Diabetes Care Volume 44, November 2021 2449

Adult-Onset Type 1 Diabetes: R. David Leslie,1

Carmella Evans-Molina,2,3
Current Understanding and Jacquelyn Freund-Brown,4
Raffaella Buzzetti,5 Dana Dabelea,6
Challenges Kathleen M. Gillespie,7 Robin Goland,8
Angus G. Jones,9 Mark Kacher,4
Diabetes Care 2021;44:2449–2456 | https://doi.org/10.2337/dc21-0770 Lawrence S. Phillips,10 Olov Rolandsson,11
Jana L. Wardian,12 and Jessica L. Dunne4

Recent epidemiological data have shown that more than half of all new cases of
type 1 diabetes occur in adults. Key genetic, immune, and metabolic differences
exist between adult- and childhood-onset type 1 diabetes, many of which are not
well understood. A substantial risk of misclassification of diabetes type can
result. Notably, some adults with type 1 diabetes may not require insulin at diag-
1
nosis, their clinical disease can masquerade as type 2 diabetes, and the conse- Centre for Immunobiology, Blizard Institute,
quent misclassification may result in inappropriate treatment. In response to this Queen Mary University of London, London, U.K.
2
Departments of Pediatrics and Medicine and
important issue, JDRF convened a workshop of international experts in Novem- Center for Diabetes and Metabolic Diseases,
ber 2019. Here, we summarize the current understanding and unanswered ques- Indiana University School of Medicine,
tions in the field based on those discussions, highlighting epidemiology and Indianapolis, IN
3
immunogenetic and metabolic characteristics of adult-onset type 1 diabetes as Richard L. Roudebush VA Medical Center,
Indianapolis, IN
well as disease-associated comorbidities and psychosocial challenges. In adult- 4
JDRF, New York, NY
onset, as compared with childhood-onset, type 1 diabetes, HLA-associated risk is 5
Department of Experimental Medicine, Sapienza
lower, with more protective genotypes and lower genetic risk scores; multiple University of Rome, Rome, Italy
6
Lifecourse Epidemiology of Adiposity & Diabetes
diabetes-associated autoantibodies are decreased, though GADA remains domi-
Center, Colorado School of Public Health, and
nant. Before diagnosis, those with autoantibodies progress more slowly, and at Departments of Epidemiology and Pediatrics,
diagnosis, serum C-peptide is higher in adults than children, with ketoacidosis University of Colorado Anschutz Medical Campus,
being less frequent. Tools to distinguish types of diabetes are discussed, including Aurora, CO
7
Translational Health Sciences, Bristol Medical
body phenotype, clinical course, family history, autoantibodies, comorbidities,
School, University of Bristol, Bristol, U.K.
and C-peptide. By providing this perspective, we aim to improve the manage- 8
Naomi Berrie Diabetes Center, Columbia

PERSPECTIVES IN CARE
ment of adults presenting with type 1 diabetes. University, New York, NY
9
Institute of Biomedical and Clinical Science,
University of Exeter, Exeter, U.K.
10
Atlanta VA Medical Center and Division of
Endocrinology, Metabolism, and Lipids, Department
Clinically, it has been relatively easy to distinguish the acute, potentially lethal, of Medicine, Emory University School of Medicine,
childhood-onset diabetes from the less aggressive condition that affects adults. Atlanta, GA
11
However, experience has taught us that not all children with diabetes are insulin Department of Public Health and Clinical
dependent and not all adults are non–insulin dependent. Immune, genetic, and Medicine, Umeå University, Umeå, Sweden
12
College of Medicine, University of Nebraska
metabolic analysis of these two, apparently distinct, forms of diabetes revealed Medical Center, Omaha, NE
inconsistencies, such that insulin-dependent and immune-mediated diabetes was Corresponding author: Carmella Evans-Molina,
redefined as type 1 diabetes, while most other forms were relabeled as type 2 dia- cevansmo@iu.edu, or R. David Leslie, r.d.g.
betes. Recent data suggest a further shift in our thinking, with the recognition that leslie@qmul.ac.uk
more than half of all new cases of type 1 diabetes occur in adults. However, many Received 7 April 2021 and accepted 12 August
adults may not require insulin at diagnosis of type 1 diabetes and have a more 2021
gradual onset of hyperglycemia, often leading to misclassification and inappropriate © 2021 by the American Diabetes Association.
care. Indeed, misdiagnosis occurs in nearly 40% of adults with new type 1 diabetes, Readers may use this article as long as the
work is properly cited, the use is educational
with the risk of error increasing with age (1,2). To consider this important issue,
and not for profit, and the work is not altered.
JDRF convened a workshop of international experts in November 2019 in New More information is available at https://www.
York, NY. In this Perspective, based on that workshop, we outline the evidence for diabetesjournals.org/content/license.
2450 Adult-Onset Type 1 Diabetes Diabetes Care Volume 44, November 2021

a new viewpoint, suggesting future direc- study in China, the fraction (8.6%) with the term latent autoimmune diabetes in
tions of research and ways to alter dis- diabetes not requiring insulin yet with adults (LADA) has been used to describe
ease management to help adults living type 1 diabetes–associated autoantibod- adults with slowly progressive autoimmu-
with type 1 diabetes. ies was similar to that in Europe, implying nity, sometimes exhibiting features over-
that there could be over 6 million Chi- lapping with those of type 2 diabetes
UNDERSTANDING ADULT-ONSET nese with adult-onset type 1 diabetes (9,14,18). At the outset of the workshop
TYPE 1 DIABETES (10). While there is a wide range in the and for the purposes of this Perspective,
Incidence of Type 1 Diabetes Among incidence of type 1 diabetes across differ- LADA was not considered a unique entity;
Adults Worldwide ent ethnic groups, even using differing rather, we considered the classification of
Adult-onset type 1 diabetes is more com- methods of case identification (7), these type 1 diabetes to include all individuals
mon than childhood-onset type 1 diabe- data support the notion that, worldwide, with evidence of autoimmunity, regardless
tes, as shown from epidemiological data over half of all new-onset type 1 diabetes of the trajectory of disease development
from both high-risk areas such as North- cases occur in adults. (i.e., rapid or slowly progressive) or other
ern Europe and low-risk areas such as associated demographic and/or clinical fea-
China (3–8). In southeastern Sweden, the Natural History Studies of Type 1 tures (e.g., obesity).
disease incidence among individuals aged Diabetes
0–19 years is similar to that among indi- Our understanding of the natural history Age-Related Genetic Heterogeneity
viduals 40–100 years of age (37.8 per of type 1 diabetes has been informed by Type 1 diabetes shows heterogeneity
100,000 persons per year and 34.0/ a number of longitudinal and cross-sec- across a broad range of clinical, genetic,
tional studies. At one end of the spec- immune, histological, and metabolic
100,000/year, respectively) (3). Given that
trum are prospective birth cohort studies, features (20). Childhood-onset type 1
the comparable incidence spans only two
such as the BABYDIAB study in Germany diabetes is most often attributed to sus-
decades in children, it follows that adult-
and The Environmental Determinants of ceptibility alleles in human leukocyte
onset type 1 diabetes is more prevalent.
Similarly, analysis of U.S. data from com-
Diabetes in the Young (TEDDY) study, antigen (HLA), which contribute 50%
which includes sites in Germany, Finland, of the disease heritability. Whereas eth-
mercially insured individuals demon-
Sweden, and the U.S. While these studies nic differences exist, notably for specific
strated an overall lower incidence in
now have the potential to explore the HLA genotypes, several broad principles
individuals 20–64 years of age (18.6/
pathogenesis of islet autoimmunity by apply. Compared with childhood-onset
100,000/year) than in youth aged 0–19
being extended into adulthood, they have disease, adult-onset type 1 diabetes
years (34.3/100,000/year), but the total
primarily focused on events occurring in cases show lower type 1 diabetes con-
number of new cases in adults over a 14-
childhood (11). Clinical centers in North cordance rates in twins (21), less high-
year period was 19,174 compared with
America, Europe, and Australia collabo- risk HLA heterozygosity (19), lower HLA
13,302 in youth (4). Despite the incidence rate within Type 1 Diabetes TrialNet, a class I (14), more protective genotypes
of childhood-onset type 1 diabetes in study that identifies autoantibody-positive (14,15), and lower GRS (6,22), which
China being among the lowest in the adults and children in a cross-sectional are calculated by summing the odds
world, prevalence data show similar manner to examine the pathogenesis of ratios (OR) for disease-risk alleles.
trends across the life span. From type 1 diabetes and to perform clinical tri-
2010–2013, the incidence was 1.93/ als on those at high risk in order to pre- Diabetes-Associated Immune
100,000 among individuals aged 0–14 serve b-cell function (12). At the other Changes
years and 1.28/100,000 among those end of the spectrum, the European Pro- Adult-onset type 1 diabetes, like child-
15–29 years of age versus 0.69/100,000 spective Investigation into Cancer and hood-onset type 1 diabetes, is associated
among older adults (5). In aggregate, Nutrition (EPIC)-InterAct study is a case- with the presence of serum autoantibod-
adults comprised 65.3% of all clinically cohort study nested in the U.K. prospec- ies against b-cell antigens. Serum glutamic
defined newly diagnosed type 1 diabetes tive adult population-based EPIC study acid decarboxylase (GADA) autoantibodies
cases in China, which is similar to esti- (13), while the clinical, immunogenetic, may be useful as a predictor of type 1
mates using genetically stratified data and metabolic characteristics of autoim- diabetes in adults, as adult-onset cases
from the population-based UK Biobank mune adult-onset type 1 diabetes have most often present with GADA positivity
using a childhood-onset polygenic genetic been extensively studied in large Ameri- (9,10,15,17,18,20,22) and possess an HLA-
risk score (GRS) (6). It is important to can, European, and Chinese studies, DR3 genotype (9,14,15, 20,21,23). In one
note that the proportion would likely be including UK Prospective Diabetes Study prospective study of a general population,
higher if autoimmune cases not requiring (UKPDS), Action LADA, Scandia, Non the hazard risk of incident diabetes in
insulin initially were classified as type 1 Insulin Requiring Autoimmune Diabetes those with a high type 1 diabetes GRS and
diabetes. For example, in a clinic-based (NIRAD), and LADA China (9,14–19). Based GADA positivity was 3.23 compared with
European study, the proportion of adults on these cross-sectional and prospective all other individuals, suggesting that
with diabetes not initially requiring insulin studies, considerable data have been gen- 1.8% of incident diabetes in adults was
yet with type 1 diabetes–associated auto- erated to define differences within type 1 attributable to that combination of risk
antibodies was even higher than those diabetes according to the age at onset. factors (13). In adult-onset type 1 diabetes,
started on insulin at diagnosis with a Here, we highlight key aspects of age- multiple diabetes-associated autoantibod-
defined type 1 diabetes diagnosis (9). related genetic, immune, and metabolic ies tend to be less prevalent with increas-
Moreover, in an adult population-based heter-ogeneity in type 1 diabetes. Of note, ing age at diagnosis (1,8), yet GADA
care.diabetesjournals.org Leslie and Associates 2451

remains the dominant autoantibody irre- However, there is a paucity of immune Association (ADA) targets for glycemia
spective of the need for insulin treatment studies on adult-onset type 1 diabetes are higher in children, so that in this
at diagnosis and irrespective of ethnicity and few histologic studies. An analysis of same cohort, 17% of children, com-
(9,17,18,24, tissues from the Network for Pancreatic pared with 21% of adults, achieved the
25), even despite a paucity of HLA DR3, as Organ Donors with Diabetes (nPOD) ADA hemoglobin A1c (HbA1c) goal of
in Japan and China (17,18). In contrast, showed no relationship between age at <7.5% and <7.0%, respectively (37).
childhood-onset type 1 diabetes cases diabetes onset and the frequency of islet Other factors confound this relationship
often have insulin autoantibodies and an insulitis (31). The composition of islet between age at diagnosis and metabolic
HLA-DR4 genotype, higher identical twin insulitis differs in very young children control. First, individuals with adult-
disease concordance, more HLA heterozy- compared with older individuals, with the onset type 1 diabetes are more likely to
gosity, and higher GRS (20). Taken former having an increased frequency of have residual insulin-producing b-cells
together, these data indicate that type 1 B cells in islet infiltrates (32). However, and persistent measurable C-peptide in
diabetes is heterogeneous across the spec- relating pancreatic histological changes to disease of long duration, the latter of
trum of diagnoses, suggesting that patho- changes in peripheral blood remains a which has been linked to improved gly-
genesis and optimal therapy are also challenge. cemic control (38,39). Second, individuals
diverse. Adults with new-onset type 1 diabetes with adult-onset type 1 diabetes, initially
Data from the TrialNet Pathway to are at increased risk of other autoim- not on insulin therapy, tend to have
Prevention cohort demonstrated lower mune conditions. About 30% of individu- worse metabolic control than people
risk of progression to type 1 diabetes in als with adult-onset type 1 diabetes have with type 2 diabetes, even when receiv-
adults than children, even when both thyroid autoimmunity (27,29). In addition, ing insulin treatment (9,40). The sole
show multiple autoantibodies on a sin- adults with type 1 diabetes who possess exception is the LADA China study,
gle occasion and are monitored over 10 high-titer GADA and/or multiple islet where worse control was noted only
years (12). One recent analysis found autoantibodies are at increased risk of among those with a high GAD titer (18).
that the 5-year rate of progression to progression to hypothyroidism (24,33). In Metabolic differences between adults
diabetes in multiple autoantibody–posi- a large population-based Chinese study, and children extend beyond C-peptide.
tive adults was only 15%, with a num- the prevalence of adult-onset type 1 dia- Adults with autoantibody positivity who
ber of them remaining diabetes-free for betes was 6% among initially non-insulin- progressed to type 1 diabetes were less
decades (26). A combined cohort study, requiring diabetes cases, and 16.3% of likely than very young children to exhibit
known as the Slow or Nonprogressive them had thyroid autoimmunity (OR 2.4) elevated proinsulin/C-peptide ratios prior
Autoimmunity to the Islets of Langer- (10). Of note, those with islet antigen 2 to stage 3 disease onset (41). In addition,
hans (SNAIL) study, is following such autoantibodies had a high risk of tissue in individuals with disease of long dura-
“slow progressors” with multiple auto- transglutaminase autoantibodies, a marker tion, those diagnosed at an older age
antibodies who have yet to progress to for celiac disease (OR 19.1) (10). Thus, in had evidence of improved proinsulin
stage 3 type 1 diabetes (i.e., clinical the clinical setting, there should be a high processing and nutrient-induced proinsu-
diagnosis) over at least a 10-year period index of suspicion for other autoimmune lin secretory capacity (42).
(27). Many of these slow progressors conditions in individuals with adult-onset
lose disease-associated autoantibodies type diabetes, and associated autoimmu- Diagnosis and Management of Adult-
over time, adding complexity to cross- nity should be screened where clinically Onset Type 1 Diabetes
sectional classification (28). Based on indicated. Correctly identifying diabetes etiology
estimates from natural history studies, and type is difficult, and misclassifica-
slow progressors, even if identified Metabolic Characteristics of Adult- tion may occur in up to 40% of adults
when young, cannot account for all Onset Type 1 Diabetes presenting with type 1 diabetes (1,2).
autoimmune adult-onset diabetes, indi- Age-related differences in type 1 diabe- Reasons underlying misclassification are
cating that autoantibodies must develop tes extend to metabolic parameters. C- multiple and include 1) lack of aware-
at all ages (11). However, little is known peptide at diagnosis is higher in adults ness that the onset of type 1 diabetes is
about those who initially develop auto- than children, driven in part by higher not limited to children; 2) the over-
immunity as adults, mostly due to the BMI (34). Analysis of U.K., TrialNet, and whelming majority of people developing
lack of longitudinal studies focusing on Chinese cohorts has identified two dis- diabetes as older adults have type 2 dia-
this population. tinct phases of C-peptide decline in betes, contributing to a confirmation
People with type 1 diabetes, in con- stage 3 disease: an initial exponential bias (2); 3) typical clinical criteria, such
trast to the majority of those with type 2 fall followed by a period of relative sta- as BMI and metabolic syndrome, can be
diabetes, have altered adaptive immunity bility. Along with initial differences at poor discriminators, especially as rates
(i.e., islet autoantibodies and T-cell activa- the time of clinical diagnosis, the rate of of obesity in the overall population
tion), while innate immune changes, decline over 2–4 years was inversely increase (9,43); 4) clinical characteristics
including cytokine changes, are common related to age at onset (10,34–36). Fur- of adult-onset type 1 diabetes can mas-
to both (29). Increased T-cell activation by thermore, the U.S. T1D Exchange Study querade as type 2 diabetes, given their
islet proteins has also been found in a found that glycemic control was better slow metabolic progression and risk of
proportion of adults with initially non- in adults with type 1 diabetes than in metabolic syndrome (which occurs in
insulin-requiring diabetes, even when children and adolescents with type 1 about 40%), so that the distinction
they lack diabetes autoantibodies (30). diabetes (37). The American Diabetes between types of diabetes may be
2452 Adult-Onset Type 1 Diabetes Diabetes Care Volume 44, November 2021

blurred (43–45); and 5) lack of awareness utility in disease of long duration when been tested in multiethnic populations,
of and accessibility to biomarkers that levels fall below 300 pmol/L (39,47). we suggest, as an approach to aid the
may serve as tools to distinguish type 1 However, C-peptide levels are typically practicing physician, assessment of age,
diabetes and type 2 diabetes. higher at presentation and may be diffi- autoimmunity, body habitus/BMI, back-
Tools to distinguish type 1 and type 2 cult to distinguish from levels in type 2 ground, control, and comorbidities, using
diabetes are under active development. diabetes, which are usually >600 pmol/ the acronym AABBCC (Table 2). This
For example, classification models inte- L. Thus, thresholds of C-peptide that approach includes the clinical consider-
grating up to five prespecified predictor clearly delineate type 1 diabetes from ation of autoimmunity and other clinical
variables, including clinical features (age type 2 diabetes at diagnosis cannot be features suggestive of type 1 diabetes,
of diagnosis and BMI) and clinical bio- categorically defined, and C-peptide including age at diagnosis, low BMI, an
markers (autoantibodies and GRS) in a must be interpreted within the context unexplained or rapid worsening of clini-
White European population, had high of other clinical and laboratory features. cal course manifesting as a lack of
accuracy to identify adults with recently Measurement of a random nonfasting response or rising HbA1c with type 2 dia-
diagnosed diabetes with rapid insulin C-peptide is superior to fasting C-pep- betes medications, and a rapid require-
requirement despite using GRS derived tide in identifying type 1 diabetes (48) ment for insulin therapy, especially
from childhood-onset type 1 diabetes. and is well correlated with stimulated within 3 years of diagnosis. It should be
While GRS have the potential to assist C-peptide levels measured during a emphasized that among these features,
diagnosis of type 1 diabetes in uncertain mixed-meal tolerance test, which is con- age at diagnosis (<40 years), low BMI
cases, they are not yet widely available sidered the gold standard assessment of (<25 kg/m2), and rapid need for insulin
in clinical practice. Moreover, it is insulin secretory function in established therapy are the most discriminatory (43).
important to note that while the model type 1 diabetes (49). A recent analysis We recommend measurement of islet
was optimized with the inclusion of all found that concomitant blood glucose antibodies and C-peptide be considered
five variables, the addition of GRS had $144 mg/dL (8 mmol/L) increased the in all older people with clinical features
only a modest effect on overall model specificity of random C-peptide in pre- that suggest type 1 diabetes, with islet
performance (22). dicting a stimulated C-peptide level autoantibodies being the initial test of
Classification can be aided by the <600pmol/L, suggesting this is a rea- choice in short-duration disease (<3
measurement of autoantibodies and C- sonable threshold of blood glucose to years) and C-peptide the test of choice
peptide. Recommended autoantibodies employ for C-peptide interpretation at longer durations.
to assay at the time of diagnosis include (49).
those to insulin (insulin autoantibody), C-peptide also can be used to guide Diabetes-Associated Comorbidities
glutamate decarboxylase isoform 65 therapy (50). Individuals with a random and Complications
(GAD65A), insulinoma antigen 2, and C-peptide level #300 pmol/L should be DKA
zinc transporter isoform 8 (Znt8A), with managed mainly with insulin. For those The U.S. SEARCH for Diabetes in Youth
GAD65A being the most prevalent auto- with random C-peptide levels >300 study reported that nearly 30% of youth
antibody among adults. High levels or pmol/L, insulin could be combined with with newly diagnosed type 1 diabetes
the presence of more than one anti- other diabetes therapies, although evi- age <20 years presented with DKA (52).
body increases the likelihood of type 1 dence about safety and efficacy is limited. The frequency of DKA among adults at
diabetes. However, it is important to It is generally agreed that sulfonylureas diagnosis with type 1 diabetes is
realize that islet autoantibodies are a should be avoided because of the poten- unknown but is believed to be lower
continuous marker that can also occur tial to hasten b-cell failure (50). There is given that they often have higher C-
in the population without diabetes. As concern for increased risk of diabetic peptide levels at diagnosis and a slower
with many other tests, an abnormal test ketoacidosis (DKA) with sodium–glucose decline in b-cell function over time,
is usually based on a threshold signal cotransporter 1 (SGLT1) and SGLT2 inhibi- even in those requiring insulin initially
from control populations without diabe- tors when these agents are used in type (34). Among childhood-onset type 1 dia-
tes, usually the 97.5th or the 99th cen- 1 diabetes, especially in nonobese individ- betes, most episodes of DKA beyond
tile. Therefore, false-positive results with uals who may need only low dosages of diagnosis are associated with insulin
these assays can occur and can be insulin (51). All other agents could be omission, pump failure, or treatment
reduced by using higher-specificity assays considered for therapy in those not error (53). However, for adults with
or thresholds and targeting testing requiring insulin initially. In individuals type 1 diabetes, the primary risk factors
toward those with clinical features sug- with random C-peptide levels exceeding are noncompliance and infections (54),
gestive of type 1 diabetes (46). Finally, 600 pmol/L, management can be much the former sometimes due to the cost
since antibody levels can wane over time as recommended for type 2 diabetes, of insulin (55). Thus, there is a need to
in established type 1 diabetes, the with the caveats outlined above (50). An further understand DKA in adults, not
absence of autoantibodies does not rule important consideration is that loss of least because it is associated with long-
out the possibility of a diagnosis of type 1 b-cell function may be rapid in autoim- term worsening glycemic control (56).
diabetes. mune diabetes. As such, individuals
Measurement of C-peptide, paired treated without insulin should be closely Hypoglycemia
with a blood glucose in the same sam- monitored. Fear of hypoglycemia remains a major
ple, provides an estimate of endo-gen- In the absence of prospectively vali- problem in the clinical management
ous insulin production and has the most dated decision support tools that have of adults with type 1 diabetes (57),
care.diabetesjournals.org Leslie and Associates 2453

influencing quality of life and glycemic Factors that dictate use of these technol- sectional studies. In aggregate, these
control. The effect of diabetes duration or ogies are multiple and may include studies suggest that the prevalence of
age at diagnosis on hypoglycemia risk is reduced access to or acceptance of wear- nephropathy and retinopathy are lower
not consistent among different studies. able technology, challenges with insur- in adult-onset type 1 than in type 2 dia-
However, a-cell responses to hypoglyce- ance coverage, especially in the context betes, but this conclusion is potentially
mia and hypoglycemia risk are both lower of past misclassification, and/or inade- confounded by diabetes duration. For
in individuals with higher C-peptide levels quate education about hypoglycemia risk example, the prevalence of nephropathy
(38). Because residual C-peptide is more (58). A better understanding of potential and retinopathy was lower in Chinese
likely to be observed in those with a later barriers to technology use in adult-onset individuals with adult-onset type 1 diabe-
age of onset, hypoglycemia risk may be type 1 diabetes is needed. Furthermore, tes than in those with type 2, but only in
different between those with childhood- little is known about changes in hypogly- those with a disease duration <5 years,
and adult-onset diabetes. While insulin cemia risk across the life span of individu- while in the Botnia Study, retinopathy risk
pumps and continuous glucose monitors als with adult-onset disease, representing in adult-onset type 1 diabetes increased,
are associated with improved glycemic an important gap in knowledge. as expected, with disease duration (59).
control and reduced hypoglycemia (37), Two substantial prospective studies recently
adults may show reluctance or inertia in Microvascular and Macrovascular Disease reported that those adults with diabetes
adopting newer technologies. In the T1D Complications enrolled in the UKPDS who were also
Exchange study population, 63% of adults Despite the prevalence of adult-onset GADA positive (i.e., presumably with type 1
used an insulin pump while only 30% type 1 diabetes, there is a paucity of data diabetes) compared with those who were
used a continuous glucose monitor, and on the burden of microvascular complica- GADA negative (with type 2 diabetes)
use of these technologies tended to be tions in this population. Current knowl- showed a higher prevalence of retinopathy
lower in adults than in children (37). edge is largely based on small, cross- and lower prevalence of cardiovascular

Table 1—Knowledge gaps

Area of focus Description

Eliminating cultural bias in order to understand Most large-scale studies of adult type 1 diabetes have been done in
what impacts disease development Europe, North America, and China. There is a pressing need to
extend these studies to other continents and to diverse racial and
ethnic groups. Such studies could help us identify and understand
the nature and implications of diversity, whether in terms of
pathogenesis, cultural differences, or health care disparity. In
addition, prospective childhood studies of high-risk birth cohorts
could be extended into adulthood and new studies initiated to
better understand mechanisms behind disease development and
whether there is a differentiation in the disease process between
young and adult type 1 diabetes.
Population screening At present, universal childhood screening programs are being
developed in many countries. Research will be needed to develop
strategies for the follow-up of autoantibody-positive populations
throughout adulthood.
Disease-modifying therapies in early-stage disease Trials of disease-modifying therapies have generally shown better
efficacy in children (12). There are likely to be important differences
in agent selection between adult and pediatric populations, and
these differences require study.
Diagnosis and misclassification There is a need to build a diagnostic decision tree to aid in diabetes
classification. Tools are needed to estimate individual-level risk.
Adjunctive therapies There is a need to better understand the benefits and risks of using
therapies that are adjunctive to insulin in adult-onset type 1
diabetes. To this end, large-scale drug trials need to be performed,
and therapeutic decision trees are required to help health care
professionals and endocrinologists select such therapies.
Post-diagnosis education and support Improving education and support post-diagnosis is vital and should
include psychosocial support, health care provision, and analysis of
long-term outcomes (including complications) in adult-onset type 1
diabetes. Current knowledge is limited with respect to
complications, especially related to the complex mechanisms
contributing to macrovascular disease in adult-onset type 1
diabetes. Surveillance efforts based on larger and representative
cohorts of patients with clear and consistent case definitions are
needed to better understand the burden and risk of diabetes-
related chronic complications in this large population.
2454 Adult-Onset Type 1 Diabetes Diabetes Care Volume 44, November 2021

Whether these risks differ between those

diagnosed as children or as adults is
unclear and requires additional study.

CLOSING
In this Perspective, we have summarized
the current understanding of adult-onset
type 1 diabetes while identifying many
knowledge gaps (Table 1). Epidemiologi-
cal data from diverse ethnic groups
show that adult-onset type 1 diabetes is
often more prevalent than childhood-
onset type 1 diabetes. However, our
Figure 1—Proposed roadmap to better understand, diagnose, and care for adults with type 1 understanding of type 1 diabetes pre-
diabetes (T1D). Created in BioRender (BioRender.com). senting in adults is limited. This striking
shortfall in knowledge (Table 1) results
events (60,61). These results are consistent type 1 diabetes (65). In addition, there is in frequent misclassification, which may
with people with adult-onset type 1 diabe- the looming threat of complications, negatively impact disease management.
tes compared with those with type 2 diabe- including blindness and amputations (65). Here, we outline a roadmap for address-
tes, showing a general tendency to higher Adults with type 1 diabetes describe a ing these deficiencies (Fig. 1). A corner-
HbA1c levels (40,44,60,61) as well as stone of this roadmap is a renewed
sense of powerlessness, fear of hypogly-
reduced traditional cardiovascular risk fac- emphasis on the careful consideration of
cemia, and the challenges of both self-
tors, including reduced adiposity (BMI and the underlying etiology of diabetes in
management and appropriate food man-
waist circumference), metabolic (lipid lev- every adult presenting with diabetes.
agement (66). A common misunderstand-
els), and vascular (blood pressure) profiles In the absence of data-driven classifica-
ing is that while they face the same life tion tools capable of estimating individ-
(9,24,62). Nevertheless, all-cause mortality choices associated with type 2 diabetes
and cardiovascular mortality rates in such ual-level risk, we offer a simple set of
(e.g., weight loss, exercise, and limiting questions, incorporating what we have
individuals with adult-onset type 1 diabetes intake of simple sugars), adults with type
(59) are still higher than those among indi- termed the AABBCCs of diabetes classifi-
1 diabetes may require different manage- cation and management (Table 2). In par-
viduals without diabetes. In addition, there
ment skills (67). Moreover, there is a allel, we invite the research community
are discrepancies across studies, likely
strong association in adults with type 1 to join together in addressing key gaps in
related to differences in populations under
diabetes between chronic, stressful life knowledge through studies aimed at
study (i.e., age, race/ethnicity, and diabetes
events and fluctuating HbA1c, possibly defining the genetic, immunologic, and
duration), lack of consistent case definitions
due to indirect mechanisms, including metabolic phenotype of adult-onset type
(i.e., adult-onset type 1 diabetes or LADA
cases), and different outcomes, as well as adherence to diabetes management (68). 1 diabetes with the goal of using this
small sample sizes with insufficient events
on which to base strong recommendations. Table 2—AABBCC approach to diabetes classification
Parameter Description
Psychosocial Challenges Age Autoimmune diabetes is most prevalent in patients aged <50 years
Negative stressors, including pressure to at diagnosis. Those aged <35 years at diagnosis should be
achieve target HbA1c levels, lifestyle considered for maturity-onset diabetes of the young as well
considerations, and fear of complications, as type 1 diabetes
are factors leading to the increased fre- Autoimmunity Does this individual have islet autoantibodies or a history of
quency of mood disorders, attempted sui- autoimmunity (i.e., thyroid disease, celiac disease)? Is there
cide, and psychiatric care in adults with a goiter or vitiligo on exam?
diabetes (63). In individuals who have Body habitus/BMI Is the body habitus or BMI inconsistent with a diagnosis of type 2
experienced misclassification, additional diabetes, especially if BMI <25 kg/m2?
stress derives from conflicting messages Background What is the patient’s background? Is there a family history of
about the nature of their diabetes. autoimmunity and/or type 1 diabetes? Are they from a
Among adults with type 1 diabetes, those high-risk ethnic group?
with high psychological coping skills (e.g., Control Are diabetes control and HbA1c worsening on noninsulin therapies?
self-efficacy, self-esteem, and optimism) Has there been an accelerated change in HbA1c? Is the C-peptide
and adaptive skills may buffer the nega- low, that is, #300 pmol/L (especially <200 pmol/L), or is there
clinical evidence that b-cell function is declining? Was there a
tive effect of stress and should be culti- need for insulin therapy within 3 years of diabetes diagnosis?
vated (64). Relationship challenges,
Comorbidities Irrespective of immunogenetic background, coexistent cardiac or
including sexual intimacy, starting a fam-
renal disease and their risk factors impact the approach to
ily, caring for children, and relational therapy and HbA1c targets.
stress, are major stressors for adults with
care.diabetesjournals.org Leslie and Associates 2455

knowledge to develop improved approaches Nordisk, Sanofi, PhaseBio, Roche, AbbVie, Vascu- and risk of progression to diabetes in children.
for disease management and prevention lar Pharmaceuticals, Janssen, GlaxoSmithKline, JAMA 2013;309:2473–2479
and the Cystic Fibrosis Foundation. L.S.P. is also 12. Wherrett DK, Chiang JL, Delamater AM,
(Fig. 1). a cofounder and officer and board member and et al.; Type 1 Diabetes TrialNet Study Group.
stockholder for a company, Diasyst, Inc., that Defining pathways for development of disease-
markets software aimed to help improve diabe- modifying therapies in children with type 1
Acknowledgments. Sharon Saydah, Division tes management. No other potential conflicts of diabetes: a consensus report. Diabetes Care
of Diabetes Translation, National Center for interest relevant to this article were reported. 2015;38:1975–1985
Chronic Disease Prevention and Health Promo- The sponsors had no role in the design 13. Rolandsson O, Hampe CS, Sharp SJ, et al.
tion, Centers for Disease Control and Prevention, and conduct of the study, collection, manage- Autoimmunity plays a role in the onset of diabetes
attended the workshop and participated in sub- ment, analysis, and interpretation of the data, after 40 years of age. Diabetologia 2020;63:266–277
sequent discussions of the manuscript. Elizabeth and preparation, review, or approval of the 14. Mishra R, Åkerlund M, Cousminer DL, et al.
Seaquist, Division of Diabetes, Endocrinology, manuscript. This work is not intended to Genetic discrimination between LADA and
and Metabolism at the University of Minnesota, reflect the official opinion of the VA or the childhood-onset type 1 diabetes within the MHC.
participated in the workshop. The authors U.S. Government. Diabetes Care 2020;43:418–425
acknowledge Marilyn L. Wales for her assistance Author Contributions. R.D.L., C.E.M., J.F.-B., 15. Zhu M, Xu K, Chen Y, et al. Identification of
with formatting the manuscript. and J.L.D. conceived of the article and wrote novel T1D risk loci and their association with age
Funding and Duality of Interest. This manu- and edited the manuscript. All other authors and islet function at diagnosis in autoantibody-
script is the result of a one-day meeting held at were involved in the writing and editing of positive T1D individuals: based on a two-stage
JDRF headquarters in New York, NY. Financial the manuscript. R.D.L. and C.E.-M. are guaran- genome-wide association study. Diabetes Care
support for the workshop was provided by JDRF tors of this work and, as such, had full access 2019;42:1414–1421
and Janssen Research and Development, LLC. to all the data in the study and take responsi- 16. Buzzetti R, Di Pietro S, Giaccari A, et al.;
Financial support from Janssen Research and bility for the integrity of the data and the Non Insulin Requiring Autoimmune Diabetes
Development, LLC, for the workshop was in an accuracy of the data analysis. Study Group. High titer of autoantibodies to
unrestricted grant to JDRF. JDRF provided partici- GAD identifies a specific phenotype of adult-
pants with transportation, lodging, and meals to References onset autoimmune diabetes. Diabetes Care
attend the workshop. No additional support was 1. Mu~ noz C, Floreen A, Garey C, et al. 2007;30:932–938
provided for the writing of the manuscript. Misdiagnosis and diabetic ketoacidosis at diagnosis 17. Yasui J, Kawasaki E, Tanaka S, et al.; Japan
R.D.L. is supported by a grant from the Euro- of type 1 diabetes: patient and caregiver perspe- Diabetes Society Committee on Type 1 Diabetes
pean Union (contract no. QLGi-CT-2002-01886). ctives. Clin Diabetes 2019;37:276–281 Mellitus Research. Clinical and genetic characteristics
C.E.-M. is supported by National Institute for 2. Thomas NJ, Lynam AL, Hill AV, et al. Type 1 of non-insulin-requiring glutamic acid decarboxylase
Health Research grants R01 DK093954, R21DK11 diabetes defined by severe insulin deficiency (GAD) autoantibody-positive diabetes: a nationwide
9800, U01DK127786, R01DK127308, and P30DK occurs after 30 years of age and is commonly survey in Japan. PLoS One 2016;11:e0155643
097512; VA Merit Award I01BX001733; JDRF treated as type 2 diabetes. Diabetologia 2019;62: 18. Zhou Z, Xiang Y, Ji L, et al.; LADA China Study
grant 2-SRA-2019-834-S-B; and gifts from the 1167–1172 Group. Frequency, immunogenetics, and clinical
Sigma Beta Sorority, the Ball Brothers Founda- 3. Thunander M, Petersson C, Jonzon K, et al. characteristics of latent autoimmune diabetes in
tion, and the George and Frances Ball Founda- Incidence of type 1 and type 2 diabetes in adults China (LADA China study): a nationwide,
tion. R.B. is supported in part by the Italian and children in Kronoberg, Sweden. Diabetes Res multicenter, clinic-based cross-sectional study.
Ministry of University and Research (project Clin Pract 2008;82:247–255 Diabetes 2013;62:543–550
code 20175L9H7H). A.G.J. is funded by a 4. Rogers MAM, Kim C, Banerjee T, Lee JM. 19. Howson JM, Rosinger S, Smyth DJ, Boehm
National Institute for Health Research (NIHR) Fluctuations in the incidence of type 1 diabetes
BO; ADBW-END Study Group. Genetic analysis of
Clinician Scientist fellowship (CS-2015-15-018). in the United States from 2001 to 2015: a
adult-onset autoimmune diabetes. Diabetes
L.S.P. is supported in part by U.S. Department longitudinal study. BMC Med 2017;15:199
2011;60:2645–2653
of Veterans Affairs (VA) awards CSP #2008, I01 5. Weng J, Zhou Z, Guo L, et al.; T1D China Study
20. Battaglia M, Ahmed S, Anderson MS, et al.
CX001899, I01 CX001737, and Health Services Group. Incidence of type 1 diabetes in China, 2010-
Introducing the endotype concept to address the
Research & Development IIR 07-138; National 13: population based study. BMJ 2018;360:j5295
Institute for Health Research awards R21 challenge of disease heterogeneity in type 1
6. Thomas NJ, Jones SE, Weedon MN, Shields
DK099716, R18 DK066204, R03 AI133172, R21 diabetes. Diabetes Care 2020;43:5–12
BM, Oram RA, Hattersley AT. Frequency and
21. Redondo MJ, Yu L, Hawa M, et al.
AI156161, U01 DK091958, U01 DK098246, and phenotype of type 1 diabetes in the first six
UL1 TR002378; and Cystic Fibrosis Foundation Heterogeneity of type I diabetes: analysis of
decades of life: a cross-sectional, genetically
award PHILLI12A0. stratified survival analysis from UK Biobank. monozygotic twins in Great Britain and the
R.D.L. received unrestricted educational Lancet Diabetes Endocrinol 2018;6:122–129 United States. Diabetologia 2001;44:354–362
grants from Novo Nordisk, Sanofi, MSD, and 7. Diaz-Valencia PA, Bougneres P, Valleron AJ. 22. Lynam A, McDonald T, Hill A, et al.
AstraZeneca. C.E.-M. has participated in advi- Global epidemiology of type 1 diabetes in young Development and validation of multivariable
sory boards for Domp
e Pharmaceuticals, Pro- adults and adults: a systematic review. BMC clinical diagnostic models to identify type 1
vention Bio, MaiCell Technologies, and ISLA Public Health 2015;15:255 diabetes requiring rapid insulin therapy in adults
Technologies. C.E.M. is the recipient of in- 8. Gorham ED, Barrett-Connor E, Highfill-McRoy aged 18-50 years. BMJ Open 2019;9:e031586
kind research support from Nimbus Pharma- RM, et al. Incidence of insulin-requiring diabetes in 23. Cousminer DL, Ahlqvist E, Mishra R, et al.;
ceuticals and Bristol Myers Squibb and an the US military. Diabetologia 2009;52:2087–2091 Bone Mineral Density in Childhood Study. First
investigator-initiated research grant from Eli 9. Hawa MI, Kolb H, Schloot N, et al.; Action LADA genome-wide association study of latent
Lilly and Company. J.F.-B. and J.L.D. were Consortium. Adult-onset autoimmune diabetes in autoimmune diabetes in adults reveals novel
employed by JDRF during the workshop and Europe is prevalent with a broad clinical phenotype: insights linking immune and metabolic diabetes.
early stages of writing. J.F.-B. is currently an Action LADA 7. Diabetes Care 2013;36:908–913 Diabetes Care 2018;41:2396–2403
employee of Provention Bio, and J.L.D. is cur- 10. Xiang Y, Huang G, Zhu Y, et al.; China National 24. Zampetti S, Capizzi M, Spoletini M, et al.; NIRAD
rently an employee of Janssen Research and Diabetes and Metabolic Disorders Study Group. Study Group. GADA titer-related risk for organ-
Development, LLC. R.B. participated in advi- Identification of autoimmune type 1 diabetes and specific autoimmunity in LADA subjects subdivided
sory boards for Sanofi and Eli Lilly and multiple organ-specific autoantibodies in adult- according to gender (NIRAD study 6). J Clin
received honoraria for speaker bureaus from onset non-insulin-requiring diabetes in China: a Endocrinol Metab 2012;97:3759–3765
Sanofi, Eli Lilly, AstraZeneca, Novo Nordisk, population-based multicentre nationwide survey. 25. Balcha SA, Demisse AG, Mishra R, et al. Type
and Abbott. L.S.P. has served on scientific advi- Diabetes Obes Metab 2019;21:893–902 1 diabetes in Africa: an immunogenetic study in
sory boards for Janssen and has or had research 11. Ziegler AG, Rewers M, Simell O, et al. the Amhara of North-West Ethiopia. Diabetologia
support from Merck, Pfizer, Eli Lilly, Novo Seroconversion to multiple islet autoantibodies 2020;63:2158–2168
2456 Adult-Onset Type 1 Diabetes Diabetes Care Volume 44, November 2021

26. Jacobsen LM, Bocchino L, Evans-Molina C, duration of type 1 diabetes. Diabetes Care 2015; 55. Randall L, Begovic J, Hudson M, et al. Recurrent
et al. The risk of progression to type 1 diabetes is 38:476–481 diabetic ketoacidosis in inner-city minority patients:
highly variable in individuals with multiple 40. Andersen CD, Bennet L, Nystr€ om L, et al. behavioral, socioeconomic, and psychosocial factors.
autoantibodies following screening. Diabetologia Worse glycaemic control in LADA patients than in Diabetes Care 2011;34:1891–1896
2020;63:588–596 those with type 2 diabetes, despite a longer time 56. Duca LM, Reboussin BA, Pihoker C, et al.
27. Long AE, Wilson IV, Becker DJ, et al. on insulin therapy. Diabetologia 2013;56:252–258 Diabetic ketoacidosis at diagnosis of type 1
Characteristics of slow progression to diabetes in 41. Sims EK, Chaudhry Z, Watkins R, et al. diabetes and glycemic control over time: the
multiple islet autoantibody-positive individuals Elevations in the fasting serum proinsulin-to-C- SEARCH for diabetes in youth study. Pediatr
from five longitudinal cohorts: the SNAIL study. peptide ratio precede the onset of type 1 Diabetes 2019;20:172–179
Diabetologia 2018;61:1484–1490 diabetes. Diabetes Care 2016;39:1519–1526 57. Martyn-Nemeth P, Schwarz Farabi S, Mihailescu
28. Hanna SJ, Powell WE, Long AE, et al. Slow 42. Sims EK, Bahnson HT, Nyalwidhe J, et al.; D, Nemeth J, Quinn L. Fear of hypoglycemia in adults
progressors to type 1 diabetes lose islet T1D Exchange Residual C-Peptide Study Group. with type 1 diabetes: impact of therapeutic advances
autoantibodies over time, have few islet antigen- Proinsulin secretion is a persistent feature of type and strategies for prevention – a review. J Diabetes
specific CD81 T cells and exhibit a distinct CD95hi B 1 diabetes. Diabetes Care 2019;42:258–264
Complications 2016;30:167–177
cell phenotype. Diabetologia 2020;63:1174–1185 43. Shields BM, Peters JL, Cooper C, et al. Can
58. Engler R, Routh TL, Lucisano JY. Adoption
29. Schloot NC, Pham MN, Hawa MI, et al.; Action clinical features be used to differentiate type 1
barriers for continuous glucose monitoring and
LADA Group. Inverse relationship between organ- from type 2 diabetes? A systematic review of the
their potential reduction with a fully implanted
specific autoantibodies and systemic immune literature. BMJ Open 2015;5:e009088
system: results from patient preference surveys.
mediators in type 1 diabetes and type 2 diabetes: 44. Ahlqvist E, Storm P, K€ar€aj€am€aki A, et al.
Action LADA 11. Diabetes Care 2016;39:1932–1939 Novel subgroups of adult-onset diabetes and Clin Diabetes 2018;36:50–58
30. Brooks-Worrell BM, Palmer JP. Setting the their association with outcomes: a data-driven 59. Isomaa B, Almgren P, Henricsson M, et al.
stage for islet autoimmunity in type 2 diabetes: cluster analysis of six variables. Lancet Diabetes Chronic complications in patients with slowly
obesity-associated chronic systemic inflammation Endocrinol 2018;6:361–369 progressing autoimmune type 1 diabetes (LADA).
and endoplasmic reticulum (ER) stress. Diabetes 45. Hawa MI, Thivolet C, Mauricio D, et al.; Diabetes Care 1999;22:1347–1353
Care 2019;42:2338–2346 Action LADA Group. Metabolic syndrome and 60. Maddaloni E, Coleman RL, Agbaje O, Buzzetti R,
31. Campbell-Thompson M, Fu A, Kaddis JS, et al. autoimmune diabetes: action LADA 3. Diabetes Holman RR. Time-varying risk of microvascular
Insulitis and b-cell mass in the natural history of Care 2009;32:160–164 complications in latent autoimmune diabetes of
type 1 diabetes. Diabetes 2016;65:719–731 46. Jones AG, McDonald TJ, Shields BM, adulthood compared with type 2 diabetes in adults:
32. Leete P, Willcox A, Krogvold L, et al. Hagopian W, Hattersley AT. Latent autoimmune a post-hoc analysis of the UK Prospective Diabetes
Differential insulitic profiles determine the extent diabetes of adults (LADA) is likely to represent a Study 30-year follow-up data (UKPDS 86). Lancet
of b-cell destruction and the age at onset of type mixed population of autoimmune (type 1) and Diabetes Endocrinol 2020;8:206–215
1 diabetes. Diabetes 2016;65:1362–1369 nonautoimmune (type 2) diabetes. Diabetes Care 61. Maddaloni E, Coleman RL, Pozzilli P, Holman
33. Fleiner HF, Bjøro T, Midthjell K, Grill V, 2021dc202834 RR. Long-term risk of cardiovascular disease in
Åsvold BO. Prevalence of thyroid dysfunction 47. Jones AG, Hattersley AT. The clinical utility of individuals with latent autoimmune diabetes in
in autoimmune and type 2 diabetes: the C-peptide measurement in the care of patients adults (UKPDS 85). Diabetes Obes Metab
population-based HUNT study in Norway. J Clin with diabetes. Diabet Med 2013;30:803–817 2019;21:2115–2122
Endocrinol Metab 2016;101:669–677 48. Berger B, Stenstr€ om G, Sundkvist G. Random 62. Tuomi T, Carlsson A, Li H, et al. Clinical and
34. Greenbaum CJ, Beam CA, Boulware D, et al.; C-peptide in the classification of diabetes. Scand J genetic characteristics of type 2 diabetes with and
Type 1 Diabetes TrialNet Study Group. Fall in C- Clin Lab Invest 2000;60:687–693 without GAD antibodies. Diabetes 1999;48:150–157
peptide during first 2 years from diagnosis: 49. Hope SV, Knight BA, Shields BM, Hattersley 63. Robinson ME, Simard M, Larocque I, Shah J,
evidence of at least two distinct phases from AT, McDonald TJ, Jones AG. Random non-fasting Nakhla M, Rahme E. Risk of psychiatric disorders
composite Type 1 Diabetes TrialNet data. C-peptide: bringing robust assessment of endogenous and suicide attempts in emerging adults with
Diabetes 2012;61:2066–2073 insulin secretion to the clinic. Diabet Med 2016; diabetes. Diabetes Care 2020;43:484–486
35. Gong S, Wu C, Zhong T, et al. Complicated 33:1554–1558 64. Yi JP, Vitaliano PP, Smith RE, Yi JC, Weinger K.
curve association of body weight at diagnosis 50. Buzzetti R, Tuomi T, Mauricio D, et al. The role of resilience on psychological adjustment
with C-peptide in children and adults with new- Management of latent autoimmune diabetes in
and physical health in patients with diabetes. Br J
onset type 1 diabetes. Diabetes Metab Res Rev adults: a consensus statement from an international
Health Psychol 2008;13:311–325
2020;36:e3285 expert panel. Diabetes 2020;69:2037–2047
65. Trief PM, Sandberg JG, Dimmock JA, Forken PJ,
36. Hao W, Gitelman S, DiMeglio LA, Boulware D; 51. Danne T, Garg S, Peters AL, et al. International
Weinstock RS. Personal and relationship challenges
Type 1 Diabetes TrialNet Study Group. Fall in c- consensus on risk management of diabetic keto-
of adults with type 1 diabetes: a qualitative focus
peptide during first 4 years from diagnosis of type 1 acidosis in patients with type 1 diabetes treated
diabetes: variable relation to age, HbA1c, and with sodium-glucose cotransporter (SGLT) inhibitors. group study. Diabetes Care 2013;36:2483–2488
insulin dose. Diabetes Care 2016;39:1664–1670 Diabetes Care 2019;42:1147–1154 66. Fisher L, Polonsky WH, Hessler DM, et al.
37. Foster NC, Beck RW, Miller KM, et al. State of 52. Dabelea D, Rewers A, Stafford JM, et al.; Understanding the sources of diabetes distress in
type 1 diabetes management and outcomes SEARCH for Diabetes in Youth Study Group. adults with type 1 diabetes. J Diabetes
from the T1D exchange in 2016-2018. Diabetes Trends in the prevalence of ketoacidosis at Complications 2015;29:572–577
Technol Ther 2019;21:66–72 diabetes diagnosis: the SEARCH for diabetes in 67. Hilliard ME, Yi-Frazier JP, Hessler D, Butler
38. Rickels MR, Evans-Molina C, Bahnson HT, youth study. Pediatrics 2014;133:e938–e945 AM, Anderson BJ, Jaser S. Stress and A1c among
et al.; T1D Exchange b-Cell Function Study 53. Rewers A. Acute metabolic complications in people with diabetes across the lifespan. Curr
Group. High residual C-peptide likely contributes diabetes. In Diabetes in America. 3rd ed. Diab Rep 2016;16:67
to glycemic control in type 1 diabetes. J Clin Bethesda, MD, National Institute of Diabetes and 68. Lloyd CE, Dyer PH, Lancashire RJ, Harris T,
Invest 2020;130:1850–1862 Digestive and Kidney Diseases, 2018; pp. 1–19 Daniels JE, Barnett AH. Association between
39. Davis AK, DuBose SN, Haller MJ, et al.; T1D 54. Umpierrez GE, Kitabchi AE. Diabetic ketoac- stress and glycemic control in adults with type 1
Exchange Clinic Network. Prevalence of detectable idosis: risk factors and management strategies. (insulin-dependent) diabetes. Diabetes Care
C-peptide according to age at diagnosis and Treat Endocrinol 2003;2:95–108 1999;22:1278–1283