Skip to main content
PMC home page
PLOS Medicine logoLink to PLOS Medicine
. 2020 Sep 18;17(9):e1003316. doi: 10.1371/journal.pmed.1003316

Age at diagnosis, glycemic trajectories, and responses to oral glucose-lowering drugs in type 2 diabetes in Hong Kong: A population-based observational study

Calvin Ke 1,2, Thérèse A Stukel 3,4, Baiju R Shah 2,3,4,5, Eric Lau 1,6, Ronald C Ma 1,7, Wing-Yee So 1, Alice P Kong 1,7, Elaine Chow 1, Juliana C N Chan 1,6,7,*, Andrea Luk 1,6,7
Editor: Sanjay Basu8
PMCID: PMC7500681  PMID: 32946450

Abstract

Background

Lifetime glycemic exposure and its relationship with age at diagnosis in type 2 diabetes (T2D) are unknown. Pharmacologic glycemic management strategies for young-onset T2D (age at diagnosis <40 years) are poorly defined. We studied how age at diagnosis affects glycemic exposure, glycemic deterioration, and responses to oral glucose-lowering drugs (OGLDs).

Methods and findings

In a population-based cohort (n = 328,199; 47.2% women; mean age 34.6 and 59.3 years, respectively, for young-onset and usual-onset [age at diagnosis ≥40 years] T2D; 2002–2016), we used linear mixed-effects models to estimate the association between age at diagnosis and A1C slope (glycemic deterioration) and tested for an interaction between age at diagnosis and responses to various combinations of OGLDs during the first decade after diagnosis. In a register-based cohort (n = 21,016; 47.1% women; mean age 43.8 and 58.9 years, respectively, for young- and usual-onset T2D; 2000–2015), we estimated the glycemic exposure from diagnosis until age 75 years.

People with young-onset T2D had a higher mean A1C (8.0% [standard deviation 0.15%]) versus usual-onset T2D (7.6% [0.03%]) throughout the life span (p < 0.001). The cumulative glycemic exposure was >3 times higher for young-onset versus usual-onset T2D (41.0 [95% confidence interval 39.1–42.8] versus 12.1 [11.8–12.3] A1C-years [1 A1C-year = 1 year with 8% average A1C]). Younger age at diagnosis was associated with faster glycemic deterioration (A1C slope over time +0.08% [0.078–0.084%] per year for age at diagnosis 20 years versus +0.02% [0.016–0.018%] per year for age at diagnosis 50 years; p-value for interaction <0.001). Age at diagnosis ≥60 years was associated with glycemic improvement (−0.004% [−0.005 to −0.004%] and −0.02% [−0.027 to −0.0244%] per year for ages 60 and 70 years at diagnosis, respectively; p-value for interaction <0.001). Responses to OGLDs differed by age at diagnosis (p-value for interaction <0.001). Those with young-onset T2D had smaller A1C decrements for metformin-based combinations versus usual-onset T2D (metformin alone: young-onset −0.15% [−0.105 to −0.080%], usual-onset −0.17% [−0.179 to −0.169%]; metformin, sulfonylurea, and dipeptidyl peptidase-4 inhibitor: young-onset −0.44% [−0.476 to −0.405%], usual-onset −0.48% [−0.498 to −0.459%]; metformin and α-glucosidase inhibitor: young-onset −0.40% [−0.660 to −0.144%], usual-onset −0.25% [−0.420 to −0.077%]) but greater responses to other combinations containing sulfonylureas (sulfonylurea alone: young-onset −0.08% [−0.099 to −0.065%], usual-onset +0.06% [+0.059 to +0.072%]; sulfonylurea and α-glucosidase inhibitor: young-onset −0.10% [−0.266 to 0.064%], usual-onset: 0.25% [+0.196% to +0.312%]). Limitations include possible residual confounding and unknown generalizability outside Hong Kong.

Conclusions

In this study, we observed excess glycemic exposure and rapid glycemic deterioration in young-onset T2D, indicating that improved treatment strategies are needed in this setting. The differential responses to OGLDs between young- and usual-onset T2D suggest that better disease classification could guide personalized therapy.

In a population-based cohort, Calvin Ke and colleagues investigate age at diagnosis and glycemic trajectories in Type 2 diabetes patients.

Author summary

Why was this study done?

  • Young-onset type 2 diabetes (diagnosed before age 40 years) is an aggressive disease, associated with higher risks of mortality and other complications compared with usual-onset type 2 diabetes (diagnosed at age 40 years or after).

  • Although exposure to hyperglycemia is a key risk factor for type 2 diabetes complications, the magnitude of this exposure over a lifetime has never been quantified.

  • Because young people were excluded from randomized control trials of oral glucose-lowering drugs, it is unknown whether these drugs are effective for treating young-onset type 2 diabetes.

What did the researchers do and find?

  • Young-onset type 2 diabetes was associated with over triple the exposure to hyperglycemia from diagnosis until age 75 years compared with usual-onset type 2 diabetes.

  • Diabetes progressed much faster in young-onset type 2 diabetes compared with usual-onset type 2 diabetes.

  • Compared with usual-onset type 2 diabetes, people with young-onset type 2 diabetes had smaller responses to most metformin-based drug combinations and greater responses to other combinations containing sulfonylureas.

What do these findings mean?

  • Excess exposure to hyperglycemia and rapid disease progression in young-onset T2D call for better treatment strategies.

  • The differential responses to oral glucose-lowering drugs between young- and usual-onset T2D suggest that better disease classification could guide personalized therapy.

Introduction

Life-threatening complications of type 2 diabetes (T2D) are caused by long-term exposure to hyperglycemia [1]. Glycemic exposure is defined as the area under the glycated hemoglobin A1c (A1C) curve in excess of 7% over time [2–5]. A meta-analysis of randomized control trials showed that every 10 A1C-years of glycemic exposure (that is, 10 years of A1C at 8%) predicts a 25% increase in the relative risk of major adverse cardiovascular events (MACE) [2]. Young-onset T2D (defined here as age at diagnosis <40 years) is an aggressive phenotype associated with higher lifetime risks of MACE and other complications compared to usual-onset T2D (age at diagnosis ≥40 years) [6–9]. Although early age at diagnosis and poor glycemic control are important risk factors for complications [10], lifetime glycemic exposure and its relationship with age at T2D diagnosis are unknown.

Reducing glycemic exposure typically requires frequent escalation of glucose-lowering therapies to overcome the natural history of T2D, which is characterized by progressively declining β-cell function and insulin sensitivity [11]. The rate of β-cell function decline is known as glycemic deterioration [3,12,13]. This measure can be quantified using repeated measurements of A1C over time, after accounting for glucose-lowering drugs using statistical adjustment [13]. Glycemic deterioration is more precise than other measures using binary thresholds (e.g., treatment failure) [14]. Among Europeans with usual-onset T2D, younger age at diagnosis is associated with faster glycemic deterioration [13,14]. In Asia, 1 in 5 adults with T2D attending medical clinics has young-onset T2D [10], yet data on glycemic exposure and deterioration are lacking.

Similarly, pharmacologic glycemic management strategies for young-onset T2D are poorly defined because this population was excluded from trials of glucose-lowering therapies in adults [15]. In a pediatric trial, metformin alone was associated with treatment failure in 51.7% of children with T2D [16]. Another study of children with T2D found that 3 months of insulin and 9 months of metformin failed to improve β-cell function versus metformin alone [17]. Considering the lack of randomized trials among adults with young-onset T2D [15], observational studies may provide valuable insights in this understudied population.

To address these knowledge gaps, we conducted a large register- and population-based cohort study to measure how age at diagnosis affects (1) glycemic exposure, (2) glycemic deterioration, and (3) responses to oral glucose-lowering drugs (OGLDs) during the first decade after diagnosis among adults with T2D. We hypothesized that earlier age at diagnosis would be associated with increased glycemic exposure, faster glycemic deterioration, and decreased responsiveness to OGLDs.

Methods

Setting

Hong Kong has a population of 7.3 million people, 92% of whom are of Chinese ethnicity [18]. The estimated diabetes prevalence was 10.3% in 2014 [19]. The Hong Kong Hospital Authority (HA) provides universal public healthcare modeled after the British National Health Service. Because of the high out-of-pocket cost of private healthcare, 95% of people with diabetes in Hong Kong receive care in HA clinics [20]. Consultation, prescription, and medication dispensing services are all provided on site under an all-inclusive nominal user fee [21], which is waived for low-income and other vulnerable groups [22]. All hospitals and clinics managed by the HA share the same electronic health record (EHR) with data including laboratory tests, discharge summaries, and dispensed prescriptions. Dispensed prescription records are comprehensive because drugs are prescribed and dispensed on site at the time of consultation. These data are linked by the unique Hong Kong Identity Card number.

Data sources

In 1995, the Prince of Wales Hospital set up the multicenter Hong Kong Diabetes Register (HKDR) as a research-driven quality improvement program [23]. The HKDR is a prospective cohort of people with prevalent diabetes, of varying disease duration. Participants were enrolled from 1994 to 2015. All participants undergo structured assessment (eyes, feet, blood, urine) by trained nurses every 2–3 years to collect data not routinely captured in the HA EHR, including diabetes type, age at diagnosis, family history, and lifestyle habits. Enrollment was open to any adult with diabetes based on self- or physician-initiated referrals from community- and hospital-based clinics. All services were provided at the nominal fee charged by the HA.

The Hong Kong Diabetes Surveillance Database (HKDSD) is a population-based cohort of incident diabetes cases from across the entire territory, identified from the HA EHR. We defined diabetes as any person with an A1C ≥6.5% (48 mmol/mol), outpatient fasting plasma glucose ≥7 mmol/L, non-insulin glucose-lowering drug prescription, or long-term insulin prescription (≥28 days). The date of diagnosis is defined as the date of first occurrence of any of these events. According to HA regulations, these drugs are only indicated for diabetes [24]. To avoid detecting gestational diabetes, events occurring within 9 months of any pregnancy-related encounter are excluded. As diabetes type is not systematically encoded in this EHR, we previously developed and validated algorithms to classify diabetes type based on diagnosis codes and prescriptions [25]. In this study, we excluded people who received their first insulin prescription within 90 days of diabetes diagnosis as having type 1 diabetes [25] and classified the remainder as T2D. This definition has a sensitivity of 94.6% (95% confidence interval 93.9%–95.2%) and positive predictive value of 99.9% (99.8%–100.0%) for predicting T2D [25].

To estimate lifetime glycemic exposure, we required person-years of follow-up from across the life span. This requirement necessitated the inclusion of people with new (incident) and old (prevalent) diagnoses of T2D of varying disease duration. We used the HKDR (“Register cohort”) for this objective because it includes people with both new and old T2D cases with disease duration ranging from zero to more than 70 years, whereas the HKDSD only includes new T2D cases after 2002. For example, a 70-year-old person in 2010 who was diagnosed with T2D at age 30 years in 1970 would be included in the HKDR but not the HKDSD. To measure glycemic deterioration (defined as A1C slope over time, after statistical adjustment for OGLDs) and medication responses, we required person-years of follow-up from the first decade after a new T2D diagnosis. This requirement was based on a previous study showing that glycemic deterioration is a linear function of time during this period [26]. Although both cohorts included people with new diagnoses, we used the HKDSD (“Population cohort”) because it was larger and included people in the HKDR.

Study population

Register cohort (glycemic exposure)

We included adults aged 18–75 years in the HKDR with prevalent T2D observed between January 1, 2000, and December 31, 2015, and defined the index date as the earliest date within this period when the person met these criteria (Fig A in S1 Appendix). We excluded observations between 1994 to 1999 to more closely match the time period of the population cohort. We excluded people with type 1 diabetes, gestational diabetes, and non-Chinese ethnicity.

Population cohort (glycemic deterioration and medication responses)

We included adults aged 18–75 years in the HKDSD with T2D diagnosed between January 1, 2002, and December 31, 2012, (Fig A in S1 Appendix; dates chosen based on data availability). The index date was the diagnosis date. We followed people for up to 10 years after diagnosis of diabetes, censoring at the date of the first insulin prescription. The maximum follow-up date was December 31, 2016, for both the register and population cohorts, censoring at death or age 75 years.

Statistical analysis

We conducted the study according to a prospective analysis plan (S1 Analysis Plan). The primary outcome was A1C measured over the follow-up period, and the primary exposure was age at diagnosis, expressed as a continuous variable. In the register cohort, we calculated the mean A1C by attained age among all people with young-onset T2D. Using these values, we estimated the glycemic exposure, defined as the area under the A1C curve in excess of 7% (53 mmol/mol), from the mean observed age at diagnosis until age 75 years (Fig B in S1 Appendix), and repeated this procedure for usual-onset T2D.

In the population cohort, we used a linear mixed-effects model with age at diagnosis and time as independent variables and A1C as the dependent variable, with person-specific random intercepts and slopes because we assumed each person had his or her own A1C trajectory. Although we did not dichotomize age at diagnosis in the model, we presented the expected results for young- and usual-onset T2D based on the mean observed ages at diagnosis in each group. Glycemic deterioration was defined as the A1C slope over time, adjusted for model covariates. We included time-varying covariates for each OGLD based on their dispensing records, allowing for people to switch on and off a drug. We assumed the same absolute effect for each drug class, regardless of A1C level, and that drugs within the same class would lower A1C by a similar decrement [27]. We included metformin, sulfonylureas, and 10 multidrug combinations as unique variables (Table A in S1 Appendix). Effects of individual drugs taken in combination were not assumed to be additive (S1 Appendix) because OGLDs may have different effects when prescribed alone versus in combination [28,29]. Each drug combination was considered as a separate variable. When a new drug was dispensed, we excluded the first A1C measurement on the new treatment regimen to allow sufficient time for the A1C to equilibrate. We adjusted for prespecified variables (Fig C in S1 Appendix) including sex and recent comorbidities, namely, ischemic heart disease, congestive heart failure, stroke, peripheral arterial disease, and cancer, based on principal diagnoses from hospitalizations occurring within 2 years prior to the index date (Table B in S1 Appendix) and chronic kidney disease classified by the estimated glomerular filtration rate. As anthropometric data were unavailable, we included triglycerides and high-density lipoprotein cholesterol as proxies for obesity [30].

We estimated the association between age at diagnosis and A1C slope (glycemic deterioration) and tested for an interaction between age at diagnosis and responses to various combinations of OGLDs. We conducted several sensitivity analyses. To test the validity of our assumption of a linear relationship between age at diagnosis and A1C, we repeated the analysis using a nonlinear model with restricted cubic splines containing 4 knots placed at fixed quantiles [31]. We also repeated the analysis excluding A1C values from the first 6 and 12 months after the diagnosis of T2D because these measurements might be unusually elevated as a consequence of transiently depressed β-cell function (“glucotoxicity”) [32]. To test whether the baseline A1C level affected A1C lowering, we repeated the analysis (post hoc) with an interaction term between each drug combination and its observed baseline A1C level. Interaction terms between age at diagnosis and drug combinations were excluded in this model because of computational limitations.

We used the MIXED procedure (SAS version 9.4, SAS Institute, Cary, NC, www.sas.com), specifying the spatial power covariance structure to account for nonequal time intervals between A1C measurements. Missing outcome data were minimal (4.0% and 6.4% in the register and population cohorts) and handled by complete case analysis. The study was approved by The Chinese University of Hong Kong-New Territories East Cluster Clinical Research Ethics Committee and the University of Toronto Health Sciences Research Ethics Board. Data in the HKDSD were anonymized at the time of access. Individuals in the HKDR provided written informed consent for the use of their data. This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology guideline (S1 STROBE checklist).

Results

We included 21,016 people (47.1% women, 0.2 million person-years follow-up, median 8.4 years) in the register cohort and 328,199 people (47.2% women, 2.4 million person-years follow-up, median 7.9 years) in the population cohort with >3 million A1C measurements combined (Fig D in S1 Appendix). The mean age at diagnosis was similar in both cohorts (young-onset: 33.8–34.6 years, usual-onset: 53.8–59.3 years, Table 1). Recent comorbidities were more common in the register than the population cohort. The baseline A1C was higher in young- versus usual-onset T2D (register: young-onset, mean 7.8% [62 mmol/mol, standard deviation 1.5%]; usual-onset, 7.3% [56 mmol/mol, 1.2%]; population: young-onset, 7.6% [60 mmol/mol, 1.6%]; usual-onset, 7.3% [56 mmol/mol, 1.3%]). Lipid levels were similar across cohorts. Metformin (young-onset: 31.8–56.4%, usual-onset: 41.5–48.6%) and sulfonylureas (young-onset: 25.9–38.4%, usual-onset: 33.6–40.7%) were the most common OGLDs. Insulin use during the first year after diagnosis was more common in the register (young-onset: 24.2%, usual-onset: 11.7%) than the population cohort (young-onset: 3.0%, usual-onset: 1.1%).

Table 1. Baseline characteristics in the register (2000–2016) and population (2002–2016) cohorts.

Register Cohort Population Cohort
Young-Onset
(n = 4,058)		 Usual-Onset
(n = 16,958)		 Young-Onset
(n = 15,265)		 Usual-Onset
(n = 312,934)
Age at diagnosis (years) 33.8 (5.5) 53.8 (8.1) 34.6 (4.8) 59.3 (9.0)
Index age (years) 43.8 (10.5) 58.9 (8.3) 34.6 (4.8) 59.3 (9.0)
Women (n, %) 2,031 (50.0) 7,876 (46.4) 6,508 (42.6) 148,516 (47.4)
Recent comorbidities within 2 year of index* (n, %)
 Ischemic heart disease 62 (1.5) 546 (3.2) 57 (0.4) 4,780 (1.5)
 Congestive heart failure 14 (0.3) 130 (0.8) 28 (0.2) 1,244 (0.4)
 Stroke 57 (1.4) 534 (3.2) 55 (0.4) 3,809 (1.2)
 Peripheral arterial disease 11 (0.3) 33 (0.2) 2 (0.0) 105 (0.0)
 Cancer 56 (1.4) 359 (2.1) 152 (1.0) 4,794 (1.5)
Laboratory Values (within 2 years of index* for A1C; 3 years for other variables)
Hemoglobin A1C (%) 7.8 (1.5) 7.3 (1.2) 7.6 (1.6) 7.3 (1.3)
Fasting plasma glucose (mmol/L) 8.6 (2.4) 7.9 (1.9) 8.1 (2.6) 7.6 (2.0)
LDL-C (mmol/L) 2.8 (0.7) 2.7 (0.7) 2.9 (0.8) 3.0 (0.8)
HDL-C (mmol/L) 1.3 (0.3) 1.3 (0.3) 1.2 (0.3) 1.1 (0.3)
Triglycerides (mmol/L; median, IQR) 1.5 (1.2) 1.5 (1.0) 1.7 (1.4) 1.5 (1.0)
Estimated GFR (mL/min/1.73 m2) 93.2 (22.7) 78.9 (20.5) 105 (18.0) 81.2 (18.9)
 <60 mL/min/1.73 m2 (n, %) 327 (8.0) 2,787 (16.4) 404 (2.8) 36,541 (12.6)
 <15 mL/min/1.73 m2 (n, %) 41 (1.0) 163 (1.0) 59 (0.4) 2,017 (0.7)
Pharmacotherapy within 1 year after index* (n, %)
Metformin 1,292 (31.8) 7,032 (41.5) 8,611 (56.4) 152,156 (48.6)
Sulfonylurea 1,051 (25.9) 5,704 (33.6) 5,856 (38.4) 127,512 (40.7)
DPP-4 Inhibitor 16 (0.4) 68 (0.4) 417 (0.1) 43 (0.3)
Thiazolidinedione 38 (0.9) 124 (0.7) 178 (0.1) 13 (0.1)
Acarbose 24 (0.6) 96 (0.6) 42 (0.3) 1,085 (0.4)
GLP-1 receptor agonist 0 (0.0) 0 (0.0) 3 (0.0) 2 (0.0)
Insulin 983 (24.2) 1,983 (11.7) 464 (3.0) 3,400 (1.1)
Open in a new tab

Young-onset type 2 diabetes is defined here as age at diagnosis <40 years and usual-onset type 2 diabetes as ≥40 years. Values are means and standard deviations unless otherwise indicated. Because of the large sample size, baseline differences should be interpreted based on clinical significance rather than statistical significance.

*The index date in the register cohort was the date of enrollment in the register, whereas the index date in the population cohort was the date of diabetes diagnosis.

DPP-4, dipeptidyl peptidase-4; GFR, glomerular filtration rate; GLP-1, glucagon-like peptide 1; HDL-C, high-density lipoprotein cholesterol; IQR, interquartile range; LDL-C, low-density lipoprotein cholesterol.

Glycemic exposure in young-onset and usual-onset T2D

In the register, people with young-onset T2D had a higher mean observed A1C (mean of annual means 8.0% [64 mmol/mol], standard deviation 0.15%) versus those with usual-onset T2D (7.6% [59 mmol/mol], 0.03%) throughout the life span (p < 0.001; Fig 1). The cumulative glycemic exposure was >3 times higher for young- versus usual-onset T2D (41.0 [39.1–42.8] versus 12.1 [11.8–12.3] A1C-years [each A1C-year equivalent to one year with an 8% (64 mmol/mol) average A1C]). In the population cohort, mean observed A1C levels were more elevated among people with young-onset compared with usual-onset T2D, both at diagnosis and throughout the first decade after diagnosis.

Fig 1. Observed A1C and glycemic exposure among adults in the register and population cohorts.

Fig 1

Open in a new tab

These A1C values are not adjusted for medications. (A) Mean observed A1C across the age span (attained age) in the register cohort (2000–2016), stratified by age at diagnosis (smoothed using 3-year moving averages). The shaded areas indicate glycemic exposure in excess of 7%. Sample sizes are indicated for each 5-year age group. (B) Cumulative glycemic exposure in the register cohort (2000–2016), defined as area under the A1C curve in excess of 7%, by age at diagnosis. One A1C-year is equivalent to one year of exposure to an average A1C of 8%. For example, a person diagnosed with diabetes at age 30 years, with an average A1C of 8% from age 30 to 75 years, would have been exposed to a 1% excess in A1C over 45 years, which is equivalent to 45 A1C-years of glycemic exposure. (C) Mean observed A1C by years since type 2 diabetes diagnosis in the population cohort (2002–2016), stratified by age at diagnosis. A1C, hemoglobin A1c.

Glycemic deterioration in young-onset and usual-onset T2D

In the population cohort, glycemic deterioration differed significantly across age at diagnosis (Fig 2, Table C in S1 Appendix). Younger age at diagnosis was associated with faster glycemic deterioration (p-value for interaction <0.001). People aged ≤30 years at diagnosis had the most rapid deterioration (+0.08% [95% confidence interval 0.078 to 0.084%] per year for age 20 years at diagnosis) as compared with no deterioration in those aged 60 years at diagnosis (0.00% [−0.005 to −0.004%] per year), and glycemic improvement (−0.02% [−0.027 to −0.0244%] per year) in people aged 70 years at diagnosis. In a sensitivity analysis, we observed similar results allowing for a nonlinear relationship between age at diagnosis and glycemic deterioration (Fig E in S1 Appendix) and excluding A1C values from the first 6 to 12 months after diagnosis (Fig F in S1 Appendix).

Fig 2. Glycemic deterioration during the first decade after type 2 diabetes diagnosis.

Fig 2

Open in a new tab

Results are stratified by age at diagnosis (population cohort, Hong Kong Diabetes Surveillance Database, 2002–2016). Glycemic deterioration is the modeled slope of the A1C over time after adjusting for oral glucose-lowering drug prescriptions. The sample size (n) is indicated for each age group (age at diagnosis <25, 25–34, 35–44, 45–54, 55–64, ≥65 years). See Table C in S1 Appendix for numeric values. A1C, hemoglobin A1c.

Responses to OGLDs in young-onset and usual-onset T2D

There was a statistically significant difference between people with young- and usual-onset T2D in their responses to OGLDs (p-value <0.001 for omnibus test across all combinations; Fig 3). Combinations containing metformin were associated with lower A1C values, but these decrements appeared slightly smaller for young-onset T2D. Metformin alone was associated with an A1C decrement of −0.15% (95% confidence interval −0.105% to −0.080%) in young-onset and −0.17% (−0.179% to −0.169%) in usual-onset T2D. The combination of metformin, a sulfonylurea and a dipeptidyl peptidase-4 inhibitor had the greatest A1C decrement (young-onset: −0.44% [−0.476% to −0.405%], usual-onset: −0.48% [−0.498% to −0.459%]). The combination of metformin and an α-glucosidase inhibitor (class exclusively consisting of acarbose) was associated with a greater A1C decrement in young-onset (−0.40% [−0.660% to −0.144%]) than usual-onset T2D (−0.25% [−0.420% to −0.077%]).

Fig 3. A1C responses to oral glucose-lowering drugs among people with young- and usual-onset type 2 diabetes.

Fig 3

Open in a new tab

These insulin-naive individuals were observed during the first decade after diabetes diagnosis (population cohort, Hong Kong Diabetes Surveillance Database, 2002–2016). Sample sizes (number of A1C measurements) are indicated for each combination. Error bars indicate 95% confidence intervals. Differences between young- and usual-onset type 2 diabetes were statistically significant across all combinations (omnibus test p < 0.001). A, acarbose; D, dipeptidyl peptidase-4 inhibitor; M, metformin; S, sulfonylurea; T, thiazolidinedione.

Most combinations containing a sulfonylurea (without metformin) were associated with reduced A1C values in young-onset T2D (−0.08% [−0.099% to −0.065%] for sulfonylurea alone) but increased A1C values in usual-onset T2D (+0.06% [+0.059% to +0.072%] for sulfonylurea alone). The combination of a sulfonylurea and an α-glucosidase inhibitor was associated with the largest A1C decrement in young-onset T2D (−0.10% [−0.266% to 0.064%]; usual-onset: +0.25% [+0.196% to +0.312%]). Sensitivity analyses excluding A1C values from the first 6 and 12 months after diagnosis yielded relatively similar findings for metformin-based combinations, although combinations containing a sulfonylurea were associated with increased A1C values in young-onset T2D (Fig G in S1 Appendix). Adjustment for the baseline A1C with each OGLD combination yielded similar results to the main analysis (Fig H in S1 Appendix).

Discussion

In this large population- and register-based study, we found that people with young-onset T2D had poorly controlled hyperglycemia throughout their life span, resulting in more than triple the cumulative glycemic exposure versus usual-onset T2D. This disparity was driven by rapid glycemic deterioration, which was particularly steep among people diagnosed with T2D before age 30 years. Conversely, we revealed that T2D diagnosis after age 60 years was associated with glycemic improvement—a novel finding that has not been previously reported to our knowledge. In this real-world study, people with young-onset T2D had slightly smaller A1C decrements compared with usual-onset T2D for most combinations of OGLDs including metformin, whereas young-onset T2D was unexpectedly associated with greater responsiveness to sulfonylureas and α-glucosidase inhibitors than usual-onset T2D. Although most OGLD combinations appear to lower A1C by similar decrements across age at diagnosis, the rapidity of glycemic deterioration in young-onset T2D suggests that early combination therapy [33] and aggressive treatment escalation are needed to reduce the massive excess in glycemic exposure that we observed.

Excess glycemic exposure in young-onset T2D

Our study is the first, to our knowledge, to describe the greater than 3-fold disparity in lifetime glycemic exposure between young- and usual-onset T2D. We estimated that people with young-onset T2D will accumulate nearly 40 A1C-years of glycemic exposure by age 75 years. Considering that every 10 A1C-years predicts a 25% increase in relative risk of MACE [2], this excess glycemic exposure is consistent with the high complication risks of young-onset T2D that we [6,7] and others [8,9] have reported. In the United Kingdom Prospective Diabetes Study [1], early reduction of glycemic exposure in T2D has been proven to prevent complications in future decades. Our real-world findings emphasize the urgent need for a paradigm shift toward early and intensive glycemic management to improve survival in the young-onset T2D population.

Rapid glycemic deterioration in young-onset T2D

Glycemic deterioration was particularly rapid in young-onset T2D, peaking at 0.09% per year for an age at diagnosis of 20 years. This rapid progression is thought to be caused by declining β-cell function in the setting of peripheral and hepatic insulin resistance [34,35]. Because of genetic factors, impaired β-cell secretion is an especially important driver of T2D in East Asians [36]. Our findings are consistent with the 20% to 35% per year decline in β-cell function reported among children aged 10 to 17 years in the Treatment Options for type 2 Diabetes in Adolescents and Youth study [34]. By contrast, we observed glycemic improvement among adults diagnosed at age 60 years or above. This finding supports the principle of therapy de-intensification among older adults [37], who may be predisposed to developing a milder phenotype of T2D [38,39]. However, previous studies of usual-onset T2D in European populations reported different results. In Scotland, glycemic deterioration rates were 0.09% and 0.20% per year for ages at diagnosis of 70 and 50 years, respectively, according to the Genetics of Diabetes Audit and Research in Tayside Study [13]. In A Diabetes Outcome Progression Trial, the rate of glycemic deterioration in a largely European cohort was 0.07%–0.14% per year [12]. It is unclear why these rates differed from our study. Rates of β-cell decline might be lower among older Chinese people [40] compared with Europeans [41]. Lifestyle might also play a role, as Hong Kong has the world’s most physically active population [42,43].

Differential responses to OGLDs between young- and usual-onset T2D

In this observational study, A1C decrements associated with OGLD combinations were relatively modest compared with clinical trials. Although we adjusted for adherence using time-varying covariates based on prescription dispensing records, this discrepancy may relate to people not consuming their dispensed medications [44,45]. Rather than focusing on the absolute magnitude of A1C decrements, we compared relative changes in A1C across different ages at diagnosis and drug classes to identify any potentially informative response differences in this real-world setting. In both young- and usual-onset T2D, the combination of metformin, a sulfonylurea and a DDP-4 inhibitor was associated with the greatest A1C reduction. These findings are consistent with evidence suggesting that dipeptidyl peptidase-4 inhibitors lower blood glucose especially well among of East Asians and other people with low body mass indices [46,47]. This combination might be particularly effective in the context of β-cell dysfunction in East Asians, especially in the presence of obesity [48]. Thiazolidinediones were associated with relatively large A1C decrements when used in combination with metformin and a sulfonylurea, which is consistent with the findings of the Treatment Options for type 2 Diabetes in Adolescents and Youth study [34].

Overall, young-onset T2D was associated with slightly smaller A1C decrements versus usual-onset T2D for most—but not all—regimens. As observed in previous studies [12], combinations containing a sulfonylurea were associated with increased A1C values in usual-onset T2D (0.06%–0.25%). However, in young-onset T2D, sulfonylureas were associated with reduced A1C values overall (−0.08% to −0.10%) and relatively smaller A1C increases compared with usual-onset T2D when excluding the initial 6-month period after diagnosis. The reason for this differential response is unclear. In our setting, we previously reported that 5% of people in a young-onset T2D cohort actually had maturity onset diabetes of the young type 3 [49]—which is known to respond well to sulfonylureas [50]—although genetic data were unavailable in the present study. Additionally, α-glucosidase inhibitors, which reduce glycemic excursion and demand on β-cells, were associated with greater A1C reductions in young- versus usual-onset T2D, albeit with wide confidence intervals. This class is efficacious and well tolerated in East Asians [51–54] and merits further investigation among people with young-onset T2D and poor β-cell reserve.

Strengths and limitations

Strengths of our study include its large sample size, complementary cohorts, and long follow-up duration. Unlike previous studies, our detailed register enabled us to capture A1C levels across the lifetime and quantify glycemic exposure at various ages at diagnosis. Using a territory-wide laboratory database containing over 3 million A1C measurements and 2.8 million dispensing records, we were able to demonstrate the important associations between age at diagnosis and glycemic deterioration and responses to OGLDs. Future studies are required to determine whether these findings extend to sodium-glucose cotransporter-2 inhibitors and glucagon-like peptide 1 receptor agonists, which were not widely available in our setting during the study period. As this was a real-world study, we could not confirm whether people consumed the medications dispensed. We cannot exclude the possibility of confounding by indication or by unmeasured variables. Further research is needed to confirm whether these findings are generalizable outside of Hong Kong’s healthcare system and to other ethnic groups. The differential responses to OGLDs between young- and usual-onset T2D are hypothesis-generating, and experimental studies are needed to confirm these findings.

Conclusion

Young-onset T2D progresses rapidly and remains poorly controlled throughout the life span, resulting in more than triple the lifetime glycemic exposure versus usual-onset T2D. The larger A1C reductions associated with sulfonylureas and α-glucosidase inhibitors in young-onset T2D emphasize the importance of β-cell dysfunction in this condition and calls for more precise phenotyping to personalize therapy [55–58]. Pending such evidence, our data suggest that most OGLD regimens are associated with lower glucose levels in young-onset T2D. However, better strategies are needed to apply and rapidly escalate these treatments to overcome therapeutic inertia and mitigate the aggressive trajectory of glycemic deterioration in this high-risk population—especially considering the lack of pharmacologic approaches to improve underlying β-cell function [17]. Given the complex biological and behavioral determinants of young-onset T2D, it is especially important to deploy these strategies in the context of integrated team-based care, disease classification, and patient empowerment to achieve better outcomes [59].

Supporting information

S1 Appendix

(PDF)

Click here for additional data file. (435.2KB, pdf)
S1 Analysis plan

(PDF)

Click here for additional data file. (265.8KB, pdf)
S1 STROBE checklist. STROBE, Strengthening the Reporting of Observational Studies in Epidemiology.

(DOCX)

Click here for additional data file. (33KB, docx)

Acknowledgments

We thank Dr. Allan Detsky for his useful comments on an earlier version of this manuscript and the Hong Kong Hospital Authority for providing the data for this study. Part of this work was presented at the American Diabetes Association’s 79th Scientific Sessions in San Francisco, California, June 7–11, 2019.

Abbreviations

A1C

hemoglobin A1c

MACE

major adverse cardiovascular events

OGLD

oral glucose-lowering drug

T2D

type 2 diabetes

Data Availability

Due to local laws and regulations regarding the use and distribution of personal data, the data used in the present study cannot be deposited in a public repository. Data access can be applied for through the Data Sharing Portal of the Hong Kong Hospital Authority (https://www3.ha.org.hk/data/DCL/Index/).

Funding Statement

CK is supported by the Canadian Institutes of Health Research (https://cihr-irsc.gc.ca/) Canada Graduate Scholarship and Michael Smith Foreign Study Supplements, the University of Toronto Clinician Investigator Program (https://cip.utoronto.ca/), the Canadian Society of Endocrinology and Metabolism (https://www.endo-metab.ca/) Dr. Fernand Labrie Research Fellowship Grant, and the Royal College of Physicians and Surgeons of Canada (http://www.royalcollege.ca/) Detweiler Traveling Fellowship. (Grant numbers not applicable.) The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

  • 1.Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HAW. 10-Year Follow-up of Intensive Glucose Control in Type 2 Diabetes. New England Journal of Medicine. 2008;359(15):1577–89. 10.1056/NEJMoa0806470 [DOI] [PubMed] [Google Scholar]
  • 2.Roussel R, Steg PG, Mohammedi K, Marre M, Potier L. Prevention of cardiovascular disease through reduction of glycaemic exposure in type 2 diabetes: A perspective on glucose-lowering interventions. Diabetes Obes Metab. 2018. February 1;20(2):238–44. 10.1111/dom.13033 [DOI] [PubMed] [Google Scholar]
  • 3.Brown JB, Nichols GA, Perry A. The Burden of Treatment Failure in Type 2 Diabetes. Diabetes Care. 2004. July 1;27(7):1535–40. 10.2337/diacare.27.7.1535 [DOI] [PubMed] [Google Scholar]
  • 4.Nichols GA, Rosales AG, Perrin NA, Fortmann SP. The Association Between Different A1C-Based Measures of Glycemia and Risk of Cardiovascular Disease Hospitalization. Diabetes Care. 2014. January;37(1):167–72. 10.2337/dc13-1300 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.van Wijngaarden RPT, Overbeek JA, Heintjes EM, Schubert A, Diels J, Straatman H, et al. Relation Between Different Measures of Glycemic Exposure and Microvascular and Macrovascular Complications in Patients with Type 2 Diabetes Mellitus: An Observational Cohort Study. Diabetes Ther. 2017. October 1;8(5):1097–109. 10.1007/s13300-017-0301-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Chan JCN, Lau ESH, Luk AOY, Cheung KKT, Kong APS, Yu LWL, et al. Premature Mortality and Comorbidities in Young-onset Diabetes: A 7-Year Prospective Analysis. The American Journal of Medicine. 2014. July 1;127(7):616–24. 10.1016/j.amjmed.2014.03.018 [DOI] [PubMed] [Google Scholar]
  • 7.Ke C, Lau E, Shah BR, Stukel TA, Ma RC, So W-Y, et al. Excess Burden of Mental Illness and Hospitalization in Young-Onset Type 2 Diabetes: A Population-Based Cohort Study. Annals of Internal Medicine. 2019. February 5;170(3):145–54. 10.7326/M18-1900 [DOI] [PubMed] [Google Scholar]
  • 8.Pavkov ME, Bennett PH, Knowler WC, Krakoff J, Sievers ML, Nelson RG. Effect of Youth-Onset Type 2 Diabetes Mellitus on Incidence of End-Stage Renal Disease and Mortality in Young and Middle-Aged Pima Indians. JAMA. 2006. July 26;296(4):421–6. 10.1001/jama.296.4.421 [DOI] [PubMed] [Google Scholar]
  • 9.Steinarsson AO, Rawshani A, Gudbjörnsdottir S, Franzén S, Svensson A-M, Sattar N. Short-term progression of cardiometabolic risk factors in relation to age at type 2 diabetes diagnosis: a longitudinal observational study of 100,606 individuals from the Swedish National Diabetes Register. Diabetologia. 2018. March 1;61(3):599–606. 10.1007/s00125-017-4532-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Yeung RO, Zhang Y, Luk A, Yang W, Sobrepena L, Yoon K-H, et al. Metabolic profiles and treatment gaps in young-onset type 2 diabetes in Asia (the JADE programme): a cross-sectional study of a prospective cohort. The Lancet Diabetes & Endocrinology. 2014. January 12;2(12):935–43. [DOI] [PubMed] [Google Scholar]
  • 11.van Raalte DH, Diamant M. Glucolipotoxicity and beta cells in type 2 diabetes mellitus: Target for durable therapy? Diabetes Research and Clinical Practice. 2011. August 1;93:S37–46. 10.1016/S0168-8227(11)70012-2 [DOI] [PubMed] [Google Scholar]
  • 12.Kahn SE, Haffner SM, Heise MA, Herman WH, Holman RR, Jones NP, et al. Glycemic Durability of Rosiglitazone, Metformin, or Glyburide Monotherapy. New England Journal of Medicine. 2006. December 7;355(23):2427–43. 10.1056/NEJMoa066224 [DOI] [PubMed] [Google Scholar]
  • 13.Donnelly LA, Zhou K, Doney ASF, Jennison C, Franks PW, Pearson ER. Rates of glycaemic deterioration in a real-world population with type 2 diabetes. Diabetologia. 2018. March;61(3):607–15. 10.1007/s00125-017-4519-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Wallace TM, Matthews DR. Coefficient of failure: a methodology for examining longitudinal β-cell function in Type 2 diabetes. Diabetic Medicine. 2002. June;19(6):465–9. 10.1046/j.1464-5491.2002.00718.x [DOI] [PubMed] [Google Scholar]
  • 15.Ke C, Shah BR, Luk AO, Di Ruggiero E, Chan JCN. Cardiovascular Outcomes Trials in Type 2 Diabetes: Time to Include Young Adults. Diabetes, Obesity and Metabolism. 2020. January 20;22(1):3–5. 10.1111/dom.13874 [DOI] [PubMed] [Google Scholar]
  • 16.TODAY Study Group, Zeitler P, Hirst K, Pyle L, Linder B, Copeland K, et al. A clinical trial to maintain glycemic control in youth with type 2 diabetes. N Engl J Med. 2012. June 14;366(24):2247–56. 10.1056/NEJMoa1109333 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Consortium RISE. Effects of Treatment of Impaired Glucose Tolerance or Recently Diagnosed Type 2 Diabetes With Metformin Alone or in Combination With Insulin Glargine on β-Cell Function: Comparison of Responses In Youth And Adults. Diabetes. 2019. August;68(8):1670–80. 10.2337/db19-0299 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.2016 Population By-census Office. 2016 Population By-Census—Main Results [Internet]. Hong Kong: Census and Statistics Department, Hong Kong Special Administrative Region Government. 2017 Nov [cited 2019 Jan 7]. https://www.bycensus2016.gov.hk/data/16bc-main-results.pdf
  • 19.Quan J, Li TK, Pang H, Choi CH, Siu SC, Tang SY, et al. Diabetes incidence and prevalence in Hong Kong, China during 2006–2014. Diabet Med. 2017. July 1;34(7):902–8. 10.1111/dme.13284 [DOI] [PubMed] [Google Scholar]
  • 20.Lau IT. A Clinical Practice Guideline to Guide a System Approach to Diabetes Care in Hong Kong. Diabetes Metab J. 2017. April;41(2):81–8. 10.4093/dmj.2017.41.2.81 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Leung GM, Wong IO.L., Chan W-S, Choi S, Lo S-V. The ecology of health care in Hong Kong. Social Science & Medicine. 2005. August;61(3):577–90. [DOI] [PubMed] [Google Scholar]
  • 22.Legislative Council Panel on Health Services. Enhancements of Medical Fee Waiver Mechanism of Hospital Authority [Internet]. Hong Kong: Legislative Council Panel on Health Services; 2017 Jul [cited 2020 Jun 8] p. 8. (LC Paper). Report No.: CB(2)1843/16-17(02). https://www.legco.gov.hk/yr16-17/english/panels/hs/papers/hs20170717cb2-1843-2-e.pdf
  • 23.Chan JCN, So W, Ma RCW, Tong PCY, Wong R, Yang X. The Complexity of Vascular and Non-Vascular Complications of Diabetes: The Hong Kong Diabetes Registry. Curr Cardiovasc Risk Rep. 2011. June;5(3):230–9. 10.1007/s12170-011-0172-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Hospital Authority. Hospital Authority Drug Formulary [Internet]. 15.5. Hong Kong: Hospital Authority; 2020 [cited 2020 Jun 9]. https://www.ha.org.hk/hadf/Portals/0/Docs/HADF_List/External%20list%2020190413/6%20%20%20ENDOCRINE%20SYSTEM.pdf
  • 25.Ke C, Stukel TA, Luk A, Shah BR, Jha P, Lau E, et al. Development and validation of algorithms to classify type 1 and 2 diabetes according to age at diagnosis using electronic health records. BMC Med Res Methodol. 2020. February 24;20(35):1–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Levy J, Atkinson AB, Bell PM, McCance DR, Hadden DR. Beta-cell deterioration determines the onset and rate of progression of secondary dietary failure in Type 2 diabetes mellitus: the 10-year follow-up of the Belfast Diet Study. Diabetic Medicine. 1998. April 1;15(4):290–6. [DOI] [PubMed] [Google Scholar]
  • 27.Sherifali D, Nerenberg K, Pullenayegum E, Cheng JE, Gerstein HC. The Effect of Oral Antidiabetic Agents on A1C Levels: A systematic review and meta-analysis. Diabetes Care. 2010. August 1;33(8):1859–64. 10.2337/dc09-1727 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Mearns ES, Sobieraj DM, White CM, Saulsberry WJ, Kohn CG, Doleh Y, et al. Comparative Efficacy and Safety of Antidiabetic Drug Regimens Added to Metformin Monotherapy in Patients with Type 2 Diabetes: A Network Meta-Analysis. PLoS ONE. 2015. April 28;10(4):e0125879 10.1371/journal.pone.0125879 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Maruthur NM, Tseng E, Hutfless S, Wilson LM, Suarez-Cuervo C, Berger Z, et al. Diabetes Medications as Monotherapy or Metformin-Based Combination Therapy for Type 2 Diabetes: A Systematic Review and Meta-analysis. Annals of Internal Medicine. 2016. June 7;164(11):740 10.7326/M15-2650 [DOI] [PubMed] [Google Scholar]
  • 30.Alberti KGMM Eckel RH, Grundy SM Zimmet PZ, Cleeman JI Donato KA, et al. Harmonizing the Metabolic Syndrome: A Joint Interim Statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation. 2009. October 20;120(16):1640–5. 10.1161/CIRCULATIONAHA.109.192644 [DOI] [PubMed] [Google Scholar]
  • 31.Harrell FE. SAS macros and data step programs useful in survival analysis and logistic regression [Internet]. Duke University Medical Center: Takima West Corporation; 1991 [cited 2018 Dec 28]. http://biostat.mc.vanderbilt.edu/wiki/pub/Main/SasMacros/survrisk.txt
  • 32.Turner RC, McCarthy ST, Holman RR, Harris E. Beta-cell function improved by supplementing basal insulin secretion in mild diabetes. Br Med J. 1976. May 22;1(6020):1252–4. 10.1136/bmj.1.6020.1252 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Matthews DR, Paldánius PM, Proot P, Chiang Y, Stumvoll M, Del Prato S. Glycaemic durability of an early combination therapy with vildagliptin and metformin versus sequential metformin monotherapy in newly diagnosed type 2 diabetes (VERIFY): a 5-year, multicentre, randomised, double-blind trial. The Lancet. 2019. October 26;394(10208):1519–29. [DOI] [PubMed] [Google Scholar]
  • 34.TODAY Study Group. Effects of metformin, metformin plus rosiglitazone, and metformin plus lifestyle on insulin sensitivity and β-cell function in TODAY. Diabetes Care. 2013. June;36(6):1749–57. 10.2337/dc12-2393 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Lascar N, Brown J, Pattison H, Barnett AH, Bailey CJ, Bellary S. Type 2 diabetes in adolescents and young adults. The Lancet Diabetes & Endocrinology. 2018. January;6(1):69–80. [DOI] [PubMed] [Google Scholar]
  • 36.Ma RCW, Chan JCN. Type 2 diabetes in East Asians: similarities and differences with populations in Europe and the United States. Ann NY Acad Sci. 2013. April 1;1281(1):64–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Lipska KJ, Ross JS, Miao Y, Shah ND, Lee SJ, Steinman MA. Potential Overtreatment of Diabetes Mellitus in Older Adults With Tight Glycemic Control. JAMA Intern Med. 2015. March 1;175(3):356–62. 10.1001/jamainternmed.2014.7345 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Ahlqvist E, Storm P, Käräjämäki A, Martinell M, Dorkhan M, Carlsson A, et al. Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables. The Lancet Diabetes & Endocrinology. 2018;6(5):361–369. 10.1016/S2213-8587(18)30051-2 [DOI] [PubMed] [Google Scholar]
  • 39.Wu H, Lau ESH, Yang A, Ma RCW, Kong APS, Chow E, et al. Trends in diabetes-related complications in Hong Kong, 2001–2016: a retrospective cohort study. Cardiovascular Diabetology. 2020. May 12;19(1):60 10.1186/s12933-020-01039-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Gao Z, Yan W, Fang Z, Zhang Z, Yuan L, Wang X, et al. Annual decline in β-cell function in patients with type 2 diabetes in China. Diabetes Metab Res Rev. 2020. June 8;e3364 10.1002/dmrr.3364 [DOI] [PubMed] [Google Scholar]
  • 41.U. K. Prospective Diabetes Study Group. U.K. Prospective Diabetes Study 16: Overview of 6 Years’ Therapy of Type II Diabetes: A Progressive Disease. Diabetes. 1995. November 1;44(11):1249–58. [PubMed] [Google Scholar]
  • 42.Althoff T, Sosič R, Hicks JL, King AC, Delp SL, Leskovec J. Large-scale physical activity data reveal worldwide activity inequality. Nature. 2017. July 10;547(7663):336–9. 10.1038/nature23018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Lu Y, Chen L, Yang Y, Gou Z. The Association of Built Environment and Physical Activity in Older Adults: Using a Citywide Public Housing Scheme to Reduce Residential Self-Selection Bias. Int J Environ Res Public Health. 2018. September 10;15(9):1973 10.3390/ijerph15091973 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Edelman SV, Polonsky WH. Type 2 Diabetes in the Real World: The Elusive Nature of Glycemic Control. Diabetes Care. 2017. November 1;40(11):1425–32. 10.2337/dc16-1974 [DOI] [PubMed] [Google Scholar]
  • 45.Carls GS, Tuttle E, Tan R-D, Huynh J, Yee J, Edelman SV, et al. Understanding the Gap Between Efficacy in Randomized Controlled Trials and Effectiveness in Real-World Use of GLP-1 RA and DPP-4 Therapies in Patients With Type 2 Diabetes. Diabetes Care. 2017. November 1;40(11):1469–78. 10.2337/dc16-2725 [DOI] [PubMed] [Google Scholar]
  • 46.Kim YG, Hahn S, Oh TJ, Kwak SH, Park KS, Cho YM. Differences in the glucose-lowering efficacy of dipeptidyl peptidase-4 inhibitors between Asians and non-Asians: a systematic review and meta-analysis. Diabetologia. 2013. April 1;56(4):696–708. 10.1007/s00125-012-2827-3 [DOI] [PubMed] [Google Scholar]
  • 47.Davis TME, Mulder H, Lokhnygina Y, Aschner P, Chuang L-M, Grado CAR, et al. Effect of race on the glycaemic response to sitagliptin: Insights from the Trial Evaluating Cardiovascular Outcomes with Sitagliptin (TECOS). Diabetes, Obesity and Metabolism. 2018;20(6):1427–34. 10.1111/dom.13242 [DOI] [PubMed] [Google Scholar]
  • 48.Ohn JH, Kwak SH, Cho YM, Lim S, Jang HC, Park KS, et al. 10-year trajectory of β-cell function and insulin sensitivity in the development of type 2 diabetes: a community-based prospective cohort study. The Lancet Diabetes & Endocrinology. 2016. January 1;4(1):27–34. [DOI] [PubMed] [Google Scholar]
  • 49.Ng MCY, Lee S-C, Ko GTC, Li JKY, So W-Y, Hashim Y, et al. Familial Early-Onset Type 2 Diabetes in Chinese Patients: Obesity and genetics have more significant roles than autoimmunity. Diabetes Care. 2001. April 1;24(4):663–71. 10.2337/diacare.24.4.663 [DOI] [PubMed] [Google Scholar]
  • 50.Hattersley AT, Patel KA. Precision diabetes: learning from monogenic diabetes. Diabetologia. 2017. May 1;60(5):769–77. 10.1007/s00125-017-4226-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Holman RR, Coleman RL, Chan JCN, Chiasson J-L, Feng H, Ge J, et al. Effects of acarbose on cardiovascular and diabetes outcomes in patients with coronary heart disease and impaired glucose tolerance (ACE): a randomised, double-blind, placebo-controlled trial. The Lancet Diabetes & Endocrinology. 2017. November;5(11):877–86. [DOI] [PubMed] [Google Scholar]
  • 52.Chiasson J-L, Josse RG, Gomis R, Hanefeld M, Karasik A, Laakso M. Acarbose for prevention of type 2 diabetes mellitus: the STOP-NIDDM randomised trial. The Lancet. 2002. June 15;359(9323):2072–7. [DOI] [PubMed] [Google Scholar]
  • 53.Lam KS, Tiu SC, Tsang MW, Ip TP, Tam SC. Acarbose in NIDDM patients with poor control on conventional oral agents. A 24-week placebo-controlled study. Diabetes Care. 1998. July;21(7):1154–8. 10.2337/diacare.21.7.1154 [DOI] [PubMed] [Google Scholar]
  • 54.Chan JC, Chan KW, Ho LL, Fuh MM, Horn LC, Sheaves R, et al. An Asian multicenter clinical trial to assess the efficacy and tolerability of acarbose compared with placebo in type 2 diabetic patients previously treated with diet. Asian Acarbose Study Group. Diabetes Care. 1998. July;21(7):1058–61. 10.2337/diacare.21.7.1058 [DOI] [PubMed] [Google Scholar]
  • 55.Davies MJ, D’Alessio DA, Fradkin J, Kernan WN, Mathieu C, Mingrone G, et al. Management of Hyperglycemia in Type 2 Diabetes, 2018. A Consensus Report by the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care. 2018. December;41(12):2669–701. 10.2337/dci18-0033 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 56.American Diabetes Association. 6. Glycemic Targets: Standards of Medical Care in Diabetes—2019. Diabetes Care. 2019. January 1;42(Supplement 1):S61–70. [DOI] [PubMed] [Google Scholar]
  • 57.Imran SA, Agarwal G, Bajaj HS, Ross S. Targets for Glycemic Control. Canadian Journal of Diabetes. 2018. April 1;42:S42–6. 10.1016/j.jcjd.2017.10.030 [DOI] [PubMed] [Google Scholar]
  • 58.National Institute for Health and Care Excellence. Type 2 diabetes in adults: management [Internet]. National Institute for Health and Care Excellence; 2015 Dec [cited 2019 Feb 26] p. 1–44. https://www.nice.org.uk/guidance/ng28/chapter/1-Recommendations#hba1c-measurement-and-targets
  • 59.Lim LL, Lau ESH, Kong APS, Davies MJ, Levitt NS, Eliasson B, et al. Aspects of Multicomponent Integrated Care Promote Sustained Improvement in Surrogate Clinical Outcomes: A Systematic Review and Meta-analysis. Diabetes Care. 2018. June 1;41(6):1312–20. 10.2337/dc17-2010 [DOI] [PMC free article] [PubMed] [Google Scholar]

Decision Letter 0

Artur Arikainen

24 Apr 2020

Dear Dr Ke,

Thank you for submitting your manuscript entitled "Age at Diagnosis, Glycemic Control, and Responses to Oral Glucose-Lowering Drugs in Type 2 Diabetes: A Population-Based Study" for consideration by PLOS Medicine.

Your manuscript has now been evaluated by the PLOS Medicine editorial staff as well as by an academic editor with relevant expertise and I am writing to let you know that we would like to send your submission out for external peer review.

However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.

Please re-submit your manuscript within two working days, i.e. by .

Login to Editorial Manager here: https://www.editorialmanager.com/pmedicine

Once your full submission is complete, your paper will undergo a series of checks in preparation for peer review. Once your manuscript has passed all checks it will be sent out for review.

Feel free to email us at plosmedicine@plos.org if you have any queries relating to your submission.

Kind regards,

Artur Arikainen,

Associate Editor

PLOS Medicine

Decision Letter 1

Artur Arikainen

1 Jun 2020

Dear Dr. Ke,

Thank you very much for submitting your manuscript "Age at Diagnosis, Glycemic Control, and Responses to Oral Glucose-Lowering Drugs in Type 2 Diabetes: A Population-Based Study" (PMEDICINE-D-20-01525R1) for consideration at PLOS Medicine.

Your paper was evaluated by a senior editor and discussed among all the editors here. It was also discussed with an academic editor with relevant expertise, and sent to independent reviewers, including a statistical reviewer. The reviews are appended at the bottom of this email and any accompanying reviewer attachments can be seen via the link below:

[LINK]

In light of these reviews, I am afraid that we will not be able to accept the manuscript for publication in the journal in its current form, but we would like to consider a revised version that addresses the reviewers' and editors' comments. Obviously we cannot make any decision about publication until we have seen the revised manuscript and your response, and we plan to seek re-review by one or more of the reviewers.

In revising the manuscript for further consideration, your revisions should address the specific points made by each reviewer and the editors. Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments, the changes you have made in the manuscript, and include either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please submit a clean version of the paper as the main article file; a version with changes marked should be uploaded as a marked up manuscript.

In addition, we request that you upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, please upload your figure files to the PACE digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at PLOSMedicine@plos.org.

We expect to receive your revised manuscript by Jun 22 2020 11:59PM. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

Please use the following link to submit the revised manuscript:

https://www.editorialmanager.com/pmedicine/

Your article can be found in the "Submissions Needing Revision" folder.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. For instructions see http://journals.plos.org/plosmedicine/s/submission-guidelines#loc-methods.

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

We look forward to receiving your revised manuscript.

Sincerely,

Artur Arikainen,

Associate Editor

PLOS Medicine

plosmedicine.org

-----------------------------------------------------------

Requests from the editors:

1. Please address all of the reviewers’ comments below.

2. Title: Please describe your study as ‘observational’, and include the cohort location ‘Hong Kong’.

3. Abstract:

a. Please move the limitations to the end of the ‘Methods and findings’ subsection.

b. Please include basic demographic information for the cohort (age, sex etc.).

c. Please quantify the main results with 95% CIs and p values.

4. Please include line numbers in the margin throughout your manuscript.

5. At this stage, we ask that you include a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract. Please see our author guidelines for more information: https://journals.plos.org/plosmedicine/s/revising-your-manuscript#loc-author-summary

6. Results: Please quantify all results with 95% CIs and p values.

7. Please avoid using the word “effect” or similar words when describing associative findings from your observational study.

8. Please remove the Data Availability, Funding, Competing Interests, and Author contributions from after your Acknowledgements – these will be taken from the submission form instead.

9. When completing the STROBE checklist, please use section and paragraph numbers, rather than page numbers. Please add the following statement, or similar, to the Methods: "This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 Checklist)."

10. Did your study have a prospective protocol or analysis plan? Please state this (either way) early in the Methods section.

a. If a prospective analysis plan (from your funding proposal, IRB or other ethics committee submission, study protocol, or other planning document written before analyzing the data) was used in designing the study, please include the relevant prospectively written document with your revised manuscript as a Supporting Information file to be published alongside your study, and cite it in the Methods section. A legend for this file should be included at the end of your manuscript.

b. If no such document exists, please make sure that the Methods section transparently describes when analyses were planned, and when/why any data-driven changes to analyses took place.

c. In either case, changes in the analysis-- including those made in response to peer review comments-- should be identified as such in the Methods section of the paper, with rationale.

11. Please state whether the patient data were anonymised at the time of access, or whether patients provided written informed consent for the use of their data.

Comments from the reviewers:

Reviewer #1: I confine my remarks to statistical aspects of this paper.

These were very well done and I recommend publication

Peter Flom

Reviewer #2:

This is a manuscript which describes the relationship between age at diagnosis, glycemic control, and responses to oral glucose-lowering drugs in type 2 diabetes in a population-based study from Hong Kong.

Analyses examining poor outcomes ( diabetes complications) among those with young onset diabetes versus older onset diabetes are tricky because is difficult to understand whether any relationship observed are due to someone particular about young onset diabetes or merely due to duration. This study partly avoids this conundrum and uses lifetime glycaemia as outcome.

The sample used in this study is very large and extracted from the Hong Kong Diabetes Surveillance Database and Hong Kong Diabetes Registry. The size of the sample is a real strength as other studies on young versus older onset type 2 diabetes are based on clinic samples of much smaller size.

Specific comments:

There are several statements in the abstract which do not have a statement of significance.

Figure 2 is difficult to understand. Does the direction and the colour of the arrow mean something? I assume it does but and I can guess what it means, but it should be specified more clearly. Have you thought about plotting trajectories of HbA1c in classes perhaps? I did not find this figure helpful.

Although the work on glucose lowering drugs among these groups is interesting, it seems a bit like a bit of an add on. Are medications free of charge in Hong Kong? Are there any health access issues which could influence these results? Some context should be provided for those who are not familiar with the health system in Hong Kong. When were SGLT2 inhibitor available in Hong Kong?

How was adherence and persistence dealt with?

The % missing on the tables could be removed. It does not really add anything.

The discussion could include some text about the mechanisms explaining why HbA1c deterioration is faster in young onset type 2 versus older onset type 2 diabetes.

Reviewer #3: The effect of age of onset of diabetes on long term outcomes is an interesting topic given that it may require differential management approaches based on individual risk. The rising prevalence of type 2 diabetes in youth is a matter of major concern. The clinical observation that diabetes in youth has a more aggressive phenotype requires better scrutiny especially to differentiate it from older onset type 2 diabetes. This study addresses 3 major aspects of this issue in relation to the glycemic burden, trajectory and response to treatment. The authors have done well in exploring these key issues in tis study.The manuscript is well written and the methods and results are presented with good clarity. The large sample size and the use of information from well established registers is a strength of this study. Discussion is well written and is balanced.

Couple of minor points : 1. As the cohort is mainly east asian, it would be good to add a statement re generalisability of these findings in other populations .

2. With regards to treatment comparisons, although the focus was on OGLD was insulin used in combination over the duration of follow up in this cohort? If so, it is worth mentioning given the rapid progression to insulin requirement in younger age groups. It is also important to discuss the point that 24% of younger diabetics were requiring insulin within 1 year of index date.

3. Reference to the effects of TZDs in younger age groups ( found to be effective on this group in combination with SU and Metformin) in the context of findings of the TODAY trial would be good.

Reviewer #4: PMEDICINE-D-20-01525R1

This is a nice well written paper that explores an area for which there are few data. A significant strength is the large study population and that the study employs valuable and well defined cohorts. The study adds to new knowledge by quantifying excess glycaemia exposure for younger onset type 2 diabetes and further examines the trajectory of glycaemic decline and relative efficacy of therapy. Data are real world and observational so that all conclusions should be seen in this context. Nevertheless the data as presented do align with previous studies in adolescent type 2 diabetes as expressed in the landmark TODAY study and the more recent RISE data. I cannot critique the appropriateness of the statistical methods in full. I do have a number of largely minor comments

1.The authors should comment on the possibility of ascertainment bias, especially the possibility of more severe young diabetes being referred to relevant service and be preferentially represented in the two datasets (especially the HKDR). Also in the HKDSD what is the possibility of those captured by metformin use having pre-diabetes and or PCOD and not diabetes per se.

2. Re glucose exposure outcome. Re mean observed HbA1c measured at standard intervals and number of observations standardised. How many HbAic observations were taken by age of diagnosis category? That the lifetime cumulative glucose exposure is greater in young onset compared to usual onset is not counter-intuitive given the longer duration of disease. Or was glucose exposure comparisons equated for duration in this study? Could this be clarified in the manuscript? It might be more informative to express this excess equated for duration of observation. Further how many younger onset individuals contributed to the HbAic data at different current ages up to age 75. This would add perspective on the strength of these data.

3. Re Response to OGLD . The delta Hba1c for any treatment is dependent on starting Hba1c. Can this be or was this adjusted for in this analysis. It is a significant effect modifier. Comment on the quality of the dispensing records would be helpful

4. Re the exclusion of patients in the HKDSD Appendix fig 4 . A large number (>20,000) were excluded as they had incomplete HbAic. It would be nice to know that these data were missing at random and would not have biased outcomes in any way.

5. The data are from a HK Chinese population. Could the authors comment on generalizability

6. The excess glycaemic deterioration in the young onset could be due to failure of treatment escalation, therapeutic inertia. This is an important point as data from primary care suggest a greater failure to recognise the importance of treatment escalation in youth. Could the authors comment or shed light on this possibility.

Any attachments provided with reviews can be seen via the following link:

[LINK]

Decision Letter 2

Artur Arikainen

14 Jul 2020

Dear Dr. Ke,

Thank you very much for re-submitting your manuscript "Age at Diagnosis, Glycemic Trajectories, and Responses to Oral Glucose-Lowering Drugs in Type 2 Diabetes: A Population-Based Observational Study in Hong Kong" (PMEDICINE-D-20-01525R2) for review by PLOS Medicine.

I have discussed the paper with my colleagues and the academic editor and it was also seen again by 3 reviewers. I am pleased to say that provided the remaining editorial and production issues are dealt with we are planning to accept the paper for publication in the journal.

The remaining issues that need to be addressed are listed at the end of this email. Any accompanying reviewer attachments can be seen via the link below. Please take these into account before resubmitting your manuscript:

[LINK]

Our publications team (plosmedicine@plos.org) will be in touch shortly about the production requirements for your paper, and the link and deadline for resubmission. DO NOT RESUBMIT BEFORE YOU'VE RECEIVED THE PRODUCTION REQUIREMENTS.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

In revising the manuscript for further consideration here, please ensure you address the specific points made by each reviewer and the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. If you haven't already, we ask that you provide a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract.

We expect to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We ask every co-author listed on the manuscript to fill in a contributing author statement. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT.

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Jul 21 2020 11:59PM.

Sincerely,

Artur Arikainen,

Associate Editor

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

1. Please update the Title to: “Age at Diagnosis, Glycemic Trajectories, and Responses to Oral Glucose-Lowering Drugs in Type 2 Diabetes in Hong Kong: A Population-Based Observational Study”

2. Please update the Short Title to: “Age at Diagnosis and Glycemic Trajectories in Type 2 Diabetes”

3. Abstract:

a. Please add additional limitation(s), eg. generalisability outside Hong Kong, to the ‘Methods and findings’ subsection.

b. Please do not use italics for emphasis, e.g., at line 38.

c. Please quote exact p values or p<0.001 (e.g., at line 49).

d. Line 51: Delete second “versus”.

e. Please provide quantitative data with 95% CIs and p values for each of the following statements:

i. “Age at diagnosis ≥60 years was associated with glycemic improvement.”

ii. “Responses to OGLDs differed by age at diagnosis (p<0.0001).”

iii. “Those with young-onset T2D had smaller A1C decrements for metformin-based combinations versus usual-onset T2D, but greater responses to other combinations containing sulfonylureas.”

f. Line 59: we suggest adapting the text to "In this study, we observed excess ..." or similar, and segueing into "... indicating that improved treatment strategies are needed in this setting."

4. Lines 423-424: Please avoid using the word ‘effectively’, and re-word as follows: “…our data suggest that most OGLDs regimens are associated with lower glucose levels in young-onset T2D.”

5. Discussion:

a. Line 344: Please clarify that similar findings have not been reported “to the authors’ knowledge”.

b. We note that treatment with GLP-1 agonists seemed low in this cohort - please perhaps comment on lack of information about and/or availability of newer drugs, either as a limitation of the study if these have become more widely used post-study, or of the health system if cost is an issue.

6. Re: reference 39 listed as in press, papers cannot be listed in the reference list until they have been accepted for publication or are otherwise publicly accessible (for example, in a preprint archive). The information may be cited in the text as a personal communication with the author if the author provides written permission to be named. Alternatively, please provide a different appropriate reference.

7. In the attached STROBE checklist, please remove page numbers, and refer to individual items by section (e.g., "Methods") and by paragraph number as at present.

----

Comments from Reviewers:

Reviewer #1: I had already accepted the earlier revision, so I recommend publication

Peter Flom

Reviewer #2: Thanks for addressing the comments and simplifying the figure.

Reviewer #3: I have now reviewed the manuscript after the revisions. I can see that the authors have addressed the comments from the reviewers satisfactorily. Figure 1 and Figure 2 are much better now. Discussion is well balanced. I would be delighted to see this work published.

Any attachments provided with reviews can be seen via the following link:

[LINK]

Decision Letter 3

Artur Arikainen

14 Aug 2020

Dear Dr. Ke,

On behalf of my colleagues and the academic editor, Dr. Sanjay Basu, I am delighted to inform you that your manuscript entitled "Age at Diagnosis, Glycemic Trajectories, and Responses to Oral Glucose-Lowering Drugs in Type 2 Diabetes in Hong Kong: A Population-Based Observational Study" (PMEDICINE-D-20-01525R3) has been accepted for publication in PLOS Medicine.

PRODUCTION PROCESS

Before publication you will see the copyedited word document (in around 1-2 weeks from now) and a PDF galley proof shortly after that. The copyeditor will be in touch shortly before sending you the copyedited Word document. We will make some revisions at the copyediting stage to conform to our general style, and for clarification. When you receive this version you should check and revise it very carefully, including figures, tables, references, and supporting information, because corrections at the next stage (proofs) will be strictly limited to (1) errors in author names or affiliations, (2) errors of scientific fact that would cause misunderstandings to readers, and (3) printer's (introduced) errors.

If you are likely to be away when either this document or the proof is sent, please ensure we have contact information of a second person, as we will need you to respond quickly at each point.

PRESS

A selection of our articles each week are press released by the journal. You will be contacted nearer the time if we are press releasing your article in order to approve the content and check the contact information for journalists is correct. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximize its impact.

PROFILE INFORMATION

Now that your manuscript has been accepted, please log into EM and update your profile. Go to https://www.editorialmanager.com/pmedicine, log in, and click on the "Update My Information" link at the top of the page. Please update your user information to ensure an efficient production and billing process.

Thank you again for submitting the manuscript to PLOS Medicine. We look forward to publishing it.

Best wishes,

Artur Arikainen,

Associate Editor

PLOS Medicine

plosmedicine.org

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    S1 Appendix

    (PDF)

    Click here for additional data file. (435.2KB, pdf)
    S1 Analysis plan

    (PDF)

    Click here for additional data file. (265.8KB, pdf)
    S1 STROBE checklist. STROBE, Strengthening the Reporting of Observational Studies in Epidemiology.

    (DOCX)

    Click here for additional data file. (33KB, docx)

    Data Availability Statement

    Due to local laws and regulations regarding the use and distribution of personal data, the data used in the present study cannot be deposited in a public repository. Data access can be applied for through the Data Sharing Portal of the Hong Kong Hospital Authority (https://www3.ha.org.hk/data/DCL/Index/).

    Articles from PLoS Medicine are provided here courtesy of PLOS

    RESOURCES