Current Diabetes Reports (2022) 22:227–235

https://doi.org/10.1007/s11892-022-01462-3

MACROVASCULAR COMPLICATIONS IN DIABETES (R SHAH, SECTION EDITORS)

Genetics of Type 2 Diabetes: Implications from Large‑Scale Studies

Natalie DeForest1,2 · Amit R. Majithia1

Accepted: 8 February 2022 / Published online: 19 March 2022

This is a U.S. government work and not under copyright protection in the U.S.; foreign copyright protection may apply 2022

Abstract
Purpose of Review Type 2 diabetes (T2D) is a multifactorial, heritable syndrome characterized by dysregulated glucose
homeostasis that results from impaired insulin secretion and insulin resistance. Genetic association studies have successfully
identified hundreds of T2D risk loci implicating many genes in disease pathogenesis. In this review, we provide an overview
of the recent T2D genetic studies from the past 3 years with particular focus on the effects of sample size and ancestral
diversity on genetic discovery as well as discuss recent work on the use and limitations of genetic risk scores (GRS) for
T2D risk prediction.
Recent Findings Recent large-scale, multi-ancestry genetic studies of T2D have identified over 500 novel risk loci. The
genetic variants (i.e., single nucleotide polymorphisms (SNPs)) marking these novel loci in general have smaller effect sizes
than previously discovered loci. Inclusion of samples from diverse ancestral backgrounds shows a few ancestry specific loci
marked by common variants, but overall, the majority of loci discovered are common across ancestries. Inclusion of com-
mon variant GRS, even with hundreds of loci, does not substantially increase T2D risk prediction over standard clinical risk
factors such as age and family history.
Summary Common variant association studies of T2D have now identified over 700 T2D risk loci, half of which have been
discovered in the past 3 years. These recent studies demonstrate that inclusion of ancestrally diverse samples can enhance
locus discovery and improve accuracy of GRS for T2D risk prediction. GRS based on common variants, however, only
minimally enhances risk prediction over standard clinical risk factors.

Keywords Type 2 diabetes · Human genetics · GWAS · Genetic risk score · Polygenic risk score · Multi-ancestry

Type 2 Diabetes Is a Heterogeneous months via glycated hemoglobin (HbA1c) tests. Clinical
but Heritable Syndrome presentation and disease progression may vary consider-
ably among individuals, and the prevalence of T2D varies
Type 2 diabetes (T2D) is characterized by impaired glu- between different ethnic groups; for example, Hispanic and
cose metabolism arising from defects in insulin resistance Black populations have higher age-adjusted T2D preva-
and secretion [1]. In clinical practice, T2D is diagnosed by lence compared to White and Asian groups [2, 3]. Clinical
elevated blood glucose levels most commonly assessed via complications of T2D include microvascular complications
point measurements in the fasting state or averaged over such as retinopathy, neuropathy, and nephropathy as well as
macrovascular complications such as myocardial infarction
and stroke [4]. Cardiovascular disease (CVD) is the leading
This article is part of the Topical Collection on Macrovascular cause of death in people with T2D who have up to a three-
Complications in Diabetes fold increase in CVD risk as compared to people without
T2D [5].
* Amit R. Majithia The pathogenesis of T2D involves both environmental
[email protected]
and genetic causes. Environmental factors including obesity,
1
Department of Medicine, Division of Endocrinology, stress, and lifestyle choices such as an unhealthy, energy-
University of California San Diego, La Jolla, CA, USA dense diet, and a sedentary lifestyle have been closely asso-
2
Biomedical Sciences Graduate Program, University ciated with the development of T2D [6]. The heritability of
of California San Diego, La Jolla, CA, USA T2D ranges from 30 to 70% [7] and family history of T2D

13
Vol.:(0123456789)
228 Current Diabetes Reports (2022) 22:227–235

is a significant risk factor, with an approximate two-fold effect sizes for T2D in African Americans compared with
relative risk for siblings [8] and a three-fold increased risk Europeans, but the majority of loci (92.1%) showed no sig-
for first-degree relatives of a T2D individual [9]. A handful nificant heterogeneity in effect estimates between Europeans
of robust disease genes were identified by early small-scale and African Americans.
genetic association studies for T2D [10, 11] and related The most recently published multi-ancestry T2D
Mendelian diabetes syndromes [12]. With the advent of case–control genetic study illustrates the dominant effect of
genotyping arrays and the systematic cataloging of com- sample size in driving locus discovery [20]. Polfus et al.
mon genetic variation by the International HapMap project, conducted a GWAS meta-analysis of 53,102 T2D cases and
population-scale genome-wide association studies (GWAS) 193,679 controls from the multi-ethnic Population Archi-
became feasible, leading to the identification of hundreds of tecture Genomics and Epidemiology (PAGE) consortium
T2D associated loci [13]. This review focuses on the T2D along with the DIAGRAM consortium, and replicated their
genetic association studies conducted over the past 3 years. findings in independent ancestry-specific samples from mul-
tiple T2D consortia including DIAMANTE, Asian Genetic
Epidemiology Network (AGEN), Slim Initiative in Genomic
T2D Risk Loci Marked by Common Genetic Medicine for the Americas (SIGMA), and African Ameri-
Variants Are Mostly Shared Across cans from the MEta-analysis of type 2 DIabetes in Afri-
Ancestries can Americans (MEDIA) [20]. They identified four novel
loci from the discovery PAGE + DIAGRAM GWAS, two
In the early 2000s, collaborative efforts spanning multiple of which replicated in single ancestry replication GWAS:
institutions across the globe coalesced into several interna- (1) rs11466334 near the transforming growth factor beta-1
tional consortia focused on genetic mapping of T2D along (TGFB1) gene and (2) rs13052926 near beta-secretase
with its related traits and even complications (summarized 2 (BACE2). Only the TGFB1 locus (rs11466334) was an
in Table 1). Initially, consortia such as DIAGRAM [14] and ancestry-specific variant occurring more commonly in Afri-
MAGIC [15] aggregated participants of a single ancestry can (minor allele frequency (MAF) = 6.8%) and Hispanic
(mostly northern European), but more recently, they have populations (MAF = 1.3%) as compared with other ances-
included participants across a variety of ancestries [16, 17•]. tries (MAF < 1%). The single nucleotide polymorphism
As of 2018, the list of T2D associations included over 200 (SNP) was also predicted to be functionally consequential
independent loci [18] (Table 2). Subsequent studies over the via disrupting a CCCTC-binding factor (CTCF) binding
past 3 years have built upon earlier work by meta-analyzing motif potentially leading to altered enhancer-promoter inter-
previously collected samples with samples obtained across actions. Although this study identified four novel loci, it did
multiple ancestries to identify an additional 500 T2D risk not re-identify over 90% of the genome-wide significant loci
loci, defined as those > 500 kb and linkage disequilibrium identified in previous studies [18, 19•] (Table 2). The critical
(LD) ­r2 < 0.05 from previously reported loci (Table 2). distinguishing factor was the sample size, highlighting this
The largest T2D genetic association study to date meta- as the major determinant of genetic discovery in common
analyzed GWAS from eight cohorts including population- variant association studies for T2D.
based biobanks such as the Million Veteran Program (MVP) To examine the effect of ancestry on loci associated with
and Biobank Japan as well as dedicated T2D case–control glycemic traits (fasting glucose, fasting insulin, 2-h glucose,
cohorts such as DIAMANTE [19•]. These cohorts contained and HbA1c) in non-diabetic individuals, Chen et al. and
individuals from five different ancestral groups (European, the MAGIC investigators first conducted meta-analyses of
African American, Hispanic, South Asian, and East Asian) GWAS within each of the following single-ancestry popula-
for a total of 228,499 T2D cases and 1,178,783 controls. A tions: European, African American, Hispanic, East Asian,
total of 568 T2D risk loci were identified at genome-wide or South Asian. They then meta-analyzed these “single-
significance, 293 of which were novel in this study [19•]. ancestry GWAS” in a “trans-ancestry” GWAS consisting of
These newly identified loci had smaller effect sizes (aver- a total of 281,416 non-diabetic individuals [17•]. From the
age beta regression coefficient of 0.032 ± 0.012 per allele) trans-ancestry GWAS, they identified 235 loci associated
than previously discovered T2D risk loci (average beta with at least one glycemic trait, and 7 additional loci from
of 0.054 ± 0.045 per allele), demonstrating that increased the single-ancestry GWAS that did not rise to genome-wide
sample size enhanced statistical power to detect association significance in the trans-ancestry analysis. Interestingly, the
signals with smaller biological effects. Additionally, within single-ancestry loci had similar allele frequencies across the
the MVP cohort, Vujkovic et al. performed ancestry-specific sampled ancestries, potentially suggesting epistatic effects
GWAS which identified an additional 21 loci in Europeans with other ancestry-specific variants or that they rose to sig-
and 4 loci in African Americans not initially identified in nificance in a particular single-ancestry analysis simply by
the original meta-analysis. A few loci demonstrated higher chance.

13
 Table 1  Overview of T2D-specific and disease agnostic large-scale consortia
Abbreviation Name Phenotype(s) Sequencing type Sample size Ancestry Citation Website

DIAGRAM DIAbetes Genetics T2D Genotype 26,676 T2D cases and European Scott et al. Diabetes https://​www.​diagr​am-​
Replication And 132,532 controls 2017 conso​rtium.​org/​index.​
Meta-analysis html
DIAMANTE DIAbetes Meta-ANal- T2D Genotype European: 74,124 T2D African–American, Mahajan et al. Nat. -
ysis of Trans-Ethnic cases and 824,006 East Asian, Euro- Gen 2018;
association studies controls pean, Hispanic, and Mahajan et al. medRxiv
South Asian 2020
MAGIC Meta-Analyses Glycemic traits (fast- Genotype 281,416 non-diabetic East Asian, Hispanic, Chen et al. Nat. Gen https://​magic​inves​tigat​
of Glucose and ing glucose, fasting individuals African American, 2021 ors.​org/
Current Diabetes Reports (2022) 22:227–235

Insulin-related traits insulin, 2 h glucose, South Asian, and

Consortium HbA1c) sub-Saharan African
SIGMA Slim Initiative in T2D, cancer, kidney Genotype, exome chip 8,227 T2D cases and Hispanic The SIGMA Type 2 https://​kp4cd.​org/​sigma
Genomic Medicine disease 12,966 controls Diabetes Consortium.
for the Americas Nature 2014;
Mercader et. al. Diabe-
tes 2017
AGEN Asian Genetic Epide- T2D, cardiovascular Genotype 77,418 T2D cases and East Asian Spracklen et al. Nature https://​blog.​nus.​edu.​sg/​
miology Network disease 356,122 controls 2020 agen/
T2D-GENES Type 2 Diabetes T2D Whole exome sequenc- 20,791 T2D cases and Hispanic/Latino, Euro- Flannick et al. Nature https://​www.​kp4cd.​org/​
Genetic Exploration ing 24,440 controls pean, African-Amer- 2019 about/​t2d
by Next-generation ican, East-Asian, and
sequencing in multi- South-Asian
Ethnic Samples
MEDIA African Americans T2D Genotype 8,284 T2D cases and African American Ng et al. PLoS Genet -
from the MEta-analy- 15,543 controls 2015
sis of type 2 DIabetes
in African Americans
MVP Million Veteran Pro- Disease agnostic Genotype 825,000 Predominantly Euro- Gaziano et al. J Clin https://​www.​resea​rch.​va.​
gram pean, also African Epidemiol 2016 gov/​mvp/
American, Hispanic,
and Asian
UKB UK Biobank Disease agnostic Genotype, whole Genotype: 500,000; Predominantly Euro- Bycroft et al. Nature https://​www.​ukbio​bank.​
exome sequenc- Whole exome: pean, also African 2018 ac.​uk/
ing, whole genome 450,000; American, Hispanic,
sequencing Whole genome: Asian, African, and
150,000 Carribbean
BBJ BioBank Japan Disease agnostic Genotype, whole Genotype: 200,000; Japanese Nagai et al. J Epide- http://​jenger.​riken.​jp/​en/
genome sequencing Whole genome: 2,000 miol 2017
PAGE Population Architec- Disease agnostic Genotype 50,000 European, African Wojcik et al. Nature https://​www.​pages​tudy.​
ture Genomics and American, Asian, 2019 org/
Epidemiology Native Hawaiian, and
Hispanic

13
229
 230

13
Table 2  Summary of recent large-scale T2D genetic association studies
Citation Cohort(s) Meta-analysis sample size Ancestry Phenotype(s) Total risk loci Novel risk loci

Mahajan et al. Nat. Gen 2018 DIAGRAM 74,124 T2D cases and European T2D 243 135
824,006 controls
Suzuki et al. Nat. Gen 2019 BBJ, Osaka-Midousuji 36,614 T2D cases and Japanese T2D 88 28
Rotary Club, Pharma SNP, 155,150 controls
ToMMo, IIMM, JPHC,
and J-MICC
Spracklen et al. Nature 2020 AGEN, DIAMANTE, CKB, 77,418 T2D cases and East Asian T2D 183 61
KBA, and BBJ 356,122 controls
Vujkovic et al. Nat. Gen MVP, DIAMANTE, Penn 228,499 T2D cases and European, African Ameri- T2D 568 318
2020 Medicine Biobank, Paki- 1,178,783 controls can, Hispanic, South
stan Genomic Resource, Asian, and East Asian
BBJ, Malmö Diet and
Cancer Study, Medstar,
and PennCath
Polfus et al. HGG Advances PAGE, DIAGRAM; 53,102 T2D cases and European, African Ameri- T2D 39 4; 2 of which replicated
2021 Replication: 23andMe, DIA- 193,679 controls can, Hispanic, Asian, and
MANTE, SIGMA, AGEN, Native Hawaiian
MEDIA
Chen et al. Nat. Gen 2021 MAGIC 281,416 non-T2D individu- European, African Ameri- Fasting glucose, fasting 242 72 (99 at time of writing)
als can, sub-Saharan African, insulin, 2-h glucose,
Hispanic, South Asian, and HbA1c
East Asian

Abbreviations: BBJ, Biobank Japan; ToMMo, Tohoku Medical Megabank Organization; IMM, Iwate Tohoku Medical Megabank Organization; JPHC, Japan Public Health Center–based Pro-
spective; J-MICC, Japan Multi-Institutional Collaborative Cohort; AGEN, Asian Genetic Epidemiology Network; DIAMANTE, DIAbetes Meta-ANalysis of Trans-Ethnic association studies;
CKB, China Kadoorie Biobank; KBA, Korea Biobank Array; SIGMA, Slim Initiative in Genomic Medicine for the Americas; MEDIA, African Americans from the MEta-analysis of type 2 DIa-
betes in African Americans
Current Diabetes Reports (2022) 22:227–235
Current Diabetes Reports (2022) 22:227–235 231

Of the 235 trans-ancestry glycemic trait-associated loci, associated with T2D in both East Asian and European
93 were novel at the time of publication and Chen et al. ancestry were strongly correlated (r = 0.87). Furthermore,
performed an instructive simulation to quantify the ben- the authors find that only 8.4% of variants showed signifi-
efit of including multiple ancestries as opposed to simply cant heterogeneity in effect size between the East Asian and
increasing sample size to enhance novel locus discovery. European GWAS results, and the variants which have the
By re-scaling the standard errors of the European single- greatest differences in effect sizes between the two popula-
ancestry GWAS to simulate the trans-ancestry sample size, tions are those that are common or low-frequency in East
Chen et al. found that that 21 out of the 93 (22.6%) newly Asians but rare in Europeans (MAF < 0.1%). Overall, these
discovered trans-ancestry loci would not have been identified recent T2D genetic association studies in East Asian ances-
in a GWAS restricted to European ancestry. This suggests try cohorts underscore the finding that genetic susceptibil-
that while the majority of novel loci were identified due to ity to T2D captured by common genetic variation is mostly
increase in sample size, a modest benefit was obtained by shared across ancestries.
including non-European samples.
Furthermore, this study examined the effect of single-
versus trans-ancestry analyses on the resolution of genetic Genetic Risk Scores for T2D Do Not
fine-mapping to identify causal variants. To do this, the Substantially Enhance Risk Prediction
authors identified 98 locus-trait associations that had a single over Traditional Clinical Risk Factors
causal variant from both single- and trans-ancestry fine-map-
ping and found that 72 (73%) locus-trait associations showed While over 700 loci identified by common variant associa-
improvements in the resolution of fine-mapping, as quanti- tion studies (i.e., GWAS) combine to explain almost 20%
fied by a decreased number of variants in the 99% credible of T2D heritability [19•], each individual common variant
sets. Of these 72 locus-trait associations, 53% were improved (i.e., SNP) has a small to modest effect (10–30%) on disease
due to larger sample size in the trans-ancestry analysis and risk as compared to simply knowing family history of T2D,
47% were improved due to the inclusion of diverse ancestries which if present in a parent confers a large increase in risk
as demonstrated by a decrease in the median number of vari- (~ two–threefold)[23]. Combining multiple variants geno-
ants in the 99% credible sets from 24 to 15 variants (37.5% typed in a single person into a genetic risk score (GRS, also
median reduction). Thus, for about half the loci identified, commonly referred to as polygenic risk score) is a logical
inclusion of diverse ancestries enabled a reduction of about strategy to enhance the clinical utility of genetic information
10 variants from the final 99% credible sets for the causal from common variants to identify individuals at high risk
variant. [24]. GRS combining multiple loci were initially tested in
In addition to the above-described multi-ancestry stud- the early 2000s with the first T2D GWAS studies. One of the
ies, recent large-scale T2D genetic studies have also been first studies calculated a T2D GRS from a combination of
performed in East Asian populations which have been previ- 18 loci finding that genetic information minimally enhanced
ously under-represented in GWAS. In a T2D case–control risk prediction when combined with traditional clinical risk
GWAS meta-analysis including Biobank Japan participants, factors such as age, sex, or family history of diabetes [25]. In
Suzuki et al. examined 36,614 T2D cases and 155,150 con- the past few years, there has been a resurgence of interest in
trols of Japanese ancestry and identified 88 T2D risk loci, GRS leveraging many more loci identified from large-scale,
28 of which were novel [21]. The majority (77%) of the multi-ancestry cohorts.
identified lead variants are common (MAF > 0.05) in both In the largest T2D GWAS to date (Table 2), Vujkovic
Japanese and European populations, and Suzuki et al. dem- et al. used results from a previous European GWAS [18] to
onstrated that effect sizes are strongly correlated (Pearson’s calculate GRS for participants in the MVP and demonstrated
r = 0.83, P = 8.7e-51) and directly consistent (94%) between that individuals with the highest T2D GRS (90–100% GRS
the Japanese GWAS and a comparable T2D European percentile) presented the highest risk for T2D (OR = 5.21,
GWAS, indicating that the majority of genetic susceptibil- 95% CI = 4.94–5.49) compared to those with the lowest T2D
ity between Japanese and European ancestry is shared. In GRS (0–10% GRS percentile) [19•]. Using the GWAS effect
addition to this study in Japanese individuals, the largest estimates from the T2D GWAS conducted by Vujkovic et al.,
meta-analysis of T2D GWAS in individuals of East Asian Polfus et al. computed a GRS for T2D in a multi-ethnic
ancestry to date examined 77,418 T2D cases and 356,122 cohort [20]. From this, they found that GRS constructed
controls across 23 studies including AGEN and Biobank from multi-ethnic computed weights demonstrated nominal
Japan to identify 183 loci, of which 61 were novel [22•]. increases in predictive power compared to single-ancestry
Upon comparison with a previously published T2D GWAS computed weights, and observed strongly significant hetero-
in European individuals of similar sample size, Spracklen geneity across ancestries for accuracy of T2D risk predic-
et al. demonstrated that effect sizes of variants significantly tion. For instance, the multi-ethnic GRS without adjustment

13
232 Current Diabetes Reports (2022) 22:227–235

for clinical risk factors performed best in European and East Perspective on Future Genetic Mapping
Asian populations (AUC = 0.66 and 0.63, respectively) and Studies in T2D
most poorly in African Americans (AUC = 0.57).
These recent studies which have generated GRS for T2D With over 700 T2D risk loci identified by common variant
and its related phenotypes have demonstrated that GRS has genetic association studies (i.e., GWAS), decades of fol-
the highest discriminative ability when applied to European low-up biological studies in cellular and organismal model
populations and that performance is subsequently improved systems will be required to fully understand the causal
in non-European ancestries when GRS is computed using genes and molecular mechanisms of disease pathogenesis.
multi-ancestry weights [20, 26]. However, even after a dec- Thus, it is unlikely that simply aggregating larger T2D
ade of methodological refinement as well as an increase in case:control cohorts for association analysis will provide
the number of loci to calculate GRS, the predictive power scientific and clinical insight into T2D. Here, we expect
of GRS for T2D is comparable to discrimination by clinical that an enhanced focus on T2D complications, which are
risk factors alone (Fig. 1). However, GRS may have a role the leading cause of death in T2D[5] and are indepen-
in detecting individuals at high risk before clinical risk fac- dently heritable of diabetes[38], using common variant
tors become apparent. Whether information from GRS can association methodology will advance understanding and
motivate preventative therapy to meaningfully reduce rates treatment as has been ongoing for T1D [39].
of future incident T2D remains to be studied. GRS have also It has long been appreciated that T2D is a highly het-
been applied widely beyond T2D to other heritable diseases erogeneous disorder classically defined along a spectrum
such as heart disease and cancer [27, 28] and even been of insulin secretion and insulin resistance which ultimately
offered as a tool for embryo screening during in vitro ferti- belies differences in clinical presentation, disease progres-
lization [29–31]. But the current consensus among geneti- sion, response to treatment, and susceptibility to compli-
cists, ethicists, and clinicians is that the scientific and tech- cations [40]. Recent work added four clinically available
nical uncertainty in GRS and their limited predictive power variables to insulin and glucose to refine T2D subtypes
should limit their use in genetic screening [32].
1e7
N variants

1e5
in GRS

1000

10

1 GRS
GRS+clinical risk factors
0.85
0.8

0.6
AUC

0.4

0.2

6 8 8 8 4 016 21 018 017 018

200 200 200 200 201 ct 2 s 20 ol 2
Med JM JM e tes e tes n Pra nce G en 2 emi en 2
P L o S a l. N E
a l . N E
. D i a b
D i a b
e s C l i
A d v a
l . N a t
. E p i d
. N at G
t t l l . n l
al. ko e gs e ta ta tes
R HG
G et a Ge ra e
ta
n et sen Mei go e sy e iabe t al. ajan t al. Khe
edo Lys Lan Vas l. D us e Mah na e
We e t a P o l f
z − L u
ore que
Chi
kow Mar

Fig. 1  Genetic risk scores for T2D do not substantially enhance risk tors to predict T2D from each study, quantified by the area under the
prediction over traditional clinical risk factors. Displayed are the out- receiver operating characteristic curve (AUC). The dashed red line
comes from independent, large-scale studies which have constructed and shaded red box represent the current predictive power and 95%
genetic risk scores (GRS) for T2D [18, 20, 25–27, 33–37]. Studies confidence interval respectively of T2D clinical risk factors (age, sex,
are shown along the x-axis, ordered by the number of variants used parental T2D, BMI, systolic blood pressure, fasting glucose, HDL
to construct the genetic risk scores (top panel). (bottom panel) The cholesterol, and triglycerides) to predict T2D [36]
accuracy of the GRS alone and the GRS with clinical T2D risk fac-

13
Current Diabetes Reports (2022) 22:227–235 233

which were shown to differ in patient characteristics and Declarations

risk of comorbidities [41]. Genetic association analysis of
these T2D subtypes has revealed partially distinct genetic Competing Interests The authors declare no competing interests.
backgrounds and heritability demonstrating progress in
refining T2D classification to reduce clinical heterogeneity Open Access This article is licensed under a Creative Commons Attri-
[42]. We expect that the use of omics measurements such bution 4.0 International License, which permits use, sharing, adapta-
tion, distribution and reproduction in any medium or format, as long
as transcriptomics, proteomics, and metabolomics applied
as you give appropriate credit to the original author(s) and the source,
to blood samples will enable the identification of novel provide a link to the Creative Commons licence, and indicate if changes
patterns to resolve T2D heterogeneity and in combina- were made. The images or other third party material in this article are
tion with genetic association methodologies enable iden- included in the article's Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in
tification of distinct biological pathways. Early efforts in
the article's Creative Commons licence and your intended use is not
the application of metabolic measurements to fasting and permitted by statutory regulation or exceeds the permitted use, you will
postprandial samples in concert with GWAS have shown need to obtain permission directly from the copyright holder. To view a
the potential of such omics approaches [43]. copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.
In contrast to common variants which were generated
millions of years ago in an genetically equilibrated ances-
tral human population, rare genetic variants (MAF << References
0.01) which arose during the “out of Africa” human pop-
ulation expansion [44] potentially offer different mecha- Papers of particular interest, published recently, have
nisms of disease causation. As exome and whole genome been highlighted as:
sequencing are becoming more commonplace, investiga- • Of importance
tors have begun to examine rare variant associations with
T2D [45]. The challenge with rare variant association 1. DeFronzo RA, Ferrannini E, Groop L, Henry RR, Herman WH,
studies is that the sample size requirement vastly increases Holst JJ, et al. Type 2 diabetes mellitus. Nat Rev Dis Primers.
2015;1:15019.
due to the low allele frequency and increase in multiple
2. American Diabetes Association. 2. Classification and diagnosis
hypothesis testing burden from the large number of rare of diabetes: standards of medical care in diabetes—2021. Dia-
variants [46]. Using a combination of methodological betes Care. American Diabetes Association; 2021;44:S15–33.
enhancements such as “burden tests” which aggregate rare 3. Spanakis EK, Golden SH. Race/ethnic difference in diabetes and
diabetic complications. Curr Diab Rep. 2013;13:814–23.
variants across a gene to reduce the multiple hypothesis
4. Morrish NJ, Wang SL, Stevens LK, Fuller JH, Keen H. Mortality
testing burden and population-scale biobanks like the UK and causes of death in the WHO Multinational Study of Vascular
Biobank to increase sample size, investigators have identi- Disease in Diabetes. Diabetologia. 2001;44(Suppl 2):S14-21.
fied novel T2D loci such as GIGYF1 [47] and FAM234A 5. Fox CS, Coady S, Sorlie PD, D’Agostino RB Sr, Pencina MJ,
Vasan RS, et al. Increasing cardiovascular disease burden due
[48] which were not marked by common variant signals.
to diabetes mellitus: the Framingham Heart Study. Circulation.
In summary, we expect that large-scale exome and whole- 2007;115:1544–50.
genome sequencing of population scale biobanks will facili- 6. Chatterjee S, Khunti K, Davies MJ. Type 2 diabetes. Lancet.
tate rare-variant association studies of T2D to identify novel 2017;389:2239–51.
7. Almgren P, Lehtovirta M, Isomaa B, Sarelin L, Taskinen MR,
loci beyond what has been identified by common variant
Lyssenko V, et al. Heritability and familiality of type 2 diabetes
association studies thus far. Additionally, focusing genetic and related quantitative traits in the Botnia Study. Diabetologia.
mapping efforts on micro- and macrovascular diabetes com- 2011;54:2811–9.
plications is likely to maximize the value of novel locus 8. Hemminki K, Li X, Sundquist K, Sundquist J. Familial risks for
type 2 diabetes in Sweden. Diabetes Care. 2010;33:293–7.
discovery to further understand and treat T2D.
9. Lyssenko V, Almgren P, Anevski D, Perfekt R, Lahti K, Nissén
M, et al. Predictors of and longitudinal changes in insulin sensi-
tivity and secretion preceding onset of type 2 diabetes. Diabetes.
Author Contribution N.D. and A.R.M. conceived the manuscript and 2005;54:166–74.
figures. N.D. drafted the manuscript. All authors were involved in criti- 10. Altshuler D, Hirschhorn JN, Klannemark M, Lindgren CM, Vohl
cal manuscript revision. M-C, Nemesh J, et al. The common PPARγ Pro12Ala polymor-
phism is associated with decreased risk of type 2 diabetes [Inter-
Funding This work was supported by grants from the National Institute net]. Nature Genetics. 2000. p. 76–80. Available from: https://​
of Diabetes and Digestive and Kidney Diseases (1R01DK123422-01 to doi.​org/​10.​1038/​79216
A.R.M), a UCSD/UCLA Pilot and Feasibility grant (P30 DK063491 11. Grant SFA, Thorleifsson G, Reynisdottir I, Benediktsson R,
to A.R.M), and a Ruth L. Kirschstein Institutional National Research Manolescu A, Sainz J, et al. Variant of transcription factor 7-like
Service Award T32 GM008666 from the National Institute of General 2 (TCF7L2) gene confers risk of type 2 diabetes. Nat Genet.
Medical Sciences (to N.D.). 2006;38:320–3.

13
234 Current Diabetes Reports (2022) 22:227–235

12. Bonnefond A, Froguel P. Rare and common genetic events 23. Meigs JB, Cupples LA, Wilson PW. Parental transmission of
in type 2 diabetes: what should biologists know? Cell Metab. type 2 diabetes: the Framingham Offspring Study. Diabetes.
2015;21:357–68. 2000;49:2201–7.
13. Flannick J, Florez JC. Type 2 diabetes: genetic data shar- 24. Wray NR, Goddard ME, Visscher PM. Prediction of indi-
ing to advance complex disease research. Nat Rev Genet. vidual genetic risk to disease from genome-wide association
2016;17:535–49. studies. Genome Res. Cold Spring Harbor Laboratory Press;
14. Scott RA, Scott LJ, Mägi R, Marullo L, Gaulton KJ, Kaakinen 2007;17:1520–8.
M, et al. An expanded genome-wide association study of type 2 25. Meigs JB, Shrader P, Sullivan LM, McAteer JB, Fox CS, Dupuis
diabetes in Europeans. Diabetes. 2017;66:2888–902. J, et al. Genotype score in addition to common risk factors for
15. Dupuis J, Langenberg C, Prokopenko I, Saxena R, Soranzo N, prediction of type 2 diabetes. N Engl J Med. 2008;359:2208–19.
Jackson AU, et al. New genetic loci implicated in fasting glucose 26 Márquez-Luna C, Loh P-R. South Asian Type 2 Diabetes
homeostasis and their impact on type 2 diabetes risk. Nat Genet. (SAT2D) Consortium, SIGMA Type 2 Diabetes Consortium,
2010;42:105–16. Price AL. Multiethnic polygenic risk scores improve risk predic-
16. Mahajan A, Spracklen CN, Zhang W, Ng MCY, Petty LE, Kita- tion in diverse populations. Genet Epidemiol. 2017;41:811–23.
jima H, et al. Trans-ancestry genetic study of type 2 diabetes 27. Khera AV, Chaffin M, Aragam KG, Haas ME, Roselli C, Choi
highlights the power of diverse populations for discovery and SH, et al. Genome-wide polygenic scores for common diseases
translation [Internet]. bioRxiv. medRxiv; 2020. Available from: identify individuals with risk equivalent to monogenic muta-
http://medrxiv.org/lookup/doi/https://​doi.​org/​10.​1101/​2020.​09.​ tions. Nat Genet. 2018;50:1219–24.
22.​20198​937 28. Zhang YD, Hurson AN, Zhang H, Choudhury PP, Easton DF,
17• Chen J, Spracklen CN, Marenne G, Varshney A, Corbin LJ, Milne RL, et al. Assessment of polygenic architecture and risk
Luan J, et al. The trans-ancestral genomic architecture of gly- prediction based on common variants across fourteen cancers.
cemic traits. Nat Genet. 2021;53:840–60. In this most recent Nat Commun. 2020;11:3353.
study from the MAGIC consortium, investigators examined 29. LifeView: Successful pregnancy, Healthy baby [Internet]. [cited
the effect of ancestry on loci associated with glycemic traits 2021 Dec 29]. Available from: https://​genom​icpre​dicti​on.​com/
(fasting glucose, fasting insulin, 2-hour glucose, and HbA1c) 30. Orchid: Have healthy babies [Internet]. [cited 2021 Dec 29].
in 281,416 non-diabetic individuals. Additionally, this study Available from: https://​www.​orchi​dheal​th.​com/
quantified the improvements in statistical power for genetic 31. MyOme – Harnessing the true power of genetics [Internet].
discovery and in fine-mapping that resulted from performing [cited 2021 Dec 29]. Available from: https://​myome.​com/
a trans-ancestry analysis compared to single-ancestry. 32. Turley P, Meyer MN, Wang N, Cesarini D, Hammonds E, Mar-
18. Mahajan A, Taliun D, Thurner M, Robertson NR, Torres JM, tin AR, et al. Problems with using polygenic scores to select
Rayner NW, et al. Fine-mapping type 2 diabetes loci to single- embryos. N Engl J Med. 2021;385:78–86.
variant resolution using high-density imputation and islet-spe- 33. Weedon MN, McCarthy MI, Hitman G, Walker M, Groves CJ,
cific epigenome maps. Nat Genet [Internet]. Nat Genet; 2018 Zeggini E, et al. Combining information from common type 2
[cited 2021 Dec 9];50. Available from: https://p​ ubmed.n​ cbi.n​ lm.​ diabetes risk polymorphisms improves disease prediction. PLoS
nih.​gov/​30297​969/ Med. 2006;3:e374.
19.• Vujkovic M, Keaton JM, Lynch JA, Miller DR, Zhou J, Tchean- 34. Lyssenko V, Jonsson A, Almgren P, Pulizzi N, Isomaa B, Tuomi
djieu C, et al. Discovery of 318 new risk loci for type 2 diabetes T, et al. Clinical risk factors, DNA variants, and the development
and related vascular outcomes among 1.4 million participants of type 2 diabetes. N Engl J Med. 2008;359:2220–32.
in a multi-ancestry meta-analysis. Nat Genet. Nature Publish- 35. Lango H, UK Type 2 Diabetes Genetics Consortium, Palmer
ing Group; 2020;52:680–91. This study represents the largest CNA, Morris AD, Zeggini E, Hattersley AT, et al. Assessing
T2D genetic association study to date in which investigators the combined impact of 18 common genetic variants of modest
meta-analyzed GWAS from eight cohorts composed of five effect sizes on type 2 diabetes risk. Diabetes. 2008;57:3129–35.
different ancestral groups (European, African American, 36. Vassy JL, Hivert M-F, Porneala B, Dauriz M, Florez JC, Dupuis
Hispanic, South Asian, and East Asian) for a total of 228,499 J, et al. Polygenic type 2 diabetes prediction at the limit of com-
T2D cases and 1,178,783 controls to identify 318 novel T2D mon variant detection. Diabetes. 2014;63:2172–82.
risk loci. 37. Chikowore T, van Zyl T, Feskens EJM, Conradie KR. Predictive
20. Polfus LM, Darst BF, Highland H, Sheng X, Ng MCY, Below utility of a genetic risk score of common variants associated with
JE, et al. Genetic discovery and risk characterization in type type 2 diabetes in a black South African population. Diabetes
2 diabetes across diverse populations. HGG Adv [Internet]. Res Clin Pract. 2016;122:1–8.
2021;2. Available from: https://​doi.​org/​10.​1016/j.​xhgg.​2021.​ 38. Seaquist ER, Goetz FC, Rich S, Barbosa J. Familial clustering of
100029 diabetic kidney disease. N Engl J Med. Massachusetts Medical
21. Suzuki K, Akiyama M, Ishigaki K, Kanai M, Hosoe J, Shojima Society; 1989;320:1161–5.
N, et al. Identification of 28 new susceptibility loci for type 2 39. Cole JB, Florez JC. Genetics of diabetes mellitus and diabe-
diabetes in the Japanese population. Nat Genet. 2019;51:379–86. tes complications [Internet]. Nature Reviews Nephrology.
22.• Spracklen CN, Horikoshi M, Kim YJ, Lin K, Bragg F, Moon S, 2020. p. 377–90. Available from: https:// ​d oi.​o rg/​1 0.​1 038/​
et al. Identification of type 2 diabetes loci in 433,540 East Asian s41581-​020-​0278-5
individuals. Nature. 2020;582:240–5. In this study which is the 40 Defronzo RA. Banting Lecture. From the triumvirate to the omi-
largest meta-analysis of T2D GWAS in individuals of East nous octet: a new paradigm for the treatment of type 2 diabetes
Asian ancestry to date, investigators examined 77,418 T2D mellitus. Diabetes. 2009;58:773–95.
cases and 356,122 controls across 23 studies to identify 183 41. Ahlqvist E, Storm P, Käräjämäki A, Martinell M, Dorkhan M,
loci, of which 61 were novel. This study demonstrates that Carlsson A, et al. Novel subgroups of adult-onset diabetes and
genetic susceptibility to T2D captured by common genetic their association with outcomes: a data-driven cluster analysis
variation is mostly shared between East Asian and European of six variables. Lancet Diabetes Endocrinol. 2018;6:361–9.
ancestries. 42. Mansour Aly D, Dwivedi OP, Prasad RB, Käräjämäki A,
Hjort R, Thangam M, et al. Genome-wide association analyses

13
Current Diabetes Reports (2022) 22:227–235 235

highlight etiological differences underlying newly defined sub- rare variant association studies. Proc Natl Acad Sci U S A.
types of diabetes. Nat Genet. 2021;53:1534–42. 2014;111:E455–64.
43. Li-Gao R, Hughes DA, van Klinken JB, de Mutsert R, Rosendaal 47. Deaton AM, Parker MM, Ward LD, Flynn-Carroll AO, Bon-
FR, Mook-Kanamori DO, et al. Genetic studies of metabolomics Durant L, Hinkle G, et al. Gene-level analysis of rare variants
change after a liquid meal illuminate novel pathways for glucose in 379,066 whole exome sequences identifies an association of
and lipid metabolism. Diabetes. 2021;70:2932–46. GIGYF1 loss of function with type 2 diabetes. Sci Rep Nature
44. Majithia AR, Altshuler D, Hirschhorn JN. Genetics of endocri- Publishing Group. 2021;11:1–16.
nology [Internet]. Williams Textbook of Endocrinology. 2016. 48. Backman JD, Li AH, Marcketta A, Sun D, Mbatchou J, Kes-
p. 49–68. Available from: https://​doi.​org/​10.​1016/​b978-0-​323-​ sler MD, et al. Exome sequencing and analysis of 454,787
29738-7.​00004-6 UK Biobank participants. Nature Nature Publishing Group.
45. Tennessen JA, Bigham AW, O’Connor TD, Fu W, Kenny EE, 2021;599:628–34.
Gravel S, et al. Evolution and functional impact of rare cod-
ing variation from deep sequencing of human exomes. Science. Publisher's Note Springer Nature remains neutral with regard to
2012;337:64–9. jurisdictional claims in published maps and institutional affiliations.
46. Zuk O, Schaffner SF, Samocha K, Do R, Hechter E, Kathire-
san S, et al. Searching for missing heritability: designing

13