred Pakistani or Bangladeshi ancestry. Imputation was performed using TopMED R2 panel. We excluded SNPs with low imputation scores for INFO (\\u0026lt;\\u0026thinsp;0.3) or MAF\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.1%. Further details on genetic data processing, curation, and ancestry inference are presented in the supplementary information.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec6\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eElectronic Health Record data processing and curation\\u003c/h2\\u003e \\u003cp\\u003eWe curated routine UK electronic health record (EHR) data from primary care (SNOMED coded) and secondary care (ICD10 coded) sources. Data were combined without mapping between coding formats. For each clinical code, we took the earliest ever measure recorded in a participant\\u0026rsquo;s medical records, excluding erroneous code dates preceding the participants\\u0026rsquo;s recorded date of birth.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec7\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eExposures:\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec8\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003ePartitioned polygenic risk score (pPS) construction and ancestry correction\\u003c/h2\\u003e \\u003cp\\u003eWe calculated pPS for twelve diabetes-associated genetically determined endotypes described by Smith et al derived from high-throughput genetic clustering techniques in White European, East Asian and African ancestry individuals, using only SNPs above the authors\\u0026rsquo; specified inclusion threshold (cluster weight\\u0026thinsp;\\u0026gt;\\u0026thinsp;0.78)\\u003csup\\u003e18\\u003c/sup\\u003e. These comprise three endotypes related to glucose sensing, insulin secretion and insulin production (Beta cell 1, Beta cell 2, Proinsulin respectively), three clusters related to insulin resistance and unfavourable adiposity (Obesity, Lipodystrophy 1, Lipodystrophy 2) and 6 clusters with unclear effects on insulin resistance and deficiency (Liver/lipid, ALP [alkaline phosphatase] negative, Hyper insulin secretion, Cholesterol, Sex Hormone Binding Globulin / Lipoprotein A (SHBG/LpA), and bilirubin). We additionally calculated scores for a global T2D polygenic risk score (T2D PRS), using a previously-published score\\u003csup\\u003e\\u003cspan citationid=\\\"CR24\\\" class=\\\"CitationRef\\\"\\u003e24\\u003c/span\\u003e\\u003c/sup\\u003e with the highest beta and area under the receiver-operator curve for predicting incident T2D in the Genes \\u0026amp; Health cohort.\\u003c/p\\u003e \\u003cp\\u003ePrincipal components analysis of genetic data shows distinct population structure for people of Bangladeshi and Pakistani ancestries\\u003csup\\u003e\\u003cspan citationid=\\\"CR25\\\" class=\\\"CitationRef\\\"\\u003e25\\u003c/span\\u003e\\u003c/sup\\u003e, and we observed differences in pPS distribution between these groups. Therefore, to maximise power and facilitate combined analyses of all individuals we regressed out the effect of the first 5 genetic principal components from each pPS, using an approach described in\\u003csup\\u003e\\u003cspan citationid=\\\"CR25\\\" class=\\\"CitationRef\\\"\\u003e25\\u003c/span\\u003e\\u003c/sup\\u003e (further details presented in \\u003cb\\u003eSupplementary methods\\u003c/b\\u003e and \\u003cb\\u003eFig \\u003cspan refid=\\\"MOESM1\\\" class=\\\"InternalRef\\\"\\u003eS1\\u003c/span\\u003e).\\u003c/b\\u003e\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec9\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003epPS extremes\\u003c/h2\\u003e \\u003cp\\u003eWe defined pPS \\u0026ldquo;extremes\\u0026rdquo; as scores in the top or bottom 10% of each residualized pPS distribution, and \\u0026ldquo;combined extremes\\u0026rdquo; as individuals with scores in the top or bottom 10% of multiple pPS distributions.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec10\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eUKBiobank - cross-ancestry differences in genetic burden\\u003c/h2\\u003e \\u003cp\\u003eWe used UK Biobank to compare the distribution of pPS between individuals with T2D of European (EUR) and South Asian (SAS) ancestry. Differences in distribution of pPS across ancestry groups were assessed using ANOVA for normally distributed and Kruskall-Wallis testing for all other pPS. All analyses in UK Biobank were conducted under application ID 44448.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec11\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eMedications\\u003c/h2\\u003e \\u003cp\\u003eDiabetes-controlling medication classes were defined according to method of action: insulin secretagogues (sulfonylureas and meglitinide), incretin mimetics (GLP1 receptor agonists and DDP4 inhibitors), insulin sensitizers (pioglitazone and thiazolidinediones), renal tubular glucose reabsorption modifiers (SGLT2 inhibitors), in addition to metformin and insulin. Initiation of medication was defined as the first instance of medication prescription in the EHR. For treatment response analyses (described in further detail in the \\u0026ldquo;outcomes\\u0026rdquo; section below), concurrently prescribed medications were defined as medications from another class prescribed within a conservative window of 6 months before or after initiation of the medication being treated as the exposure.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec12\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eOutcomes:\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec13\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eDiabetes phenotypes\\u003c/h2\\u003e \\u003cp\\u003eThese categorical outcome variables were defined using ICD-10 and SNOMED codelists adapted from the AI-MULTIPLY resource\\u003csup\\u003e\\u003cspan citationid=\\\"CR26\\\" class=\\\"CitationRef\\\"\\u003e26\\u003c/span\\u003e\\u003c/sup\\u003e (\\u003cb\\u003eTable \\u003cspan refid=\\\"MOESM1\\\" class=\\\"InternalRef\\\"\\u003eS1\\u003c/span\\u003e).\\u003c/b\\u003e Diabetes phenotypes included type 2 diabetes, gestational diabetes mellitus, and incident type 2 diabetes after gestational diabetes mellitus. Diabetes-related complications were defined as microvascular (nephropathy, neuropathy, and retinopathy) and macrovascular (coronary artery disease, cerebrovascular disease, and peripheral vascular disease).(\\u003cb\\u003eTable \\u003cspan refid=\\\"MOESM1\\\" class=\\\"InternalRef\\\"\\u003eS1\\u003c/span\\u003e).\\u003c/b\\u003e\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec14\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eAge at Diagnosis\\u003c/h2\\u003e \\u003cp\\u003eDate of diagnosis for each outcome was defined as the earliest recorded date in either primary or secondary care above the age of 18.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec15\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eQuantitative traits\\u003c/h2\\u003e \\u003cp\\u003eQuantitative outcomes were, unless otherwise stated, defined as the measure taken closest to the date of type 2 diabetes within 1 year (before and/or after) and included age, body mass index (BMI), waist circumference, glycated haemoglobin (HbA1c), Fasting and random blood glucose, low- and high-density lipoprotein cholesterol (LDL-C and HDL-C), serum triglycerides, alkaline phosphatase (ALP), and alanine transaminase (ALT). Because diabetes-related traits may rapidly change after diagnosis and/or initiation of treatment, for each trait, the value closest to the time of diagnosis was used. In addition to traits at diagnosis, we explored the number of medication classes (as defined above) an individual was prescribed in 5 years, and the change in HbA1c from time of diagnosis to 5 years (the HbA1c closest in time to 5 years from diagnosis date was taken, and only values between 4\\u0026ndash;6 years post-diagnosis included in analysis).\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec16\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eResponse to glucose-lowering treatment\\u003c/h2\\u003e \\u003cp\\u003eFor treatment response analyses, medication data was extracted from the primary care EHR. In line with pharmacogenomic studies\\u003csup\\u003e\\u003cspan citationid=\\\"CR27\\\" class=\\\"CitationRef\\\"\\u003e27\\u003c/span\\u003e\\u003c/sup\\u003e, treatment response was defined as the percentage change between the most recent HbA1c in 6 months before medication initiation, and lowest HbA1c in the 1 year after initiation, as a proportion of pre-medication HbA1c (i.e., %change from prior to initiation).\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec17\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eStatistical analyses\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec18\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eDescriptive Analysis\\u003c/h2\\u003e \\u003cp\\u003eWe calculated mean values for quantitative traits at diagnosis and five years post-diagnosis, stratified by ancestral group, and compared these using ANOVA.\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec19\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eMultivariable Analysis\\u003c/h2\\u003e \\u003cp\\u003eWe described the association of each pPS (the exposure) with each diabetes phenotype outcome, using multivariable logistic regression models adjusted for age, sex, and ancestry, to estimate the per-standard-deviation increase in odds of diabetes phenotype between diabetes phenotype cases and non-diabetic controls. We estimated the association of each pPS (the exposure) with diabetes-related traits at the time of diagnosis (the outcome). To allow comparison of effects of pPS between quantitative traits at the time of diagnosis, each quantitative trait was scaled to a normal distribution, and the beta per standard deviation of pPS presented for each trait, estimated from multivariable logistic regression models adjusted for age, sex and ancestry. Multivariable linear regression was used to estimate the effect of pPS on age of T2D diagnosis, adjusted for ancestry and sex; partial R\\u003csup\\u003e\\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e\\u003c/sup\\u003e was calculated with the R package \\u0026ldquo;partialR2\\u0026rdquo;. We hypothesised a-priori that associations may differ between sexes and ancestry groups; because of this, we sex-stratified and ancestry-stratified analyses were also performed. For pharmacogenomic analysis of treatment response, association between each pPS (the exposure) and HbA1c change in response to medication (the outcome) was estimated from multivariable logistic regression models adjusted for age, sex and ancestry, as well as concurrently prescribed antidiabetic medication from all other classes within 6 months before or after initiation of each. We meta-analysed treatment response analyses from the discovery and replication samples using fixed effects models.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec20\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eSurvival analysis\\u003c/h2\\u003e \\u003cp\\u003eWe constructed survival models starting from each individual\\u0026rsquo;s date of diagnosis, running until last data extraction, for two outcome categories: initiation of insulin, and progression to diabetes-related complications. We explored the association of each pPS with each complication outcome using Cox proportional hazards models adjusted for age, sex and ancestry. We calculated Schoenfeld residuals for each model to check assumptions of proportionality.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec21\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eSoftware and statistical computing\\u003c/h2\\u003e \\u003cp\\u003eGenotype curation and partitioned polygenic risk score calculation was performed using Plink v2.0\\u003csup\\u003e28\\u003c/sup\\u003e. Statistical analyses were performed using R V4.2.3.\\u003c/p\\u003e \\u003cp\\u003eRole of the funding source\\u003c/p\\u003e \\u003cp\\u003eThe funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec22\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eReporting\\u003c/h2\\u003e \\u003cp\\u003eWe report this study following the STREGA guidelines\\u003csup\\u003e\\u003cspan citationid=\\\"CR29\\\" class=\\\"CitationRef\\\"\\u003e29\\u003c/span\\u003e\\u003c/sup\\u003e.\\u003c/p\\u003e \\u003c/div\\u003e\"},{\"header\":\"Results\",\"content\":\"\\u003cp\\u003e\\u003cem\\u003eCharacteristics of populations studied and differences in clinical outcomes between British Bangladeshis and Pakistanis.\\u003c/em\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eWe analysed data from 9,771 British Pakistani and Bangladeshi individuals with a diagnosis of type 2 diabetes and 35,190 diabetes-free controls; their demographic and clinical information stratified by ancestry and sex, is shown in \\u003cstrong\\u003eTable\\u0026nbsp;1.\\u003c/strong\\u003e We observed differences in clinical features and prevalence of diabetes-related outcomes by ancestral group: compared to Pakistani individuals, Bangladeshi individuals tended to be diagnosed with type 2 diabetes earlier in life, at lower BMI, and had higher rates of prior GDM. Consequently, Bangladeshis had a higher prevalence of type 2 diabetes. However, at diagnosis and in the 5 years following diagnosis, Pakistanis had higher HbA1c and BMI. Overall, Pakistanis were more likely to be prescribed insulin and develop nephropathy than Bangladeshis, and had higher HbA1c 5 years after diagnosis, despite being prescribed antidiabetic medications from more classes.\\u003c/p\\u003e\\n\\u003cdiv id=\\\"Sec24\\\" class=\\\"Section2\\\"\\u003e\\n\\u003ch2\\u003eAssociation of partitioned polygenic scores and quantitative traits at time of diagnosis\\u003c/h2\\u003e\\n\\u003cp\\u003eWe observed numerous expected associations between single pPS and traits at time of diagnosis (\\u003cstrong\\u003eFig \\u003cspan class=\\\"InternalRef\\\"\\u003eS2\\u003c/span\\u003e, Table \\u003cspan class=\\\"InternalRef\\\"\\u003eS2\\u003c/span\\u003e)\\u003c/strong\\u003e, including associations between beta cell related pPSs (Beta Cell 1, Beta Cell 2, Proinsulin), and lower BMI, and greater HbA1c, fasting and random blood glucose at diagnosis. Lipodystrophy 1, was associated with lower BMI and waist circumference and raised ALT at time of diagnosis. Both lipodystrophy-related pPSs were associated with lower HDL and raised triglycerides, whereas the Cholesterol pPS was associated with raised HDL and lower triglycerides. The obesity pPS was associated with greater BMI and waist circumferences.\\u003c/p\\u003e\\n\\u003cp\\u003eSeveral pPS demonstrated reasonably strong correlations with one another (\\u003cstrong\\u003eFig \\u003cspan class=\\\"InternalRef\\\"\\u003eS3\\u003c/span\\u003e, Table \\u003cspan class=\\\"InternalRef\\\"\\u003eS3\\u003c/span\\u003e)\\u003c/strong\\u003e, with strongest associations observed between pPS with similar proposed mechanisms of action, eg between Lipodystrophy 1 and Lipodystrophy 2 (Pearson R\\u003csup\\u003e2\\u003c/sup\\u003e\\u0026thinsp;=\\u0026thinsp;0.50), and for Beta Cell 1 and Beta Cell 2 (R\\u003csup\\u003e2\\u003c/sup\\u003e\\u0026thinsp;=\\u0026thinsp;0.46). Weaker correlations were observed between pPS acting via distinct pathophysiological pathways, e.g. Beta Cell 1 and Lipodystrophy 1 (R\\u003csup\\u003e2\\u003c/sup\\u003e\\u0026thinsp;=\\u0026thinsp;0.35), and Beta Cell 1 and Obesity (R\\u003csup\\u003e2\\u003c/sup\\u003e\\u0026thinsp;=\\u0026thinsp;0.31). These results suggest they are unlikely to be exerting independent effects.\\u003c/p\\u003e\\n\\u003cdiv id=\\\"Sec25\\\" class=\\\"Section3\\\"\\u003e\\n\\u003ch2\\u003eAncestral differences in pPS Distribution\\u003c/h2\\u003e\\n\\u003cp\\u003eNoting the variation in clinical features of diabetes between British Bangladeshi and Pakistani individuals in our cohort, we first characterised distributions of unmodified pPS (not corrected for PC-defined ancestry). We observed higher unmodified pPS scores in Bangladeshi individuals, notably for Beta Cell 2, Obesity, and Lipodystrophy 1 \\u003cstrong\\u003e(Fig \\u003cspan class=\\\"InternalRef\\\"\\u003eS1\\u003c/span\\u003e)\\u003c/strong\\u003e (ANOVA p between Bangladeshis and Pakistanis\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.001 for all three pPS). To maximise power in subsequent analyses, we pooled the two sub-ancestries using PC-corrected pPS. We compared distributions of pPS between individuals with type 2 diabetes of White European and South Asian ancestry in UK-Biobank, observing higher scores among South Asians for several pPS including Beta Cell 2 (p\\u0026thinsp;=\\u0026thinsp;3e-8) and Lipodystrophy 1(p\\u0026thinsp;=\\u0026thinsp;1.95e-157), while Obesity pPS was higher among Europeans(p\\u0026thinsp;=\\u0026thinsp;1.09e-47) \\u003cstrong\\u003e(Fig \\u003cspan class=\\\"InternalRef\\\"\\u003eS4\\u003c/span\\u003e, Table \\u003cspan class=\\\"InternalRef\\\"\\u003eS4\\u003c/span\\u003e).\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cem\\u003eAssociations between Partitioned polygenic scores are associated with type 2 diabetes, GDM, and transition from GDM to T2D\\u003c/em\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eFor all pPS other than bilirubin, and a type 2 diabetes PRS, scores were higher among T2D cases, GDM cases, and individuals developing T2D after GDM, compared to non-diabetic controls \\u003cstrong\\u003e(Fig \\u003cspan class=\\\"InternalRef\\\"\\u003eS5\\u003c/span\\u003e)\\u003c/strong\\u003e. We observed associations between pPS and T2D risk,, after adjustment for sex and ancestry in 44,189 individuals (number of T2D cases\\u0026thinsp;=\\u0026thinsp;9771) (Fig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e). The strongest associations between pPS and T2D were for beta cell function- mediated endotypes Beta Cell 1 - related to glucose sensing (beta per standard deviation (SD) 0.39, 95% CI 0.35\\u0026ndash;0.42, p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.001) and Beta Cell 2 - related to insulin deficiency (beta per SD 0.32, 95% CI 0.28\\u0026ndash;0.35, p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.001).\\u003c/p\\u003e\\n\\u003cp\\u003eAll pPS, other than Bilirubin were associated with both GDM and incident T2D after GDM were observed in 5430 unrelated individuals with a history of at least one pregnancy \\u003cstrong\\u003e(\\u003c/strong\\u003eFig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e\\u003cstrong\\u003e)\\u003c/strong\\u003e. As with T2D, the strongest association between pPS and GDM was with pPSs related to Beta Cell 1 (beta per SD 0.24, 95% CI 0.18\\u0026ndash;0.29) and Beta Cell 2 (beta per SD 0.23, 95% CI 0.17\\u0026ndash;0.28). Similar patterns were observed in association of pPS with incident T2D after GDM, with strongest associations for Beta Cell 1 and 2.\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec26\\\" class=\\\"Section3\\\"\\u003e\\n\\u003ch2\\u003eGenetic predisposition to insulin deficiency drives earlier age of onset in type 2 diabetes\\u003c/h2\\u003e\\n\\u003cp\\u003eAlthough all pPS other than bilirubin were associated with earlier age of T2D onset \\u003cstrong\\u003e(Fig \\u003cspan class=\\\"InternalRef\\\"\\u003eS6\\u003c/span\\u003e)\\u003c/strong\\u003e, many of these effects were non-significant in a multivariate regression model incorporating all twelve pPS \\u003cstrong\\u003e(\\u003c/strong\\u003eFig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003eA\\u003cstrong\\u003e).\\u003c/strong\\u003e The pPS most strongly associated with age of onset were Beta Cell 2 (diagnosis 1.1 years earlier in life per SD, 95% CI 0.9\\u0026ndash;1.4 years, p\\u0026thinsp;=\\u0026thinsp;3e-16), Obesity (0.57 years per SD, 95% CI 0.30\\u0026ndash;0.83 years, p\\u0026thinsp;=\\u0026thinsp;3e-5) and Lipodystrophy 1 (0.54 years per SD, 95% 0.25\\u0026ndash;0.82 years, p\\u0026thinsp;=\\u0026thinsp;2e-4)(Fig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003eA\\u003cstrong\\u003e).\\u003c/strong\\u003e In this model, the pPS with the highest partial R\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e\\u003c/sup\\u003e was Beta Cell 2 (0.007), followed by Obesity (R\\u003csup\\u003e2\\u003c/sup\\u003e\\u0026thinsp;=\\u0026thinsp;0.002) and Lipodystrophy 1 (0.001)\\u003cstrong\\u003e(\\u003c/strong\\u003eFig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003eB\\u003cstrong\\u003e)\\u003c/strong\\u003e.\\u003c/p\\u003e\\n\\u003cp\\u003eIn ancestry- and sex-stratified analyses we found strongest effects for both beta cell 2 and lipodystrophy 1 in Bangladeshi females (Beta per SD\\u0026thinsp;=\\u0026thinsp;1.93, P\\u0026thinsp;=\\u0026thinsp;7.2e-20 and 1.70, P\\u0026thinsp;=\\u0026thinsp;9.3e-15, respectively, (\\u003cstrong\\u003eFig \\u003cspan class=\\\"InternalRef\\\"\\u003eS7\\u003c/span\\u003e\\u003c/strong\\u003e). We found no significant interaction between pPS and sex in ancestry-specific analyses in individuals of either Pakistani or Bangladeshi ancestry (\\u003cstrong\\u003eFig \\u003cspan class=\\\"InternalRef\\\"\\u003eS7\\u003c/span\\u003e)\\u003c/strong\\u003e. Subsequent analyses were focused on these three endotypes, as key drivers of early onset T2D risk.\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec27\\\" class=\\\"Section3\\\"\\u003e\\n\\u003ch2\\u003eAssociation between partitioned polygenic risk scores and treatment response\\u003c/h2\\u003e\\n\\u003cp\\u003eWe observed associations between pPS and response to oral antidiabetic medications, measured as the % change in HbA1c (in mmol/mol) following medication initiation \\u003cstrong\\u003e(\\u003c/strong\\u003eFig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003eD, \\u003cstrong\\u003eFig \\u003cspan class=\\\"InternalRef\\\"\\u003eS8\\u003c/span\\u003e, Table \\u003cspan class=\\\"InternalRef\\\"\\u003eS4\\u003c/span\\u003e)\\u003c/strong\\u003e. We replicated our initial findings in a replication sample of 7383 individuals (number of T2D cases\\u0026thinsp;=\\u0026thinsp;1907) in Genes \\u0026amp; Health, then meta-analysed our discovery and replication results. Higher scores across all pPS were generally associated with negative, or no, response to oral antidiabetic medication initiation (\\u003cstrong\\u003eFig \\u003cspan class=\\\"InternalRef\\\"\\u003eS8\\u003c/span\\u003eA)\\u003c/strong\\u003e, while across the whole sample of included participants, HbA1c typically declined after initiation of each medication class (median % change in HbA1c ranged from \\u0026minus;\\u0026thinsp;14.6% (metformin) to -20.3% (sulfolnylurea)(\\u003cstrong\\u003eFig \\u003cspan class=\\\"InternalRef\\\"\\u003eS8\\u003c/span\\u003eB, Table \\u003cspan class=\\\"InternalRef\\\"\\u003eS5\\u003c/span\\u003e).\\u003c/strong\\u003e Higher Beta Cell 2 pPS score was associated with increased HbA1c after initiation of thiazolidinediones (meta-analysed HbA1c increase 1.68% per SD, 95% CI 0.37\\u0026ndash;2.99, p\\u0026thinsp;=\\u0026thinsp;0.01), SGLT2i (1.22% per SD, 95% CI 0.56\\u0026ndash;1.89, p\\u0026thinsp;=\\u0026thinsp;3e-4), and metformin (0.51%, 95% CI 0.01\\u0026ndash;1.01, p\\u0026thinsp;=\\u0026thinsp;0.047). Higher Liver-Lipid pPS was associated with decrease in HbA1c after initiation of metformin (-0.65% per SD, 95% CI -0.16 - -1.13, p\\u0026thinsp;=\\u0026thinsp;0.008), while the T2D PRS was not associated with treatment response to metformin, thiazolidinediones, SGLT2 inhibitors, or sulfonylureas \\u003cstrong\\u003e(Fig \\u003cspan class=\\\"InternalRef\\\"\\u003eS8\\u003c/span\\u003eA)\\u003c/strong\\u003e.\\u003c/p\\u003e\\n\\u003cp\\u003eThe Beta Cell 2 and Lipodystrophy 1 pPSs were predictive of progression to insulin treatment in adjusted survival models (HR per SD 1.08, 95% CI 1.03\\u0026ndash;1.12, p\\u0026thinsp;=\\u0026thinsp;0.0008; HR per SD 1.11 95% CI 1.06\\u0026ndash;1.15, p\\u0026thinsp;=\\u0026thinsp;0.0002), while the Obesity pPS was not associated with progression to insulin treatment.(Fig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003eB\\u003cstrong\\u003e)\\u003c/strong\\u003e. Results for other pPS are shown in \\u003cstrong\\u003eFig \\u003cspan class=\\\"InternalRef\\\"\\u003eS9\\u003c/span\\u003e.\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec28\\\" class=\\\"Section2\\\"\\u003e\\n\\u003ch2\\u003eGenetic Risk Extremes Predict Lean and Early Onset Diabetes, and Progression To Complications\\u003c/h2\\u003e\\n\\u003cp\\u003eFinally, we explored whether extremes of genetic risk, defined as being in the top versus bottom decile of a pPS distribution (high or low risk, respectively) were associated with T2D phenotype \\u003cstrong\\u003e(Table\\u0026nbsp;2)\\u003c/strong\\u003e. Across all participants, T2D prevalence was 22.1% with mean age of onset of 46.6 years. Among individuals in the top decile of the Lipodystrophy 1 distribution, prevalence of T2D was 26.3% (mean age of onset 46.6 years)\\u003cstrong\\u003e(\\u003c/strong\\u003eFig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003ea\\u003cstrong\\u003e)\\u003c/strong\\u003e. Among individuals in the top 10% of the Beta Cell 2 distribution T2D prevalence was 29.6% and mean age of onset 43.6 years. Compared to individuals in the bottom 10% of the Beta Cell 2 pPS distribution, these individuals were more likely to develop nephropathy (HR 1.58, 95% CI 1.19\\u0026ndash;2.06, p\\u0026thinsp;=\\u0026thinsp;0.001)\\u003cstrong\\u003e(\\u003c/strong\\u003eFig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eC, \\u003cstrong\\u003eFig \\u003cspan class=\\\"InternalRef\\\"\\u003eS10\\u003c/span\\u003e, Table \\u003cspan class=\\\"InternalRef\\\"\\u003eS6\\u003c/span\\u003e-7)\\u003c/strong\\u003e. The obesity pPS, in contrast, was not associated with progression to complications.\\u003c/p\\u003e\\n\\u003cp\\u003eIndividuals at combined high genetic risk for Lipodystrophy 1 and Beta Cell 2 (in the top decile of both distributions, n\\u0026thinsp;=\\u0026thinsp;110) were on average diagnosed with diabetes 8 years earlier in life than those in the bottom decile of both distributions (n\\u0026thinsp;=\\u0026thinsp;304), had 3kg/m\\u003csup\\u003e2\\u003c/sup\\u003e lower BMI at diagnosis (Fig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eA\\u003cstrong\\u003e/B)\\u003c/strong\\u003e, 7.8% higher lifetime prevalence of diabetic retinopathy, 12% higher prevalence of insulin dependence, and 3.73 mmol/mol higher 5 year HbA1c in spite of similar baseline HbA1c \\u003cstrong\\u003e(Table\\u0026nbsp;2\\u003c/strong\\u003e). In survival models, we found they were more likely to progress to diabetic neuropathy (HR 2.75, 95% CI 1.10\\u0026ndash;6.9, p\\u0026thinsp;=\\u0026thinsp;0.031)(Fig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eC, \\u003cstrong\\u003eFig \\u003cspan class=\\\"InternalRef\\\"\\u003eS10\\u003c/span\\u003e)\\u003c/strong\\u003e.\\u003c/p\\u003e\\n\\u003c/div\\u003e\"},{\"header\":\"Discussion\",\"content\":\"\\u003cdiv id=\\\"Sec30\\\" class=\\\"Section2\\\"\\u003e\\n\\u003ch2\\u003eSummary of findings\\u003c/h2\\u003e\\n\\u003cp\\u003eIn this study of 51,170 British Pakistani and Bangladeshi individuals, we show that multi-ancestry T2D partitioned polygenic scores (pPS) developed in White European, East Asian and African ancestry individuals are applicable to British South Asian individuals, where they are predictive of both type 2 diabetes and gestational diabetes mellitus. We show the genetic architecture of T2D and GDM, mapped through pPS, bear striking similarities. Further, we identify beta-cell 2 and lipodystrophy 1 as being key drivers for type 2 diabetes diagnosis at a younger age with a lower BMI, and more rapid progression to microvascular complications. Based on associations used for endotype development, the beta cell 2 pPS was driven by lower HOMA- B and high proinsulin levels suggesting an insulin secretion defect. Lipodystrophy 1 was associated with lower gluteofemoral adipose tissue and adiponectin levels which indicate greater risk of unfavourable fat distribution-mediated insulin resistance. Using UK Biobank we show how the genetic burden of these pPS are greater in people of south Asian ancestry compared to white Europeans. We also show how higher genetic risk of the beta cell 2 pPS is associated with worse response to commonly-prescribed and first-line oral antidiabetic medications including metformin, SGLT2 inhibitors, and thiazolidinediones. A global T2D polygenic score does not show similar associations. Partitioned polygenic scores can help identify individuals with combined extreme genetic risk in specific pathways (lipodystrophy 1 and beta-cell 2) who have 8 year earlier onset of T2D, at 3 kg/m\\u003csup\\u003e2\\u003c/sup\\u003e lower BMI, and rapid progression to insulin use and nephropathy, with poor response to oral antidiabetic medications. We show that pPS can be used to identify individuals at high, clinically actionable risk and highlight the specific pathophysiological pathways underlying T2D risk in south Asians. These results recapitulate findings from the largest recent genetic study of T2D and are supportive of the use of genetic information, specific partitioned polygenic scores, in guiding screening and therapeutic management for type 2 diabetes, especially in south Asians.\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec31\\\" class=\\\"Section2\\\"\\u003e\\n\\u003ch2\\u003eType 2 diabetes pPS and PRS distributions differ between British Pakistani and Bangladeshi individuals\\u003c/h2\\u003e\\n\\u003cp\\u003eUsing UK Biobank, we observed higher genetic predisposition to beta cell dysfunction and unfavourable fat distribution patterns, but lower genetic predisposition to obesity, among individuals of South Asian compared to European ancestry \\u003cstrong\\u003e(Fig \\u003cspan class=\\\"InternalRef\\\"\\u003eS5\\u003c/span\\u003e).\\u003c/strong\\u003e In Genes \\u0026amp; Health, using stratified analyses, we identified higher unmodified (not corrected for principal components) genetic risk of all pPS in Bangladeshi compared to Pakistani individuals (\\u003cstrong\\u003eFig \\u003cspan class=\\\"InternalRef\\\"\\u003eS1\\u003c/span\\u003e)\\u003c/strong\\u003e. While we were under-powered to observe interactions between ancestry and sex we observed that Bangladeshi women showed the strongest associations between genetic risk and earlier onset T2D. These findings argue to an extent against the pooling of these distinct ancestral groups under the banner of \\u0026ldquo;South Asian / SAS\\u0026rdquo; in genetic epidemiological studies when eventually sample size is no longer a limitation \\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e30\\u003c/span\\u003e\\u003c/sup\\u003e. These findings also highlight important sex-stratified effects that can explain higher observed risk of T2D and GDM in women of certain ancestries.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cem\\u003eGenetic predictors of beta cell dysfunction, obesity, and lipodystrophy drive early-onset diabetes in Genes \\u0026amp; Health\\u003c/em\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eWe identify genetic propensity to impaired insulin secretion as a key driver of the genetic basis of age at diagnosis of type 2 diabetes in British Pakistanis and Bangladeshis (Fig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003e, Fig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eA\\u003cstrong\\u003e)\\u003c/strong\\u003e. Epidemiological studies have shown that south Asians are more likely to present with lower HOMA-B (an estimate of beta cell function), BMI, and greater dyslipidemia at the time of diagnosis referred to as the Severe Insulin Deficient Diabetes (SIDD) endotype, as compared to white Europeans \\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e31\\u003c/span\\u003e\\u003c/sup\\u003e. The greater genetic burden of insulin secretion and early beta-cell exhaustion is a likely reason. The observed effect of greater genetic burden of impaired insulin secretion recapitulates findings from Asian Indians, where a previously defined beta-cell dysfunction pPS was found to be associated with earlier onset T2D \\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e4\\u003c/span\\u003e\\u003c/sup\\u003e. We replicate this using a newer multi-ancestry pPS and extend the finding to highlight the role of specific aetiologies in GDM, treatment response, progression to insulin dependence and nephropathy.\\u003c/p\\u003e\\n\\u003cp\\u003eNumerous studies have reported the clinical phenotype of low BMI but larger waist circumferences in south Asian populations \\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e32\\u003c/span\\u003e\\u003c/sup\\u003e. For example, SABRE, a 20-year-long prospective cohort study comparing white Europeans, south Asians, and Afro-Caribbean groups living in the UK found that, compared to south Asian men developed T2D with lower BMI, higher waist-hip ratio, higher truncal skin-folds, higher insulin resistance, and increased (compensatory) insulin secretion\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e33\\u003c/span\\u003e\\u003c/sup\\u003e. Therefore, our observation that the aetiological pathways of unfavourable fat distribution plays a key role in early onset and rapid progression of T2D in South Asians is well-supported.\\u003c/p\\u003e\\n\\u003cp\\u003eThere are limited studies that explore the underlying architecture of age of diagnosis and by proxy, earlier onset of T2D\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e20\\u003c/span\\u003e,\\u003cspan class=\\\"CitationRef\\\"\\u003e34\\u003c/span\\u003e\\u003c/sup\\u003e. However, efforts to do so, agree with our findings that drivers of earlier age of diabetes onset do not overlap entirely with overall drivers of T2d risk. Our use of pPS aid the identification of specific aetiologies driving earlier onset in South Asians.\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec32\\\" class=\\\"Section2\\\"\\u003e\\n\\u003ch2\\u003eA limited role for genetically-determined BMI-mediated obesity\\u003c/h2\\u003e\\n\\u003cp\\u003eWe observed an effect of BMI-mediated obesity pPS in driving early-onset type 2 diabetes, however, the pPS was not associated with greater risk of insulin dependence or diabetes-related complications. This effect is mirrored in epidemiological studies that have shown south Asians who develop T2D at younger ages have increased weight over those who do not, but that this weight gain is relatively lesser than observed in white Europeans or Black individuals \\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e\\u003c/sup\\u003e. This was also observed in a national survey of Asian Indians, where 45% of young diagnosed (\\u0026lt;\\u0026thinsp;40 years) T2D was in individuals with obese (45%) or overweight (15%) BMI. This is the first study using genetics that uncovers the role of these two adiposity-related aetiologies (unfavourable fat distribution) and BMI in south Asian type 2 diabetes.\\u003c/p\\u003e\\n\\u003cdiv id=\\\"Sec33\\\" class=\\\"Section3\\\"\\u003e\\n\\u003ch2\\u003eGenetic architecture of GDM mirrors that of Type 2 Diabetes in British Pakistanis and Bangladeshis\\u003c/h2\\u003e\\n\\u003cp\\u003eIn comparison to a rich literature for type 2 diabetes, the genetics of gestational diabetes mellitus is less well understood. In this study, we show that the genetic architecture of GDM, assessed by association with genetically determined T2D endotypes, closely resembles that of the genetic basis of Type 2 Diabetes itself in British Bangladeshis and Pakistanis \\u003cstrong\\u003e(\\u003c/strong\\u003eFig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003eA, \\u003cstrong\\u003eFig \\u003cspan class=\\\"InternalRef\\\"\\u003eS5\\u003c/span\\u003e)\\u003c/strong\\u003e, in keeping with similar findings recently reported in European ancestry individuals \\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e35\\u003c/span\\u003e\\u003c/sup\\u003e. In our study, the strongest associations with GDM were observed for Beta Cell 1 and Beta Cell 2, suggesting genetic predisposition to insulin deficiency in British South Asian women may contribute to GDM risk as well as T2D risk. The association of GDM with the Obesity pPS in our study, in contrast, was relatively weaker, despite this being highlighted as a major aetiological pathway in White European\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e35\\u003c/span\\u003e\\u003c/sup\\u003e and Turkish\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e36\\u003c/span\\u003e\\u003c/sup\\u003e mothers with GDM.\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003cdiv id=\\\"Sec34\\\" class=\\\"Section3\\\"\\u003e\\n\\u003ch2\\u003eGenetic testing allows prediction of clinically relevant outcomes\\u003c/h2\\u003e\\n\\u003cp\\u003eWe show application of pPS to real-world linked electronic health records allows identification of individuals at extremes of genetic risk who are at particularly high risk of developing type 2 diabetes in early adulthood (Fig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eA\\u003cstrong\\u003e)\\u003c/strong\\u003e, responding poorly to widely-used oral antidiabetic drugs (Fig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003eA\\u003cstrong\\u003e)\\u003c/strong\\u003e, and of progressing rapidly to insulin requirement \\u003cstrong\\u003e(\\u003c/strong\\u003eFig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003eB\\u003cstrong\\u003e)\\u003c/strong\\u003e and complications \\u003cstrong\\u003e(\\u003c/strong\\u003eFig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eC\\u003cstrong\\u003e).\\u003c/strong\\u003e However, no pPS or combination of pPSs was as predictive as previously reported hazard ratios for phenotypically-derived clusters related to insulin deficiency in the IMI-RHAPSODY study.\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e37\\u003c/span\\u003e\\u003c/sup\\u003e This may reflect the fact that the 12 endotypes included in this analysis were derived in a multi-ancestral study which did not train cluster allocation on insulin-secretion related traits\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e19\\u003c/span\\u003e\\u003c/sup\\u003e. Unlike phenotypic clustering efforts, genetic data can predict risk much earlier in the life course, rather than proximate to the time of disease onset.\\u003c/p\\u003e\\n\\u003cp\\u003eAs the use of genetic testing in the real-world clinical setting increases\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e38\\u003c/span\\u003e,\\u003cspan class=\\\"CitationRef\\\"\\u003e39\\u003c/span\\u003e\\u003c/sup\\u003e, genetic risk scores may offer means to stratify individuals into those at high risk of requiring specialist input (for example, those rapidly progressing to insulin requirement, who are more likely to require specialist secondary care management)\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e40\\u003c/span\\u003e\\u003c/sup\\u003e, or who are less likely to respond to particular classes of medication. Overall, most individuals responded to the introduction of glucose lowering medication with a reduction in HbA1c (\\u003cstrong\\u003eFig \\u003cspan class=\\\"InternalRef\\\"\\u003eS8\\u003c/span\\u003eA).\\u003c/strong\\u003e However, high genetic risk of insulin deficiency determined by high beta Cell 2 pPS was associated with increased HbA1c after initiation of metformin, SGLT2 inhibitors, and thiazolidinediones (Fig.\\u0026nbsp;\\u003cspan class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003eA\\u003cstrong\\u003e)\\u003c/strong\\u003e.These findings are biologically plausible considering, for example, the mode of action of thiazolidinediones as insulin sensitizers\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e41\\u003c/span\\u003e\\u003c/sup\\u003e, which may offer little benefit if insulin deficiency is underpinning an individual\\u0026rsquo;s hyperglycaemia. While mechanistically congruent, these treatment-response findings are novel and possible only due to linkage with primary, secondary care and prescribing records in G\\u0026amp;H, and have not been shown in other pPS efforts. In contrast, we did not observe differential treatment response using a type 2 diabetes polygenic score, highlighting the value of pPS in dissecting clinically relevant pathophysiologies.\\u003c/p\\u003e\\n\\u003cp\\u003eOur study underscores the utility of pPS and extends their utility to complications such as insulin dependence and treatment response and highlights that as pPS become more robust and ancestrally diverse, their utility in guiding diabetes prevention and management strategies should increase \\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e15\\u003c/span\\u003e,\\u003cspan class=\\\"CitationRef\\\"\\u003e19\\u003c/span\\u003e\\u003c/sup\\u003e. pPS allow us to identify greater genetic burden of certain pathophysiological pathways of T2D in south Asians and the use of this under-represented ancestry allows us to uncover novel associations with beta-cell deficiency which would not be apparent in European cohorts.\\u003c/p\\u003e\\n\\u003c/div\\u003e\\n\\u003c/div\\u003e\\n\\u003ch2\\u003eStrengths and limitations\\u003c/h2\\u003e\\n\\u003cp\\u003eStrengths of this study include its exploration of an under-represented population with high burden of cardiometabolic disease, and linkage of real-world electronic health record and prescribing data, which provides a platform for real-world application and translation of clinically relevant findings, in addition to internal validation of novel findings around pharmacogenetic applications of pPS. We also demonstrate the utility of pPS derived by Smith et al\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e42\\u003c/span\\u003e\\u003c/sup\\u003e in a population not included in their pPS derivation. Weaknesses include the lack of external validation of results, which is limited in part by the paucity of non-European studies combining genetic data with health record data\\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e\\u003c/sup\\u003e; in fact, some of the results shown such as response to medication have not even been shown in European cohorts due to rarity of required clinical and prescribing data. Due to attrition of sample size for complications these analyses are quite under-powered, however where this is the case, we have used an internal replication cohort to validate results.. Another limitation is the fact that this is not an ancestry-specific pPS, which suggests we might not be using the most optimal causal variants. However, this is likely to cause an under-estimation of true genetic effects \\u003csup\\u003e\\u003cspan class=\\\"CitationRef\\\"\\u003e23\\u003c/span\\u003e,\\u003cspan class=\\\"CitationRef\\\"\\u003e43\\u003c/span\\u003e\\u003c/sup\\u003e. Finally, in common with all studies using real-world electronic health record data, there is a risk of misclassification of diabetes, miscoding, and sampling bias towards individuals with chronic disease.\\u003c/p\\u003e\"},{\"header\":\"Declarations\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eAcknowledgment/Funding Statement\\u003c/strong\\u003e\\u003cstrong\\u003e:\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eSH is funded by a Wellcome HARP Doctoral Fellowship 227532/Z/23/Z. RM and MKS are funded by Barts Charity (MGU0504).\\u003c/p\\u003e\\n\\u003cp\\u003eDS is funded by the Tackling Multimorbidity at Scale Strategic Priorities Fund programme [grant number MR/W014416/1] delivered by the Medical Research Council and the National Institute for Health Research in partnership with the Economic and Social Research Council and in collaboration with the Engineering and Physical Sciences Research Council.\\u003c/p\\u003e\\n\\u003cp\\u003eGenes \\u0026amp; Health is/has recently been core-funded by Wellcome (WT102627, WT210561), the Medical Research Council (UK) (M009017, MR/X009777/1, MR/X009920/1), Higher Education Funding Council for England Catalyst, Barts Charity (845/1796), Health Data Research UK (for London substantive site), and research delivery support from the NHS National Institute for Health Research Clinical Research Network (North Thames). Genes \\u0026amp; Health is/has recently been funded by Alnylam Pharmaceuticals, Genomics PLC; and a Life Sciences Industry Consortium of Astra Zeneca PLC, Bristol-Myers Squibb Company, GlaxoSmithKline Research and Development Limited, Maze Therapeutics Inc, Merck Sharp \\u0026amp; Dohme LLC, Novo Nordisk A/S, Pfizer Inc, Takeda Development Centre Americas Inc.\\u003c/p\\u003e\\n\\u003cp\\u003eWe thank Social Action for Health, Centre of The Cell, members of our Community Advisory Group, and staff who have recruited and collected data from volunteers. We thank the NIHR National Biosample Centre (UK Biocentre), the Social Genetic \\u0026amp; Developmental Psychiatry Centre (King\\u0026apos;s College London), Wellcome Sanger Institute, and Broad Institute for sample processing, genotyping, sequencing and variant annotation.\\u003c/p\\u003e\\n\\u003cp\\u003eWe thank: Barts Health NHS Trust, NHS Clinical Commissioning Groups (City and Hackney, Waltham Forest, Tower Hamlets, Newham, Redbridge, Havering, Barking and Dagenham), East London NHS Foundation Trust, Bradford Teaching Hospitals NHS Foundation Trust, Public Health England (especially David Wyllie), Discovery Data Service/Endeavour Health Charitable Trust (especially David Stables), Voror Health Technologies Ltd (especially Sophie Don), NHS England (for what was NHS Digital) - for GDPR-compliant data sharing backed by individual written informed consent.\\u003c/p\\u003e\\n\\u003cp\\u003eMost of all we thank all of the volunteers participating in Genes \\u0026amp; Health.\\u003c/p\\u003e\\n\\u003cp\\u003eCurrent Genes \\u0026amp; Health Research Team (in alphabetical order by surname): Shaheen Akhtar, Mohammad Anwar, Omar Asgar, Samina Ashraf, Saeed Bidi, Gerome Breen, James Broster, Raymond Chung, David Collier, Charles J Curtis, Shabana Chaudhary, Grainne Colligan, Panos Deloukas, Ceri Durham, Faiza Durrani, Fabiola Eto, Sarah Finer, Joseph Gafton, Ana Angel, Chris Griffiths, Joanne Harvey, Teng Heng, Sam Hodgson, Qin Qin Huang, Matt Hurles, Karen A Hunt, Shapna Hussain, Kamrul Islam, Vivek Iyer, Benjamin M Jacobs, Georgios Kalantzis, Ahsan Khan, Claudia Langenberg, Cath Lavery, Sang Hyuck Lee, Daniel MacArthur, Sidra Malik, Daniel Malawsky, Hilary Martin, Dan Mason, Rohini Mathur, Mohammed Bodrul Mazid, John McDermott, Caroline Morton, Bill Newman, Elizabeth Owor, Asma Qureshi, Shwetha Ramachandrappa, Mehru Raza, Jessry Russell, Nishat Safa, Miriam Samuel, Moneeza Siddiqui, Michael Simpson, John Solly, Marie Spreckley. Daniel Stow, Michael Taylor, Richard C Trembath, Karen Tricker, David A van Heel, Klaudia Walter, Caroline Winckley, Suzanne Wood, John Wright, Ishevanhu Zengeya, Julia Z\\u0026ouml;llner.\\u003c/p\\u003e\"},{\"header\":\"References\",\"content\":\"\\u003col\\u003e\\n\\u003cli\\u003eMagliano, D. \\u0026amp; Boyko, E. J. \\u003cem\\u003eIDF Diabetes Atlas\\u003c/em\\u003e. (International Diabetes Federation, 2021).\\u003c/li\\u003e\\n\\u003cli\\u003eGujral, U. P., Pradeepa, R., Weber, M. B., Narayan, K. M. V. \\u0026amp; Mohan, V. Type 2 diabetes in South Asians: similarities and differences with white Caucasian and other populations. \\u003cem\\u003eAnn. N. Y. Acad. Sci.\\u003c/em\\u003e \\u003cstrong\\u003e1281\\u003c/strong\\u003e, 51\\u0026ndash;63 (2013).\\u003c/li\\u003e\\n\\u003cli\\u003eWright, A. K. \\u003cem\\u003eet al.\\u003c/em\\u003e Age-, sex- and ethnicity-related differences in body weight, blood pressure, HbA1c and lipid levels at the diagnosis of type 2 diabetes relative to people without diabetes. \\u003cem\\u003eDiabetologia\\u003c/em\\u003e \\u003cstrong\\u003e63\\u003c/strong\\u003e, 1542\\u0026ndash;1553 (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eSiddiqui, M. K. \\u003cem\\u003eet al.\\u003c/em\\u003e Young-onset diabetes in Asian Indians is associated with lower measured and genetically determined beta cell function. \\u003cem\\u003eDiabetologia\\u003c/em\\u003e \\u003cstrong\\u003e65\\u003c/strong\\u003e, 973\\u0026ndash;983 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eFarrar, D. \\u003cem\\u003eet al.\\u003c/em\\u003e Association between hyperglycaemia and adverse perinatal outcomes in south Asian and white British women: analysis of data from the Born in Bradford cohort. \\u003cem\\u003eLancet Diabetes Endocrinol\\u003c/em\\u003e \\u003cstrong\\u003e3\\u003c/strong\\u003e, 795\\u0026ndash;804 (2015).\\u003c/li\\u003e\\n\\u003cli\\u003eSirugo, G., Williams, S. M. \\u0026amp; Tishkoff, S. A. The Missing Diversity in Human Genetic Studies. \\u003cem\\u003eCell\\u003c/em\\u003e \\u003cstrong\\u003e177\\u003c/strong\\u003e, 26\\u0026ndash;31 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eFatumo, S. \\u003cem\\u003eet al.\\u003c/em\\u003e A roadmap to increase diversity in genomic studies. \\u003cem\\u003eNat. Med.\\u003c/em\\u003e \\u003cstrong\\u003e28\\u003c/strong\\u003e, 243\\u0026ndash;250 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eDeutsch, A. J., Ahlqvist, E. \\u0026amp; Udler, M. S. Phenotypic and genetic classification of diabetes. \\u003cem\\u003eDiabetologia\\u003c/em\\u003e \\u003cstrong\\u003e65\\u003c/strong\\u003e, 1758\\u0026ndash;1769 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eAnjana, R. M. \\u003cem\\u003eet al.\\u003c/em\\u003e Novel subgroups of type 2 diabetes and their association with microvascular outcomes in an Asian Indian population: a data-driven cluster analysis: the INSPIRED study. \\u003cem\\u003eBMJ Open Diabetes Res Care\\u003c/em\\u003e \\u003cstrong\\u003e8\\u003c/strong\\u003e, (2020).\\u003c/li\\u003e\\n\\u003cli\\u003ePrasad, R. B. \\u003cem\\u003eet al.\\u003c/em\\u003e Correction to: Subgroups of patients with young-onset type 2 diabetes in India reveal insulin deficiency as a major driver. \\u003cem\\u003eDiabetologia\\u003c/em\\u003e \\u003cstrong\\u003e65\\u003c/strong\\u003e, 254 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eMansour Aly, D. \\u003cem\\u003eet al.\\u003c/em\\u003e Genome-wide association analyses highlight etiological differences underlying newly defined subtypes of diabetes. \\u003cem\\u003eNat. Genet.\\u003c/em\\u003e \\u003cstrong\\u003e53\\u003c/strong\\u003e, 1534\\u0026ndash;1542 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eDennis, J. M., Shields, B. M., Henley, W. E., Jones, A. G. \\u0026amp; Hattersley, A. T. Disease progression and treatment response in data-driven subgroups of type 2 diabetes compared with models based on simple clinical features: an analysis using clinical trial data. \\u003cem\\u003eLancet Diabetes Endocrinol\\u003c/em\\u003e \\u003cstrong\\u003e7\\u003c/strong\\u003e, 442\\u0026ndash;451 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eAhlqvist, E. \\u003cem\\u003eet al.\\u003c/em\\u003e Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables. \\u003cem\\u003eLancet Diabetes Endocrinol\\u003c/em\\u003e \\u003cstrong\\u003e6\\u003c/strong\\u003e, 361\\u0026ndash;369 (2018).\\u003c/li\\u003e\\n\\u003cli\\u003eUdler, M. S. \\u003cem\\u003eet al.\\u003c/em\\u003e Type 2 diabetes genetic loci informed by multi-trait associations point to disease mechanisms and subtypes: A soft clustering analysis. \\u003cem\\u003ePLoS Med.\\u003c/em\\u003e \\u003cstrong\\u003e15\\u003c/strong\\u003e, e1002654 (2018).\\u003c/li\\u003e\\n\\u003cli\\u003eSuzuki, K. \\u003cem\\u003eet al.\\u003c/em\\u003e Genetic drivers of heterogeneity in type 2 diabetes pathophysiology. \\u003cem\\u003eNature\\u003c/em\\u003e (2024) doi:10.1038/s41586-024-07019-6.\\u003c/li\\u003e\\n\\u003cli\\u003eUdler, M. S., McCarthy, M. I., Florez, J. C. \\u0026amp; Mahajan, A. Genetic Risk Scores for Diabetes Diagnosis and Precision Medicine. \\u003cem\\u003eEndocr. Rev.\\u003c/em\\u003e \\u003cstrong\\u003e40\\u003c/strong\\u003e, 1500\\u0026ndash;1520 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eDiCorpo, D. \\u003cem\\u003eet al.\\u003c/em\\u003e Type 2 Diabetes Partitioned Polygenic Scores Associate With Disease Outcomes in 454,193 Individuals Across 13 Cohorts. \\u003cem\\u003eDiabetes Care\\u003c/em\\u003e \\u003cstrong\\u003e45\\u003c/strong\\u003e, 674\\u0026ndash;683 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eKim, H. \\u003cem\\u003eet al.\\u003c/em\\u003e High-throughput genetic clustering of type 2 diabetes loci reveals heterogeneous mechanistic pathways of metabolic disease. \\u003cem\\u003eDiabetologia\\u003c/em\\u003e \\u003cstrong\\u003e66\\u003c/strong\\u003e, 495\\u0026ndash;507 (2023).\\u003c/li\\u003e\\n\\u003cli\\u003eSmith, K. \\u003cem\\u003eet al.\\u003c/em\\u003e Multi-ancestry polygenic mechanisms of type 2 diabetes. \\u003cem\\u003eNat. Med.\\u003c/em\\u003e (2024) doi:10.1038/s41591-024-02865-3.\\u003c/li\\u003e\\n\\u003cli\\u003eSrinivasan, S. \\u003cem\\u003eet al.\\u003c/em\\u003e Common and Distinct Genetic Architecture of Age at Diagnosis of Diabetes in South Indian and European Populations. \\u003cem\\u003eDiabetes Care\\u003c/em\\u003e \\u003cstrong\\u003e46\\u003c/strong\\u003e, 1515\\u0026ndash;1523 (2023).\\u003c/li\\u003e\\n\\u003cli\\u003eFiner, S. \\u003cem\\u003eet al.\\u003c/em\\u003e Cohort Profile: East London Genes \\u0026amp; Health (ELGH), a community-based population genomics and health study in British Bangladeshi and British Pakistani people. \\u003cem\\u003eInt. J. Epidemiol.\\u003c/em\\u003e \\u003cstrong\\u003e49\\u003c/strong\\u003e, 20\\u0026ndash;21i (2020).\\u003c/li\\u003e\\n\\u003cli\\u003eHuang, Q. Q. \\u003cem\\u003eet al.\\u003c/em\\u003e Transferability of genetic loci and polygenic scores for cardiometabolic traits in British Pakistanis and Bangladeshis. \\u003cem\\u003ebioRxiv\\u003c/em\\u003e (2021) doi:10.1101/2021.06.22.21259323.\\u003c/li\\u003e\\n\\u003cli\\u003eHodgson, S. \\u003cem\\u003eet al.\\u003c/em\\u003e Integrating polygenic risk scores in the prediction of type 2 diabetes risk and subtypes in British Pakistanis and Bangladeshis: A population-based cohort study. \\u003cem\\u003ePLoS Med.\\u003c/em\\u003e \\u003cstrong\\u003e19\\u003c/strong\\u003e, e1003981 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eMars, N. \\u003cem\\u003eet al.\\u003c/em\\u003e Systematic comparison of family history and polygenic risk across 24 common diseases. \\u003cem\\u003eAm. J. Hum. Genet.\\u003c/em\\u003e \\u003cstrong\\u003e109\\u003c/strong\\u003e, 2152\\u0026ndash;2162 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eLiu, T. \\u003cem\\u003eet al.\\u003c/em\\u003e Investigating misclassification of type 1 diabetes in a population-based cohort of British Pakistanis and Bangladeshis using polygenic risk scores. \\u003cem\\u003ebioRxiv\\u003c/em\\u003e (2023) doi:10.1101/2023.08.23.23294497.\\u003c/li\\u003e\\n\\u003cli\\u003eCreators Fabiola Eto1 Miriam Samuel finersarah Show affiliations 1. Queen Mary University of London. \\u003cem\\u003eF-eto/MULTIPLY-Initiative: Version 1.1\\u003c/em\\u003e. doi:10.5281/zenodo.7643566.\\u003c/li\\u003e\\n\\u003cli\\u003eZhou, K. \\u003cem\\u003eet al.\\u003c/em\\u003e Variation in the glucose transporter gene SLC2A2 is associated with glycemic response to metformin. \\u003cem\\u003eNat. Genet.\\u003c/em\\u003e \\u003cstrong\\u003e48\\u003c/strong\\u003e, 1055\\u0026ndash;1059 (2016).\\u003c/li\\u003e\\n\\u003cli\\u003ePurcell, S. \\u003cem\\u003eet al.\\u003c/em\\u003e PLINK: a tool set for whole-genome association and population-based linkage analyses. \\u003cem\\u003eAm. J. Hum. Genet.\\u003c/em\\u003e \\u003cstrong\\u003e81\\u003c/strong\\u003e, 559\\u0026ndash;575 (2007).\\u003c/li\\u003e\\n\\u003cli\\u003eLittle, J. \\u003cem\\u003eet al.\\u003c/em\\u003e STrengthening the REporting of Genetic Association Studies (STREGA)--an extension of the STROBE statement. \\u003cem\\u003eGenet. Epidemiol.\\u003c/em\\u003e \\u003cstrong\\u003e33\\u003c/strong\\u003e, 581\\u0026ndash;598 (2009).\\u003c/li\\u003e\\n\\u003cli\\u003eChambers, J. C. \\u003cem\\u003eet al.\\u003c/em\\u003e The South Asian genome. \\u003cem\\u003ePLoS One\\u003c/em\\u003e \\u003cstrong\\u003e9\\u003c/strong\\u003e, e102645 (2014).\\u003c/li\\u003e\\n\\u003cli\\u003eKe, C., Narayan, K. M. V., Chan, J. C. N., Jha, P. \\u0026amp; Shah, B. R. Pathophysiology, phenotypes and management of type 2 diabetes mellitus in Indian and Chinese populations. \\u003cem\\u003eNat. Rev. Endocrinol.\\u003c/em\\u003e \\u003cstrong\\u003e18\\u003c/strong\\u003e, 413\\u0026ndash;432 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eTillin, T. \\u003cem\\u003eet al.\\u003c/em\\u003e Insulin resistance and truncal obesity as important determinants of the greater incidence of diabetes in Indian Asians and African Caribbeans compared with Europeans: The Southall and Brent Revisited (SABRE) cohort. \\u003cem\\u003eDiabetes Care\\u003c/em\\u003e \\u003cstrong\\u003e36\\u003c/strong\\u003e, 383\\u0026ndash;393 (2013).\\u003c/li\\u003e\\n\\u003cli\\u003eTillin, T., Hughes, A. D., Godsland, I. F. \\u0026amp; Whincup, P. Insulin resistance and truncal obesity as important determinants of the greater incidence of diabetes in Indian Asians and African Caribbeans compared with \\u0026hellip;. \\u003cem\\u003eDiabetes\\u003c/em\\u003e (2013).\\u003c/li\\u003e\\n\\u003cli\\u003eKwak, S. H. \\u003cem\\u003eet al.\\u003c/em\\u003e Genetic architecture and biology of youth-onset type 2 diabetes. \\u003cem\\u003eNat Metab\\u003c/em\\u003e \\u003cstrong\\u003e6\\u003c/strong\\u003e, 226\\u0026ndash;237 (2024).\\u003c/li\\u003e\\n\\u003cli\\u003eElliott, A. \\u003cem\\u003eet al.\\u003c/em\\u003e Distinct and shared genetic architectures of Gestational diabetes mellitus and Type 2 Diabetes Mellitus. \\u003cem\\u003emedRxiv\\u003c/em\\u003e (2023) doi:10.1101/2023.02.16.23286014.\\u003c/li\\u003e\\n\\u003cli\\u003eBeysel, S. \\u003cem\\u003eet al.\\u003c/em\\u003e Maternal genetic contribution to pre-pregnancy obesity, gestational weight gain, and gestational diabetes mellitus. \\u003cem\\u003eDiabetol. Metab. Syndr.\\u003c/em\\u003e \\u003cstrong\\u003e11\\u003c/strong\\u003e, 37 (2019).\\u003c/li\\u003e\\n\\u003cli\\u003eSlieker, R. C. \\u003cem\\u003eet al.\\u003c/em\\u003e Replication and cross-validation of type 2 diabetes subtypes based on clinical variables: an IMI-RHAPSODY study. \\u003cem\\u003eDiabetologia\\u003c/em\\u003e \\u003cstrong\\u003e64\\u003c/strong\\u003e, 1982\\u0026ndash;1989 (2021).\\u003c/li\\u003e\\n\\u003cli\\u003eFranceschini, N., Frick, A. \\u0026amp; Kopp, J. B. Genetic Testing in Clinical Settings. \\u003cem\\u003eAm. J. Kidney Dis.\\u003c/em\\u003e \\u003cstrong\\u003e72\\u003c/strong\\u003e, 569\\u0026ndash;581 (2018).\\u003c/li\\u003e\\n\\u003cli\\u003eYu, O. H. Y. \\u0026amp; Shin, J.-Y. Treating type 2 diabetes: moving towards precision medicine. \\u003cem\\u003eThe Lancet. Digital health\\u003c/em\\u003e vol. 4 e851\\u0026ndash;e852 (2022).\\u003c/li\\u003e\\n\\u003cli\\u003eRecommendations | Type 2 diabetes in adults: management | Guidance | NICE.\\u003c/li\\u003e\\n\\u003cli\\u003eHauner, H. The mode of action of thiazolidinediones. \\u003cem\\u003eDiabetes. Metab. Res. Rev.\\u003c/em\\u003e \\u003cstrong\\u003e18 Suppl 2\\u003c/strong\\u003e, S10\\u0026ndash;5 (2002).\\u003c/li\\u003e\\n\\u003cli\\u003eUdler, M. \\u003cem\\u003eet al.\\u003c/em\\u003e Multi-ancestry Polygenic Mechanisms of Type 2 Diabetes Elucidate Disease Processes and Clinical Heterogeneity. \\u003cem\\u003eRes Sq\\u003c/em\\u003e (2023) doi:10.21203/rs.3.rs-3399145/v1.\\u003c/li\\u003e\\n\\u003cli\\u003eHuang, Q. Q. \\u003cem\\u003eet al.\\u003c/em\\u003e Transferability of genetic loci and polygenic scores for cardiometabolic traits in British Pakistani and Bangladeshi individuals. \\u003cem\\u003eNat. Commun.\\u003c/em\\u003e \\u003cstrong\\u003e13\\u003c/strong\\u003e, 4664 (2022).\\u003c/li\\u003e\\n\\u003c/ol\\u003e\"},{\"header\":\"Tables\",\"content\":\"\\u003cp\\u003eTables 1 and 2 are available in the Supplementary Files section.\\u003c/p\\u003e \"}],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":true,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":true,\"hideJournal\":false,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":true,\"isAuthorSuppliedPdf\":false,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":false,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"nature-portfolio\",\"isNatureJournal\":true,\"hasQc\":false,\"allowDirectSubmit\":false,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"\",\"title\":\"Nature Portfolio\",\"twitterHandle\":\"\",\"acdcEnabled\":false,\"dfaEnabled\":false,\"editorialSystem\":\"ejp\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":false},\"keywords\":\"\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-4027509/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-4027509/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003eSouth Asians experience a higher risk of early-onset Type 2 diabetes (T2D) with normal BMI. However, since genetic research is largely focussed on white Europeans, the reasons for this are poorly understood. We used 12 recently derived multi-ancestry partitioned polygenic risk scores (pPS) to identify the aetiological pathways underlying T2D, gestational diabetes mellitus (GDM), earlier onset, progression to complications and insulin dependence, and treatment response in a south Asian cohort. Using electronic health record and genetic data from 51,108 British Pakistani and Bangladeshi individuals with T2D (n\\u0026thinsp;=\\u0026thinsp;11,673) and GDM (n\\u0026thinsp;=\\u0026thinsp;1,965) in the Genes \\u0026amp; Health study, we explored associations between pPS, T2D, GDM, diabetes complications, and treatment response using sex- and ancestry-adjusted multivariable regression and Cox proportional-hazards models. A pPS representing insulin deficiency was most strongly associated with T2D per standard deviation (OR: odds ratio):1.46, 95%CI:1.42\\u0026ndash;1.50), GDM (OR:1.27, 95%CI: 1.20\\u0026ndash;1.34) and age at T2D diagnosis (beta = -1.7 years, 95%CI: -1.5 to -1.9), followed by a pPS representing an unfavourable fat distribution (lipodystrophy). Individuals at high genetic risk of both insulin deficiency and lipodystrophy were diagnosed with T2D 8.2 years earlier with BMI 3 kg/m\\u003csup\\u003e2\\u003c/sup\\u003e lower compared to those at low genetic risk. The insulin deficiency pPS was associated with poorer response to metformin, thiazolidinediones, and SGLT2 inhibitors (post-treatment HbA1c increased from baseline by 0.51%, 1.83%, and 1.13% respectively). Higher Insulin deficiency and lipodystrophy pPS were also associated with faster progression to insulin dependence and microvascular complications. Using UK Biobank, we found that south Asians had a greater genetic burden of both these pPS compared to white Europeans. In British Pakistani and Bangladeshi individuals, genetic predisposition to insulin deficiency and lipodystrophy helps identify individuals at risk of earlier onset of type 2 diabetes, who progress faster to complications and insulin dependence, and are less likely to respond to standard diabetes management pathways.\\u003c/p\\u003e\",\"manuscriptTitle\":\"Partitioned polygenic scores highlight role of beta-cell function and unfavourable fat distribution patterns in young onset type 2 diabetes in south Asians\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2024-03-11 18:27:30\",\"doi\":\"10.21203/rs.3.rs-4027509/v1\",\"editorialEvents\":[],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"nature-medicine\",\"isNatureJournal\":true,\"hasQc\":false,\"allowDirectSubmit\":false,\"externalIdentity\":\"nm\",\"sideBox\":\"Learn more about [Nature Medicine](http://www.nature.com/nm/)\",\"snPcode\":\"\",\"submissionUrl\":\"\",\"title\":\"Nature Medicine\",\"twitterHandle\":\"\",\"acdcEnabled\":true,\"dfaEnabled\":true,\"editorialSystem\":\"ejp\",\"reportingPortfolio\":\"Nature Research\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":false}}],\"origin\":\"\",\"ownerIdentity\":\"d3126577-d579-4fe6-a4ae-59309345c84b\",\"owner\":[],\"postedDate\":\"March 11th, 2024\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"published-in-journal\",\"subjectAreas\":[{\"id\":29199985,\"name\":\"Health sciences/Medical research/Genetics research\"},{\"id\":29199986,\"name\":\"Health sciences/Endocrinology/Endocrine system and metabolic diseases/Diabetes/Type 2 diabetes\"},{\"id\":29199987,\"name\":\"Health sciences/Pathogenesis/Clinical genetics/Disease genetics\"},{\"id\":29199988,\"name\":\"Health sciences/Diseases/Endocrine system and metabolic diseases/Diabetes/Type 2 diabetes\"},{\"id\":29199989,\"name\":\"Health sciences/Risk factors\"}],\"tags\":[],\"updatedAt\":\"2024-11-27T08:07:41+00:00\",\"versionOfRecord\":{\"articleIdentity\":\"rs-4027509\",\"link\":\"https://doi.org/10.1038/s41591-024-03317-8\",\"journal\":{\"identity\":\"nature-medicine\",\"isVorOnly\":false,\"title\":\"Nature Medicine\"},\"publishedOn\":\"2024-11-26 05:00:00\",\"publishedOnDateReadable\":\"November 26th, 2024\"},\"versionCreatedAt\":\"2024-03-11 18:27:30\",\"video\":\"\",\"vorDoi\":\"10.1038/s41591-024-03317-8\",\"vorDoiUrl\":\"https://doi.org/10.1038/s41591-024-03317-8\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-4027509\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-4027509\",\"identity\":\"rs-4027509\",\"version\":[\"v1\"]},\"buildId\":\"qtupq5eGEP_6zYnWcrvyt\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}