Article Text

Download PDFPDF

Precision Medicine to Redefine Insulin Secretion and Monogenic Diabetes-Randomized Controlled Trial (PRISM-RCT) in Chinese patients with young-onset diabetes: design, methods and baseline characteristics
  1. Chun Kwan O1,
  2. Ying Nan Fan1,2,
  3. Baoqi Fan1,2,
  4. Cadmon Lim1,2,
  5. Eric S H Lau1,2,
  6. Sandra T F Tsoi1,2,
  7. Raymond Wan1,
  8. Wai Yin Lai1,
  9. Emily WM Poon1,2,
  10. Jane Ho1,
  11. Cherry Cheuk Yee Ho1,
  12. Chloe Fung1,
  13. Eric KP Lee3,
  14. Samuel YS Wong3,
  15. Maggie Wang3,
  16. Risa Ozaki1,
  17. Elaine Cheung1,
  18. Ronald Ching Wan Ma1,2,
  19. Elaine Chow1,2,
  20. Alice Pik Shan Kong1,2,
  21. Andrea Luk1,2,
  22. Juliana C N Chan1,2
  1. 1Department of Medicine and Therapeutics, The Chinese University of Hong Kong, Hong Kong Special Administrative Region, People's Republic of China
  2. 2Li Ka Shing Institute of Health Sciences, The Chinese University of Hong Kong, Hong Kong Special Administrative Region, People's Republic of China
  3. 3JC School of Public Health & Primary Care, The Chinese University of Hong Kong Faculty of Medicine, Hong Kong Special Administrative Region, People's Republic of China
  1. Correspondence to Professor Juliana C N Chan; jchan{at}


Introduction We designed and implemented a patient-centered, data-driven, holistic care model with evaluation of its impacts on clinical outcomes in patients with young-onset type 2 diabetes (T2D) for which there is a lack of evidence-based practice guidelines.

Research design and methods In this 3-year Precision Medicine to Redefine Insulin Secretion and Monogenic Diabetes-Randomized Controlled Trial, we evaluate the effects of a multicomponent care model integrating use of information and communication technology (Joint Asia Diabetes Evaluation (JADE) platform), biogenetic markers and patient-reported outcome measures in patients with T2D diagnosed at ≤40 years of age and aged ≤50 years. The JADE-PRISM group received 1 year of specialist-led team-based management using treatment algorithms guided by biogenetic markers (genome-wide single-nucleotide polymorphism arrays, exome-sequencing of 34 monogenic diabetes genes, C-peptide, autoantibodies) to achieve multiple treatment goals (glycated hemoglobin (HbA1c) <6.2%, blood pressure <120/75 mm Hg, low-density lipoprotein-cholesterol <1.2 mmol/L, waist circumference <80 cm (women) or <85 cm (men)) in a diabetes center setting versus usual care (JADE-only). The primary outcome is incidence of all diabetes-related complications.

Results In 2020–2021, 884 patients (56.6% men, median (IQR) diabetes duration: 7 (3–12) years, current/ex-smokers: 32.5%, body mass index: 28.40±5.77 kg/m2, HbA1c: 7.52%±1.66%, insulin-treated: 27.7%) were assigned to JADE-only (n=443) or JADE-PRISM group (n=441). The profiles of the whole group included positive family history (74.7%), general obesity (51.4%), central obesity (79.2%), hypertension (66.7%), dyslipidemia (76.4%), albuminuria (35.4%), estimated glomerular filtration rate <60 mL/min/1.73 m2 (4.0%), retinopathy (13.8%), atherosclerotic cardiovascular disease (5.2%), cancer (3.1%), emotional distress (26%–38%) and suboptimal adherence (54%) with 5-item EuroQol for Quality of Life index of 0.88 (0.87–0.96). Overall, 13.7% attained ≥3 metabolic targets defined in secondary outcomes. In the JADE-PRISM group, 4.5% had pathogenic/likely pathogenic variants of monogenic diabetes genes; 5% had autoantibodies and 8.4% had fasting C-peptide <0.2 nmol/L. Other significant events included low/large birth weight (33.4%), childhood obesity (50.7%), mental illness (10.3%) and previous suicide attempts (3.6%). Among the women, 17.3% had polycystic ovary syndrome, 44.8% required insulin treatment during pregnancy and 17.3% experienced adverse pregnancy outcomes.

Conclusions Young-onset diabetes is characterized by complex etiologies with comorbidities including mental illness and lifecourse events.

Trial registration number NCT04049149.

  • genetics
  • MODY
  • randomized controlled trials as topic
  • latent autoimmune diabetes in adults

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


  • Young-onset diabetes is associated with increased lifetime risk of diabetic complications, mental illness, hospitalization and premature mortality; however, there is a lack of evidence-based practice guidelines for these high-risk patients.


  • The 3-year Precision Medicine to Redefine Insulin Secretion and Monogenic Diabetes-Randomized Controlled Trial is the first large-scale study to evaluate the effects of a multicomponent care model integrating the use of information and communication technology (JADE platform), biogenetic markers and patient-reported outcome measures in pursuit of precision care to achieve strict metabolic goals for reducing all diabetes-related clinical events in patients with young-onset type 2 diabetes.

  • Pending the primary study outcomes, analysis of the baseline data revealed complexity in etiologies and phenotypes characterized by significant lifecourse events, a high burden of mental illness, psychological distress, comorbidities and suboptimal control of cardiovascular risk factors in these young patients.


  • Results of this randomized controlled trial will provide important insights on the clinical effectiveness and cost-effectiveness of the demonstrated innovative care model for patients with young-onset diabetes.

  • The ultimate goal is to inform practice and policies in pursuit of value-based care with precision, evidence and quality.


In clinic-based settings, one in five Asians with type 2 diabetes (T2D) are diagnosed before the age of 40 years.1 Genetic and perinatal factors contribute to one’s predisposition to lifetime risk of diabetes, further influenced by autoimmunity and other precipitating factors such as medical illnesses, lifecourse events and psychosocial-behavioral factors. Collectively, these factors influence the age of onset of diabetes.2 The long disease duration puts these young patients at high risk of poor quality of life, multiple morbidities, recurrent hospitalizations and premature mortality, which can be modified by control of risk factors, medications and self-management with ongoing support.3–6 In contrast to declining trends of complications and death in older adults with T2D, population-level studies reported stagnation or even resurgence in incidence rates of cardiovascular disease and lower extremity amputation in patients with T2D in the youngest age group.7 Despite this disproportionate burden in young patients with T2D, there is a lack of evidence-based practice guidelines in these patients who were under-represented in most clinical trials.2 8 9

Current clinical practice in these young patients is largely based on evidence extrapolated from studies of patients with later-onset T2D.8 9 However, young-onset T2D differs from later-onset T2D in many aspects including etiologies, pathophysiology, clinical phenotypes and responses to treatment.1 3 8 10 In multiple large cohort studies, patients with young-onset T2D, arbitrarily defined as age of diagnosis ≤40 years, had higher glycated hemoglobin (HbA1c) despite higher use of insulin compared with patients with later-onset T2D.1 3 This poor control may reflect the complex etiologies, lack of evidence to guide treatment, suboptimal self-management and other unmet needs calling for redefining the standard of management in these young patients. These include accurate assignment of diabetes subtypes, detailed phenotypic characterization and psychosocial-behavioral assessment to enable personalized pharmacotherapy with patient empowerment and engagement. To address these multiple goals, the practice environment needs to be conducive to using a multidisciplinary team to define etiologies, promote patient-provider communication and increase the precision of diagnosis and management.11

Based on these premises, we undertook a 3-year randomized controlled trial (RCT) where we implemented a multicomponent care model led by endocrinologists in an ambulatory diabetes center setting in Hong Kong Chinese patients with young-onset non-type 1 diabetes. The care model was enhanced by use of biogenetic markers and the web-based Joint Asia Diabetes Evaluation (JADE) technology with built-in protocol for structured assessment for risk stratification and issue of personalized report aimed at attaining multiple cardiometabolic targets and improving clinical outcomes. This paper presents the protocol of the Precision Medicine to Redefine Insulin Secretion and Monogenic Diabetes-RCT (PRISM-RCT) and describes the clinical characteristics of the 884 patients recruited.

Research design and methods

Study design and participants

The PRISM-RCT is a single-center, open-label, two-arm parallel design study conducted at the Diabetes and Endocrine Research Center (DERC) at the Prince of Wales Hospital, the teaching hospital of the Chinese University of Hong Kong (CUHK). Hong Kong has a heavily subsidized healthcare system with the government-funded hospital authority (HA) operating all public hospitals and clinics, where patients only pay nominal fees for healthcare services. We identified eligible patients from our Diabetes Register, family medicine clinics, community-based general outpatient clinics and hospital-based diabetes and other medical clinics across New Territories East Cluster of HA with a catchment population of 1 million. Inclusion criteria included Chinese ethnicity, age 18–50 years, age at diabetes diagnosis ≤40 years and willingness to provide written informed consent. Exclusion criteria included type 1 diabetes defined by presentation with diabetic ketoacidosis or continuous insulin requirement within 12 months of diagnosis, reduced life expectancy due to terminal illness or otherwise deemed not appropriate per discretion of the investigator. The study is registered at (NCT04049149).

Randomization and blinding

Eligible patients were randomized in 1:1 ratio to receive either JADE-only or JADE-PRISM care. Computer-generated assignment codes were contained in individually sealed, opaque and consecutively numbered envelopes, which were then opened by non-study personnel. Patients, investigators and nurses were not blinded, given the study design.

Study procedures

JADE-assisted structured assessments

The web-based JADE portal, developed by the Asia Diabetes Foundation, governed by the CUHK Foundation, contains templates that guide structured assessment with built-in risk engines and reporting system. Sociodemographic characteristics, lifestyle, family history, medical history and medications were documented using a standardized form. Vital signs (blood pressure (BP), pulse rate) and anthropometric parameters (body weight (BW), body height, waist circumference (WC)) were measured. Fasting blood samples for HbA1c, plasma glucose, lipid, renal function, liver function, complete blood count and spot urine samples for albumin-to-creatinine ratio (ACR) were collected to define cardiovascular-kidney-metabolic risk factors (online supplemental table S1). Fundus and lower limbs were evaluated for retinopathy and sensory neuropathy. These data were entered into the JADE portal which generates a personalized clinical report displaying the risk category, 5-year probabilities of major clinical events, trends of attained and recommended values of HbA1c, BP, low-density lipoprotein (LDL)-cholesterol and BW, with individualized decision support and self-management recommendation.12

Supplemental material

Assessment for biogenetic markers and patient-reported outcome measures

Additional blood samples were taken for fasting C-peptide (CP), glutamic acid decarboxylase antibodies (GADA) and biogenetic tests including targeted sequencing for 34 monogenic diabetes genes (MDGs), quantitative PCR for one hotspot variant (mtA3243G) and whole-genome single-nucleotide polymorphism analysis using the Asian screening array (online supplemental tables S2–S4). Apart from using these data to classify diabetes subtypes, these genetic variants were used to generate risk scores for predicting risk of cardiovascular-kidney complications and glycemic trajectory, which were used to empower patients for improving self-management.13–15 Patient-reported outcome measures (PROMs) were completed by all patients using validated questionnaires including 5-item EuroQol for Quality of Life (EQ-5D),16 21-item Depression Anxiety Stress Scale for Negative Emotions (DASS-21),17 9-item Patient Health Questionnaire for Depression (PHQ-9),18 Chinese 15-item Diabetes Distress Scale (CDDS-15),19 20-item Diabetes Empowerment Scale for Self-efficacy,20 15-item Summary for Diabetes Self-care Activities21 and 4-item Compliance Questionnaire for Medication.22 Chinese-validated cut-offs were used to define psychological distress (DASS-21 score ≥17,23 PHQ-9 score ≥718 and CDDS-15 score ≥45).19

Follow-up care

All randomized patients were given their personalized JADE report and introduced to the mobile application for self-tracking of BW, BP and blood glucose. After undergoing the JADE-guided assessment, the JADE-only group continued to receive care at their usual clinics. All patients will return to DERC for BP and BW measurement, blood and urine tests to assess control of risk factors and progression of CP and completion of questionnaires annually. The JADE assessment will be repeated at year 3 for ascertainment of predefined clinical outcomes.

JADE-PRISM intervention group

The JADE-PRISM group was transferred from their usual clinics to the DERC for 1-year intensive intervention and then referred back to the original or diabetes clinic for follow-up in the following 2 years. During the 1-year intensive intervention, the endocrinologists were supported by nurses and research assistants who provided clerical support and reminders to patients. The JADE report was explained in group by CUHK research nurses at month 1 from randomization. They were then reviewed by the CUHK endocrinologist-nurse team at months 3, 6, 9 and 12. The endocrinologists individually explained the results of all biogenetic markers and adjusted treatment based on an algorithm which considered BMI, GADA status, fasting CP and mutations in MDGs (figure 1). Both RCTs and meta-analyses indicated Asians and patients with low BMI were particularly responsive to incretin-based therapies.24 25 Thus, we recommended dipeptidyl peptidase-4 inhibitors (DPP-4i) in normal-weight patients and glucagon-like peptide-1 receptor agonist (GLP-1 RA) in those with overweight/obesity. In patients with complications, sodium-glucose cotransporter-2 inhibitors (SGLT2i) and/or GLP-1 RA were used for organ protection. Patients with GADA positivity or fasting CP <0.2 nmol/L were advised to initiate insulin therapy. Low-dose sulfonylureas were preferred in patients with hepatic nuclear factor (HNF)1A-maturity-onset diabetes of the young (MODY) and HNF4A-MODY, while insulin is preferred in patients with HNF1B-MODY.

Figure 1

The treatment algorithm for patients randomized to the Joint Asia Diabetes Evaluation-Precision Medicine to Redefine Insulin Secretion and Monogenic Diabetes group who were given biogenetic information to guide personalized therapies to achieve multiple strict metabolic goals during the 1-year intensive intervention at Diabetes and Endocrine Research Center. ASA, Asian specific array; BMI, body mass index; CP, C-peptide; DPP4i, dipeptidyl peptidase-4 inhibitor; GADA, glutamic acid decarboxylase antibody; GRS, genetic riskscore; GLP-1 RA, glucagon-like peptide-1 receptor agonist; HNF, Hepatic nuclear factor; LDL-cholesterol, low-density lipoprotein-cholesterol; OAD, oral anti-diabetic drugs; SGLT2i, sodium-glucose co-transporter 2 inhibitor; SNP, single-nucleotide polymorphism.

The endocrinologist also reviewed the PROMs, obtained detailed medical history and lifecourse events and used this information to counsel the patients regarding possible causes of their young-onset diabetes (YOD) while adjusting and intensifying treatment to reach predefined treatment goals (HbA1c <6.2%, BP <120/75 mm Hg, LDL-cholesterol <1.2 mmol/L, WC <80 cm in women and <85 cm in men) within 12 months before referral back to original clinic (figure 1). The team collectively emphasized to patients the benefits of early intervention for long-term gain. All patients were given additional education materials including online resources with counseling by nurses either onsite or through WhatsApp messages/phone calls. Between visits, patients might receive reminders by the supporting team depending on control. Patients with severe obesity were given additional dietary counseling and referred for metabolic surgery if their BMI was >30 kg/m2. Patients with persistently poor glycemic control were encouraged to wear continuous glucose monitoring system which was not subsidized in our setting. After the first year, patients were referred back to their original clinics for routine follow-up. Annually, up to study end at year 3, patients would be reviewed by the endocrinologists with comments to the usual care doctors on optimisation of glycemic and metabolic risk factors, as appropriate.

JADE-control group

Patients assigned to the JADE-only group received group explanation of JADE report by nurses. They would attend usual clinic for ongoing treatment with yearly return to the DERC for assessment. The results of biogenetic markers will be released and explained to the patients on study completion at year 3.


All patients returned at year 3 to repeat the JADE assessment for outcome ascertainment. The primary outcome was defined as any diabetes-related microvascular and macrovascular end points including incident atherosclerotic-cardiovascular disease (ASCVD including coronary artery disease, congestive heart failure, stroke and peripheral artery disease), incidence or progression of chronic kidney disease (CKD), all-cause death, incidence or progression of albuminuria or retinopathy, incident visual impairment and incident sensory neuropathy. Detailed description of these outcomes is available in online supplemental table S5.

The secondary outcomes included attainment of ≥3 metabolic targets (HbA1c <6.2%, BP <120/80 mm Hg, LDL-cholesterol <1.8 mmol/L, triglyceride <1.2 mmol/L, WC <80 cm in women and <85 cm in men); attainment of ≥2 key performance indexes (reductions in HbA1c by ≥0.5%, systolic BP by ≥5 mm Hg, LDL-cholesterol by ≥0.5 mmol/L and BW by ≥3%); severe hypoglycemia events requiring hospital admission; on-treatment changes in cardiometabolic risk factors, trajectories in fasting CP, use of medications and psychological-behavioral measures.

Sample size and statistical analysis

We assumed a 6.0% annual rate of primary outcome in Hong Kong Chinese with YOD and hypothesized 40% risk reduction for the primary outcome in the JADE-PRISM group compared with the JADE-only group.26 Assuming 10% attrition rate, 430 patients in each arm gave 80% power at an α level of 0.05 for a 3-year intervention with 1:1 assignment. Analysis of baseline characteristics was performed using R software V.4.1.1 expressed as mean±SD, median (IQR) or number (percentages). Fisher’s exact test, χ2 test, t-test and Mann-Whitney U test were used for group comparison, as appropriate. A two-sided p value <0.05 was considered statistically significant.


Between 2020 and 2021, 1314 letters were sent to patients aged under 50 years enrolled in the JADE-Register and 146 patients were referred by physicians from various clinics. Among them, 884 patients with young-onset non-type 1 diabetes participated with the first patient randomized on January 14, 2020 and the last patient on September 28, 2021 (figure 2). 443 patients were assigned to JADE-only group and 441 patients were assigned to JADE-PRISM group. Just over half of the participants were men, with median age at diabetes diagnosis of 34 (29–38) years and diabetes duration of 7 (3–12) years at randomization (table 1). In the whole group, 74.7% reported positive family history of diabetes (19.7% had both parents affected), 51.4% had general obesity and 79.2% had central obesity. Additionally, 17.9% reported regular exercise, 35.2% adhered to a balanced diet and 32.5% were current/ex-smokers. More than one-third had concomitant hypertension (66.7%), dyslipidemia (76.4%) and albuminuria (35.4%) and as many as one in four patients had comorbidities including CKD (4.0%), retinopathy (13.8%), sensory neuropathy (0.3%), ASCVD (5.2%) and cancer (3.1%). 21.5% had elevated alanine transaminase. Up to 96.2% were taking glucose-lowering drugs (GLDs) with 27.7% on insulin and 48.7% taking three or more drugs. Around 50% received statins and renin-angiotensin system inhibitor (RASi). Among those with albuminuria, 71.4% were prescribed RASi. Overall, 14.3% attained all three conventional goals of HbA1c <7%, BP <130/80 mm Hg and LDL-cholesterol <2.6 mmol/L (or <1.8 mmol/L in presence of ASCVD); 13.7% attained ≥3 stricter metabolic targets defined by the secondary outcomes (HbA1c <6.2%, BP <120/80 mm Hg, LDL-cholesterol <1.8 mmol/L, triglyceride <1.2 mmol/L, WC <80 cm in women and <85 cm in men) of this trial. The median EQ-5D index value for quality of life ranging from 1 (full health) to 0 (death) was 0.88 (0.87–0.96). Using Chinese-validated cut-offs, 26%–38% of participants had emotional distress based on several PROMs. Using structured questionnaire, more than half of participants adhered to recommended diet and medication plans for more than half of the time in a week, and 56.8% had high medication adherence. The self-care efficacy in exercise, self-monitoring of blood glucose (SMBG) and foot care was suboptimal.

Figure 2

Graphical representation of the Precision Medicine to Redefine Insulin Secretion and Monogenic Diabetes-Randomized Controlled Trial (PRISM-RCT) study. JADE, Joint Asia Diabetes Evaluation; GADA, glutamic acid decarboxylase antibodies; GRS, genetic risk score; YOD, young-onset diabetes; HbA1c, glycated hemoglobin; BP, blood pressure; LDL-C, low-density lipoprotein cholesterol; WC, waist circumference; ASCVD, atherosclerotic cardiovascular disease; DERC, Diabetes and Endocrine Research Center. Constructed by Microsoft PowerPoint, graphics are either original or from Microsoft Office stock.

Table 1

Baseline characteristics of participants with young-onset diabetes in the PRISM-RCT

In the JADE-PRISM group (n=441), additional information was obtained through medical consultations (table 2). The mean age of diabetes diagnosis in affected family members ranged from 37 to 52 years. 11.1% of the participants were born pre-term or post-term and 33.4% had low or large birth weight. Half (50.7%) reported childhood obesity (top 10% in the class) with a mean peak BMI of 31.1 kg/m2 at a mean age of 29.4 years. 10.3% reported history of mental illness and 3.6% had previous suicide attempts. Among women, 17.3% had polycystic ovary syndrome, 44.8% required insulin treatment during pregnancy and 17.3% had adverse pregnancy outcomes. In patients with available biogenetic markers, 4.5% had pathogenic/likely pathogenic (P/LP) mutations of MDGs, 5.0% had GADA positivity and 8.6% had fasting CP <0.2 nmol/L. Using the Homeostasis Model Assessment model (HOMA2) (, the median HOMA2-beta-cell function (%) was 51.9 (35.1–85.0) and HOMA2-IR was 1.61 (1.12–2.27).

Table 2

Additional medical history and risk factors obtained through direct interview by endocrinologists and biogenetic markers shared with the JADE-PRISM group

The JADE-PRISM and JADE-only groups were generally balanced in baseline characteristics, except the JADE-PRISM group had a lower prevalence of HbA1c <6.2% (12.2% vs 19.0%, p=0.008), self-reported hypoglycemia in last 3 months (23.0% vs 29.9%, p=0.026), hypertension (63.0% vs 70.4%, p=0.025), use of BP-lowering drugs (55.8% vs 63.7%, p=0.020) and CKD (2.0% vs 5.9%, p=0.006); and a higher prevalence of cancer history (4.8% vs 1.4%, p=0.006).


The 3-year PRISM study is the first large-scale RCT to evaluate the effects of a multicomponent care model integrating use of information and communication technology (JADE platform), biogenetic markers and PROMs in pursuit of precision care to achieve strict metabolic targets for reducing all diabetes-related clinical events in patients with young-onset T2D. In public healthcare settings, the majority of these young patients are managed in busy clinics by non-endocrinologists with short consultation time, insufficient risk-profiling and poor engagement. We changed the setting, providers and procedures where the JADE-PRISM group was managed in a day-center by an endocrinologist-led multidisciplinary team for 1 year. Pending the outcomes at 3 years when the study will end in November 2024, analysis of our baseline data revealed complexity in etiologies and phenotypes characterized by significant lifecourse events, high burden of mental illness, psychological distress, comorbidities and suboptimal control of cardiovascular risk factors in these young patients.

Obesity as a key feature in the cohort

Young-onset T2D in Asians is characterized by low beta-cell reserve and accelerated loss of beta-cell function.3 10 In a population-based Korean study, while insulin resistance increased with age, individuals who could not increase their beta-cell capacity went on to develop diabetes.27 Globally including in Asia, childhood obesity and its tracking into early adulthood is a key driver for the rising incidence of YOD.28 In Hong Kong, 32% of those aged 15–24 years were overweight (BMI 23.0–24.9 kg/m2) or obese (BMI ≥25.0 kg/m2). The proportions of overweight/obesity increased from 44% in the 25–34 years age group to 56% in the 35–44 years age group.29 In our earlier report of the JADE-Asia cohort including 41 029 patients with T2D across 9 Asian regions, 1 in 5 had YOD who were more likely to have general and central obesity than their peers with later-onset T2D.1 Consistent with these observations, 50%–80% of our cohort had general or central obesity with half of them reporting a history of childhood/adolescent obesity. Compared with a mean BMI of 25.7 kg/m2 in an earlier cohort of Hong Kong Chinese with young-onset T2D (1994–2012), the mean BMI was 28.4 kg/m2 in the present cohort.4

The high prevalence of central obesity in this cohort is concerning. Visceral fat deposition in liver and pancreas contributes to hepatic insulin resistance and defective beta-cell function, respectively. In our cohort, the majority of patients had coexisting hypertension, dyslipidemia and/or albuminuria, which portends development of ASCVD and CKD. In a 10-year prospective cohort of 2323 patients with young-onset T2D, we reported overweight patients had 5-fold and 15-fold higher risk of end-stage kidney disease and ASCVD, respectively, compared with patients with type 1 diabetes. However, the risk differentials were rendered non-significant after adjusting for BMI, BP and lipid levels.30

Pharmacotherapy and control of cardiometabolic-renal risk factors

In this contemporary cohort of patients with YOD, 80% had been exposed to education by nurses and dietitians and the majority were receiving multiple drugs including GLDs, statins and RASi. With this intensive treatment, the mean HbA1c was 7.5% with around 40% at conventionally recommended HbA1c goal <7%, 40% BP goal <130/80 mm Hg and 70% LDL-cholesterol goal <2.6 mmol/L (or <1.8 mmol/L in presence of ASCVD). In the territory-wide Hong Kong Diabetes Surveillance Database, we reported the decline in HbA1c from a peak of 7.7% in 2002 to 7.2% in 2019, in part due to reform of the diabetes service and introduction of new medications, notably, DPP-4i. However, the 20–44 years age group continued to have suboptimal control with 16.8% patients having HbA1c >9%.31 Although use of statins and RASi had been proven to reduce major events and death rates, most RCTs recruited adults of older age which might contribute to the low usage of these drugs in young patients. To this end, our real-world evidence had confirmed RASi reduced all-events and all-cause death, especially in young patients when compared with other BP-lowering drugs such as calcium channel blockers.32 In another analysis of 360 202 Chinese patients with T2D, suboptimal BP control was the leading attributable risk factor contributing to 17% of all-cause deaths in the youngest age group (18–54 years).33 Given the prognostic significance of eGFR and ACR on major events, the high prevalence of CKD and albuminuria in this cohort, many of whom were treated with RASi and SGLT2i, although with residual albuminuria, highlighted the unmet needs in these young individuals. Our findings reaffirmed the urgency to conduct dedicated studies in these young patients to define treatment targets and strategies.9

In the PRISM study, we postulate that stricter metabolic targets could reduce the lifetime risk of cardiovascular-kidney complications driven by long disease duration in patients with YOD. In the Japan Diabetes Optimal Integrated Treatment (J-DOIT3) study involving 2452 Japanese patients with T2D (45–69 years), intensive multifactorial intervention to achieve HbA1c <6.2%, BP <120/75 mm Hg, LDL-cholesterol <2.1 mmol/L (or <1.8 mmol/L in patients with prior cardiovascular disease) reduced incident cerebrovascular and kidney events by 58% and 32%, respectively.34 35 In J-DOIT3 study, the intensively treated group had additional clinic visits and was given blood glucose testing strips and BP machines to promote self-management while the attending doctors were asked to intensify treatment early. Using data from Hong Kong Diabetes Register, we modeled that intensified control of multiple risk factors might reduce the lifetime hospitalization rate from 99 to 67 days in patients with YOD.6

Importance of psychosocial-behavioral factors and self-management

Suboptimal self-management and low medication adherence due to psychosocial-behavioral factors is a major barrier to optimize metabolic control. In this cohort, despite having received education, our young patients with diabetes were less likely to engage in healthy lifestyles with one-third being current/past smokers and one-third current drinkers. More detailed assessment showed high prevalence of anxiety, distress and depression. These patients were challenged by multiple demands and uncertainties associated with a diagnosis of diabetes during a critical period in life of education, career, relationships and family development. These commitments and stresses could elicit negative emotions while multiple priorities might limit their time and effort needed to improve self-management. Self-stigmatization or from others, and low willingness to seek support might compound these problems.5

Both younger age at diagnosis and longer disease duration of diabetes are associated with increased depression symptoms and diabetes-specific distress.5 Using validated questionnaires (PHQ-9, DASS-21, CDDS-15), around one-third of patients had negative emotions, especially in the domain related to rapport and support from care team as well as ambiguity about their future health. Over half had suboptimal diabetes self-care with lack of persistence in performing exercise, SMBG and foot care.

Concomitant medical history and lifecourse events

In the JADE-PRISM group, we obtained additional information based on prior knowledge which might contribute to the YOD. Some of these factors might be particularly relevant to Asia, for example, thalassemia-minor (8%) and positive hepatitis B surface antigen (7%).36 In this study, among the women, 17% had polycystic ovary syndrome and as many as two-thirds reported history of gestational diabetes often requiring insulin with some having pre-eclampsia and adverse pregnancy outcomes including preterm birth and fetal loss. Other significant illnesses included mental illness, self-harm, childhood cancer and steroid exposure and endocrinopathies, for example, thyroid disease, Cushing’s syndrome and primary hyperaldosteronism. This myriad of concomitant conditions could affect glycemic trajectory or interact with diabetes to amplify the future risks of vascular, cancer and non-vascular non-cancer events in these young patients.37 The high prevalence of mental illness is noteworthy. Using register data, we first reported that more than one-third of hospitalization bed-days in patients with diabetes before the age of 40 years were due to mental health conditions.6 In a territory-wide analysis, we recently confirmed that young adult patients with diabetes had threefold and fivefold higher hospitalization bed-days due to mental health disorders than those without diabetes, respectively.38

Autoimmunity and beta-cell function

In the JADE-PRISM group, we relayed the fasting CP, GADA as well as common and rare genetic variants to the patients to help them better understand their beta-cell function and risk of complications. Among these patients, 8.6% had significant insulin deficiency with fasting CP <0.2 nmol/L. After adjusting for fasting plasma glucose, nearly half of patients had HOMA2-%B <50% and 40% had HOMA2-IR >1.8. In our previous analysis, both low HOMA2-%B and high HOMA2-IR predicted the future requirement of insulin treatment.39 Besides, 5% of patients had GADA (n=22), 14 of whom were insulin-deficient (fasting CP <0.2 nmol/L) and already on insulin at recruitment.

Common and rare genetic variants

In the JADE-PRISM group, 4.5% had P/LP rare variants of MDGs, mainly due to GCK (glucokinase), HNF1A and WFS1 (Wolframin ER transmembrane glycoprotein) genes (table 3). Although MODY and latent autoimmune diabetes in adults are characterized by lean phenotypes, some of these patients were overweight or had obesity. The low prevalence of these diabetes subtypes could not explain the lean phenotype in 20%, and positive family history in 70% of these young patients. Common variants associated with T2D located in 700 genomic loci have been reported in different populations, although each of these variants only had a small effect size, especially in recent studies with increasing sample size.40 These loci only explained 20% of the heritability and the use of polygenic risk scores on top of conventional clinical risk factors only had incremental values in predicting diabetes. In this RCT, based on our previous discoveries on genetics of YOD and complications with validation in other Asian populations, we constructed genetic risk scores to communicate with the participants their personalized risks for complications and insulin requirement aimed at motivating behavioral change to achieve metabolic control (online supplemental tables S3–S4).

Table 3

Clinical profile and beta-cell function in patients with P/LP variants of monogenic diabetes genes


Despite the widespread recognition regarding the challenges in management of patients with young-onset T2D, there is a lack of evidence of using extended phenotyping and more precise classification to improve clinical outcome in these young patients. Results of this RCT will provide important insights on the clinical effectiveness and cost-effectiveness of an innovative, multicomponent, data-driven care model augmented by information technology, biogenetic markers and multidisciplinary care aimed at achieving stricter metabolic targets. Prospectively, analytics of these multidimensional data would generate algorithms to increase the precision of prognostication, classification and therapeutics in these high-risk young patients. The ultimate goal is to inform practice and policies in pursuit of value-based care with precision, evidence and quality.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Ethics statements

Patient consent for publication

Ethics approval

The study received approval from the Joint CUHK-NTEC Clinical Research Ethics Committee (CREC:2019-080-T) and complied with the Declaration of Helsinki. Participants gave informed consent to participate in the study before taking part.


We are most grateful to the participants and research staff for the PRISM project, especially Ms Cherry Chiu, Ms Maggie Lee, Ms Ally Chan and Mr Alex Ng for their dedication and professionalism. Special thanks are extended to all doctors and nurses at the Diabetes and Endocrine Center and Family Medicine Clinic at the Prince of Wales Hospital. We also acknowledge the support from the Research Grant Committee, United States National Institute of Health, Innovation and Technology Commission, Health and Medical Research Fund, Hong Kong Foundation for Research and Development, Asia Diabetes Foundation, Yao Yiu Sai Foundation and many other donations and grant support from the industry that have provided the ground work for the Genomics and Genetics Research Program in Diabetes in Hong Kong Chinese since 1995.



  • CKO and YNF are joint first authors.

  • Contributors JCNC and AL conceptualized, designed, implemented and coordinated the study. CKO and YNF wrote the first draft. CKO, BF, YNF, ESHL, CL and STFT performed the statistical analysis. RW and EWMP provided the technical support. All authors contributed to results interpretation, revised the manuscript critically and approved the final version. JCNC is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The results of this paper have been presented at the American Diabetes Association Scientific Session in June 2023 held in San Diego, USA as a published abstract.

  • Funding The study is supported by a Commissioned Grant by the Hong Kong Government Health and Medical Research Fund and the Hong Kong Genome Institute (CFS-CUHK2).

  • Competing interests JCNC and RCWM hold patents for using genetic markers to predict diabetes and its complications for personalized care. JCNC, RCWM and CL are cofounders of a start-up biotech company partially supported by the Technology Start-up Support Scheme for Universities (TSSSU) of the Hong Kong Government Innovation and Technology Commission.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.