Abstract
Background
Limited studies have explored the predictive efficiency of prediabetes based on two definitions for diabetes among Chinese middle-aged and older populations with prediabetes.
Objective
To evaluate the predictive efficiency of prediabetes based on two definitions for diabetes and the clinical and public health benefit in Chinese middle-aged and older populations.
Design
A 5-year cohort study from the China Health and Retirement Longitudinal Study.
Participants
A total of 5208 participants who had blood sample data at baseline in 2011.
Main Measures
The exposure was prediabetes based on American Diabetes Association (ADA) and World Health Organization (WHO) definition. The main outcome was incident diabetes. The ability of prediabetes for predicting diabetes was assessed by sensitivity, specificity, positive predictive value, and negative predictive value. Cox proportional hazards regression was used to explore the associations between prediabetes and the 5-year risk of diabetes and all-cause mortality.
Key Results
Among those with prediabetes according to the ADA definition, only 426 (15.45%) with baseline prediabetes progressed to total diabetes, while according to the WHO definition, 208 (21.89%) progressed to total diabetes. In terms of the ability of predicting the incident total diabetes in 5 years, the ADA definition has a higher sensitivity than the WHO definition (70.76% versus 34.55%, P < 0.001), while the WHO definition has a higher specificity than the ADA definition (84.09% versus 49.35%, P < 0.001). Positive predictive values based on the two definitions were low (< 24%); negative predictive values were high (> 90%).
Conclusions
Neither definition of prediabetes is robust for predicting diabetes development in Chineses middle-aged and older populations.
Supplementary Information
The online version contains supplementary material available at 10.1007/s11606-022-07731-x.
KEY WORDS: Prediabetes, Type 2 diabetes, Chinese population, Aging, Diagnostic criteria
INTRODUCTION
Type 2 diabetes mellitus, a chronic metabolic disorder characterized by the damage of an individual’s ability to regulate blood glucose,1 results in long-time severe dysfunction of various organs, such as the blood vessels, eyes, kidneys, and heart.2 Type 2 diabetes has become a rapidly growing global public health challenge, with the prevalence rising steeply in recent decades.3 Thus, it is significant to take effective interventions to prevent and delay its progression.
Prediabetes, defined as blood glucose concentrations between normal and clinical diabetes concentrations, is the intermediate stage of hyperglycemia.4 Previous studies indicated that individuals with prediabetes had higher risks of developing diabetes in the future.5,6 A review from the American Diabetes Association (ADA) demonstrated the association between impaired fasting glucose and impaired glucose tolerance and diabetes.5 Besides, according to a population-based cohort study, individuals with impaired fasting glucose had a higher diabetes incidence compared with individuals with normoglycemia at baseline.6 Early diagnosis of prediabetes could be beneficial in the identification of individuals who are at a higher risk of progressing to diabetes in the future, so that they may benefit from early interventions. However, another research found that among those older adults with prediabetes, regression to normoglycemia was more common than progression to diabetes. Therefore, prediabetes may not be robust for predicting diabetes progression in the older population.7
In recent years, the large increase in the prevalence of diabetes in the Chinese population has imposed a heavy burden to the Chinese public health system. Therefore, it is of great value to take effective measures to prevent the development of diabetes. In addition, compared with other populations, the Chinese population has different patterns of type 2 diabetes as they have a lower body mass index (BMI)8,9 and a higher risk of developing renal complications.10 Moreover, the incidence of transition from prediabetes to diabetes and the prognostic implications of prediabetes in Chinese middle-aged and older populations was poorly characterized. Therefore, the ability of prediabetes for predicting diabetes in Chinese middle-aged and older populations remains unclear.
This population-based study used nationally representative survey data to investigate the risk of progression from prediabetes to diabetes, and compare the predictive efficiency of two prediabetes definitions to explore which one has more clinical and public health benefit in China.
METHODS
Study Population
The China Health and Retirement Longitudinal Study (CHARLS) is a nationally representative longitudinal survey among the Chinese population aged ≥ 45 years.11 The survey adopted multistage probability sampling. Stratified by region and within region by urban districts or rural counties and per capita statistics on gross domestic product (GDP), a total of 150 county-level units were randomly chosen from all county-level units not including Tibet by a probability-proportional-to-size (PPS) sampling technique. The survey covered objective and self-reported measures of health and assessments of social, economic, and health circumstances of community residents. The participants were followed every 2 or 3 years, with physical measurements at each follow-up visit and blood sample collection every two follow-up visits. To date, four times of surveys (i.e., visit 1 in 2011, visit 2 in 2013, visit 3 in 2015, and visit 4 in 2018) have been conducted. At visit 1 in 2011 and visit 3 in 2015, the blood sample data were collected. Also, the death status of the participants was investigated during each follow-up.
All participants who had blood sample data at visit 1 were included. The exclusion criteria were participants who (1) had no hemoglobin A1c (HbA1c) or fasting plasma glucose (FPG) measurements at visit 1 or visit 3 (except for those who died after visit 1); (2) fasted for less than 8 h at visit 1 or visit 3 (except for those who died after visit 1); (3) had a history of diagnosed diabetes before visit 1 (self-reported physician diagnosis or glucose-lowering medication use); (4) had an HbA1c level ≥ 6.5% or FPG level ≥ 126 mg/dL at visit 1; and (5) were alive but did not attend visit 3. This study was approved by the Ethical Review Committee of Peking University (IRB00001052-11015), and written informed consent was obtained from each participant.
Definition of Prediabetes
The exposure considered in this study was prediabetes at baseline. We used two definitions of prediabetes. One was based on an HbA1c level of 5.7–6.4% or FPG level of 100–125 mg/dL according to the ADA definition of prediabetes,2 and the other one was based on an FPG level of 110–125 mg/dL according to the World Health Organization (WHO) definition of prediabetes.12
Outcomes
The primary outcome was incident total diabetes (i.e., a self-reported physician diagnosis, glucose-lowering medication use, HbA1c level ≥ 6.5%, or FPG level ≥ 126 mg/dL) at visit 3. The secondary outcomes were incident diagnosed diabetes (i.e., based solely on a self-reported physician diagnosis or glucose-lowering medication use) and all-cause mortality between visit 1 and visit 3. Since there was no blood sample data in visit 4 in 2018, only all-cause mortality between visit 1 and visit 4 was considered as a secondary outcome as well.
Statistical Analysis
The characteristics of participants at baseline were reported according to the ADA and WHO definitions of prediabetes. The ability of the different prediabetes definitions for predicting 5-year risk of total diabetes and diagnosed diabetes was assessed by calculating sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). The sensitivity, specificity, PPV, and NPV in different age groups (< 65 years and ≥ 65 years) were calculated as well.
Cox proportional hazards regression was used to generate hazard ratios (HRs) and 95% confidence intervals (CIs) for the associations between baseline prediabetes and the risk of progressing to total diabetes, diagnosed diabetes, and mortality, with age, sex, and body mass index (BMI) (underweight: < 18.5 kg/m2, normal: 18.5 to 24 kg/m2, overweight: 24 to 28 kg/m2, obesity: ≥ 28 kg/m2) adjusted as covariates. The competing risk of death was considered by using the Fine and Gray method to generate subhazard ratios (sHRs).13 The Fine and Gray method adopted competing-risks regression which posits a model for the subhazard function of a failure event of primary interest (i.e., diabetes in this study). The standard Cox regression is appropriate only for (1) the situation where competing events (i.e., death in this study) do not occur or (2) the effects of covariates on the incidence-rate curves are hard to quantify. Therefore, when the competing failure events (i.e., death in this study) can impede the event of primary interest (i.e., diabetes in this study), it may be more appropriate to use competing-risk Cox regression.13 Kaplan-Meier survival curves were plotted as well. To help evaluate the necessity of changing the lower limit of FPG in the prediabetes definition from 110 to 100 mg/dL, the risks of total diabetes, diagnosed diabetes, and mortality between visit 1 and visit 3 as well as the risk of mortality between visit 1 and visit 4 were also compared between participants with a baseline FPG level of 100–110 mg/dL and those with a baseline FPG level < 100 mg/dL. A sensitivity analysis was done to take the dropouts due to death or loss to follow-up into consideration by using the inverse probability weighting approach. All statistical analyses were done by Stata 16.0 (StataCorp LLC). If a participant had missing information of certain covariates, the participant was not included in the analysis containing the covariates. Two-sided P < 0.05 was considered statistically significant.
RESULTS
Characteristics of Study Participants
A total of 5208 participants (mean [SD] age, 59.59 [9.29] years; 2781 [53.40%] female) were included in this study (Table 1). The detailed flow chart of selecting participants can be found in Supplementary Figure 1. According to the ADA definition, there were 2450 participants with normoglycemia and 2758 with prediabetes (Table 1). According to the WHO definition, there were 4258 participants with normoglycemia and 950 with prediabetes (Table 1). The mean age and sex distribution between participants with normoglycemia and prediabetes was similar (Table 1).
Table 1.
The Baseline Characteristics of Study Participants
| Characteristic | ADA definition of prediabetes | WHO definition of prediabetes | Total | ||
|---|---|---|---|---|---|
| Normoglycemia | Prediabetes | Normoglycemia | Prediabetes | ||
| No. (%) of participants | 2450 (47.04) | 2758 (52.96) | 4258 (81.76) | 950 (18.24) | 5208 (100) |
| FPG, median (range), mg/dL | 93 (88–96) | 106 (102–112) | 97 (91–103) | 115 (112–119) | 100 (93–107) |
| Age, mean (SD), years | 59.06 (9.45) | 60.06 (9.12) | 59.40 (9.32) | 60.43 (9.12) | 59.59 (9.29) |
| Sex | |||||
| Male | 1138 (46.45) | 1289 (46.74) | 1966 (46.17) | 461 (48.53) | 2427 (46.60) |
| Female | 1312 (53.55) | 1469 (53.26) | 2292 (53.83) | 489 (51.47) | 2781 (53.40) |
Five-Year Development of Normoglycemia and Prediabetes
According to the ADA definition, 2450 participants were with normoglycemia at baseline. Among them, 924 (37.71%) stayed as normoglycemia, 1167 (47.63%) progressed to prediabetes, 176 (7.18%) progressed to diabetes, and 183 (7.47%) died at visit 3. According to the WHO definition, 4258 participants were with normoglycemia at baseline. Among them, 3457 (81.19%) stayed as normoglycemia, 106 (2.49%) progressed to prediabetes, 394 (9.25%) progressed to diabetes, and 301 (7.07%) died at visit 3.
A total of 547 participants (19.83%) with baseline prediabetes based on the ADA definition regressed to normoglycemia at visit 3, 426 (15.45%) progressed to total diabetes, 186 (6.74%) died, and 1599 (57.98%) stayed prediabetes (Fig. 1). Among 950 participants with baseline prediabetes based on WHO definition, 604 (63.58%) regressed to normoglycemia at visit 3, 208 (21.89%) progressed to total diabetes, 68 (7.16%) died, and 70 (7.37%) stayed prediabetes (Fig. 1). Progression from prediabetes to diabetes was more common in participants with prediabetes based on the WHO definition than the ADA definition (P < 0.001), but progression from prediabetes to mortality was comparable in participants with baseline prediabetes based on the two definitions (P = 0.656).
Fig. 1.
The 5-year progression of prediabetes among Chinese population aged ≥ 45 years.
Ability of Different Prediabetes Definitions for Predicting Diabetes in 5 Years
In terms of identifying incident total diabetes, sensitivity was higher for baseline prediabetes based on the ADA definition than that based on the WHO definition (70.76% versus 34.55%, P < 0.001, Table 2), but specificity was higher for baseline prediabetes based on the WHO definition than that based on the ADA definition (84.09% versus 49.35%, P < 0.001). Positive predictive values based on the two definitions were low (< 24%), and negative predictive values were high (> 90%). In terms of identifying incident diagnosed diabetes, sensitivity was again higher for baseline prediabetes based on the ADA definition than that based on the WHO definition (69.40% versus 36.64%, P < 0.001, Table 2), but specificity was higher for baseline prediabetes based on the WHO definition than that based on the ADA definition (82.70% versus 47.67%, P < 0.001). Positive predictive values based on the two definitions were low (< 10%), and negative predictive values were high (> 96%). The sensitivity, specificity, PPV, and NPV in different age groups (< 65 years and ≥ 65 years) can been seen in supplementary Table 1.
Table 2.
Performance of Different Prediabetes Definitions in Chinese Middle-Aged and Older Adults for Identifying Incident Total Diabetes and Incident Diagnosed Diabetes Within 5-Years
| Prediabetes definition | No. | Diagnostic performance, % (95% CI) | No. | Diagnostic performance, % (95% CI) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No incident total diabetes | Incident total diabetes | Sensitivity | Specificity | Positive predictive value | Negative predictive value | No incident diagnosed diabetes | Incident diagnosed diabetes | Sensitivity | Specificity | Positive predictive value | Negative predictive value | |
| ADA definition | ||||||||||||
| Normoglycemia | 2091 | 176 | NA | NA | NA | NA | 2196 | 71 | NA | NA | NA | NA |
| Prediabetes | 2146 | 426 | 70.76 (69.48–72.05) | 49.35 (47.94–50.76) | 16.56 (15.52–17.61) | 92.24 (91.48–92.99) | 2411 | 161 | 69.40 (68.10–70.69) | 47.67 (46.26–49.07) | 6.26 (5.58–6.94) | 96.87 (96.38–97.36) |
| WHO definition | ||||||||||||
| Normoglycemia | 3563 | 394 | NA | NA | NA | NA | 3810 | 147 | NA | NA | NA | NA |
| Prediabetes | 674 | 208 | 34.55 (33.21–35.89) | 84.09 (83.06–85.12) | 23.58 (22.39–24.78) | 90.04 (89.20–90.89) | 797 | 85 | 36.64 (35.28–38.00) | 82.70 (81.63–83.77) | 9.64 (8.81–10.47) | 96.29 (95.75–96.82) |
| P value | < 0.001 | < 0.001 | < 0.001 | 0.529 | < 0.001 | < 0.001 | 0.003 | 0.881 | ||||
NA Not applicable
Baseline Prediabetes and 5-Year Risk of Diabetes and Mortality
According to the Cox proportional hazards regressions for the association of prediabetes with incident total diabetes, incident diagnosed diabetes, and mortality, baseline prediabetes based on both definitions was associated with incident total diabetes and incident diagnosed diabetes within a 5-year follow-up, but not associated with risk of mortality within 5 years (Table 3). The HRs for all outcomes were slightly higher for baseline prediabetes based on the WHO definition than baseline prediabetes based on the ADA definition. The sHRs which were used to account for the competing risk of death were similar to the HRs (Table 3). The associations in sensitivity analysis using the inverse probability weighting approach were also similar to the main results (Table 4). In terms of baseline prediabetes and risk of mortality within 8 years, baseline prediabetes based on both definitions was not associated with a higher risk of mortality within 8 years (supplementary Table 2 and supplementary Table 3). The Kaplan-Meier survival curves can be seen in supplementary Figure 2.
Table 3.
Five-Year Risk of Incident Total Diabetes, Incident Diagnosed Diabetes, and Mortality in Chinese Middle-Aged and Older Adults with Prediabetes
| Prediabetes definition | Group | No. of events/participants | HR (95% CI) | sHR (95% CI) |
|---|---|---|---|---|
| Incident total diabetes | ||||
| ADA definition | Normoglycemia | 176/2450 | Reference | Reference |
| Prediabetes | 426/2758 | 1.83 (1.53–2.19) | 1.86 (1.57–2.21) | |
| WHO definition | Normoglycemia | 394/4258 | Reference | Reference |
| Prediabetes | 208/950 | 2.04 (1.72–2.41) | 2.06 (1.76–2.42) | |
| Incident diagnosed diabetes | ||||
| ADA definition | Normoglycemia | 71/2450 | Reference | Reference |
| Prediabetes | 161/2758 | 1.72 (1.30–2.28) | 1.74 (1.32–2.30) | |
| WHO definition | Normoglycemia | 147/4258 | Reference | Reference |
| Prediabetes | 85/950 | 2.23 (1.70–2.92) | 2.26 (1.73–2.93) | |
| All-cause mortality | ||||
| ADA definition | Normoglycemia | 183/2450 | Reference | NA |
| Prediabetes | 186/2758 | 0.85 (0.68–1.07) | NA | |
| WHO definition | Normoglycemia | 301/4258 | Reference | NA |
| Prediabetes | 68/950 | 0.89 (0.66–1.19) | NA | |
Note: The HRs and sHRs were adjusted with age, sex, and BMI.
NA Not applicable
Table 4.
Sensitivity Analysis for 5-Year Risk of Incident Total Diabetes, Incident Diagnosed Diabetes, and Mortality in Chinese Middle-Aged and Older Adults with Prediabetes
| Prediabetes definition | Group | No. of events/participants | HR (95% CI) | sHR (95% CI) |
|---|---|---|---|---|
| Incident total diabetes | ||||
| ADA definition | Normoglycemia | 176/2450 | Reference | Reference |
| Prediabetes | 426/2758 | 1.69 (1.06–2.71) | 1.71 (1.07–2.74) | |
| WHO definition | Normoglycemia | 394/4258 | Reference | Reference |
| Prediabetes | 208/950 | 1.90 (1.41–2.56) | 1.91 (1.41–2.57) | |
| Incident diagnosed diabetes | ||||
| ADA definition | Normoglycemia | 71/2450 | Reference | Reference |
| Prediabetes | 161/2758 | 1.78 (1.21–2.62) | 1.80 (1.22–2.64) | |
| WHO definition | Normoglycemia | 147/4258 | Reference | Reference |
| Prediabetes | 85/950 | 2.33 (1.67–3.24) | 2.34 (1.68–3.25) | |
| All-cause mortality | ||||
| ADA definition | Normoglycemia | 183/2450 | Reference | NA |
| Prediabetes | 186/2758 | 0.93 (0.70–1.24) | NA | |
| WHO definition | Normoglycemia | 301/4258 | Reference | NA |
| Prediabetes | 68/950 | 1.05 (0.74–1.50) | NA | |
Note: The HRs and sHRs were adjusted with age, sex, and BMI
Compared with the participants with a baseline FPG level < 100 mg/dL, the participants with a baseline FPG level of 100–110 mg/dL had a higher risk of developing incident total diabetes within 5 years (HR 1.42; 95% CI 1.17–1.73). But the risks of progressing to incident diagnosed diabetes or death within 5 years and death within 8 years were similar between the two groups (Table 5 and supplementary Table 4).
Table 5.
Five-Year Risk of Incident Total Diabetes, Incident Diagnosed Diabetes, and Mortality in Chinese Middle-Aged and Older Adults with Fasting Glucose of 100–110 mg/dL Versus Those with Fasting Glucose < 100 mg/dL
| Level of fasting glucose | No. of events/participants | HR (95% CI) | sHR (95% CI) |
|---|---|---|---|
| Incident total diabetes | |||
| < 100 mg/dL | 191/2543 | Reference | Reference |
| 100–110 mg/dL | 221/1810 | 1.42 (1.17–1.73) | 1.44 (1.19–1.74) |
| Incident diagnosed diabetes | |||
| < 100 mg/dL | 77/2543 | Reference | Reference |
| 100–110 mg/dL | 78/1810 | 1.26 (0.92–1.74) | 1.28 (0.94–1.75) |
| All-cause mortality | |||
| < 100 mg/dL | 190/2543 | Reference | NA |
| 100–110 mg/dL | 114/1810 | 0.82 (0.64-1.06) | NA |
Note: The HRs and sHRs were adjusted with age, sex, and BMI
NA Not applicable
DISCUSSION
This national population-based study indicated that prediabetes was prevalent in Chinese middle-aged and older individuals. Besides, prediabetic individuals did have a higher risk of developing diabetes in the future. However, in the prediabetic participants, regression to normoglycemia or maintaining prediabetes was more common than progression to diabetes, regardless of the definitions of prediabetes during the 5-year follow-up period. Moreover, the predictive ability of the two definitions was not good enough.
There are two definitions for prediabetes used in this analysis. According to the WHO definition, the prevalence of prediabetes in Chinese middle-aged and older people was less than 20%, while according to the ADA definition, the prevalence almost tripled; therefore, there was a large gap in the prevalence of prediabetes based on the two definitions. Similar findings indicating that prediabetes was highly prevalent were also observed in other studies conducted in Chinese or other racial groups.7,14,15 However, the prevalence of prediabetes varied between different racial groups. For example, a cohort study in the USA reported a prevalence of around 40% according to the ADA definition.15 A community-based study conducted in Saudi Arabia reported that the prevalence of prediabetes was around 10% based on the ADA definition.16 Moreover, another study suggested that around 20% of middle-aged and older individuals in Ireland developed prediabetes based on the ADA definition.17 Such great disparities in the prevalence of prediabetes suggested that race may have an influence on the prevalence of prediabetes, and the diagnostic definition formulated based on populations in other countries may not be suitable for the Chinese population.2,18
This study indicated that individuals developing prediabetes did have a higher risk of progressing into diabetes than those with normoglycemia, which was consistent with previous research.19,20 Prediabetes refers to an abnormal physiologic state with impaired glucose regulation ability which means patients suffering prediabetes have increased insulin resistance and decreased insulin secretion.4 The increase in insulin resistance in prediabetic individuals is followed by the dysfunction of β-cells, and the progression from prediabetes to diabetes occurs when the function of β-cells could not overcome insulin resistance.21,22 For some prediabetic patients, an increased blood glucose concentration can last for 13 years before developing into diabetes,23 but the complications of diabetes can still be observed in some prediabetic individuals.24,25 Thus, it is of great importance to identify middle-aged and older populations with prediabetes who are at an increasing risk of developing diabetes, and then take strict lifestyle interventions to reduce the risk of developing diabetes.
Although it is beneficial to recognize those suffering from prediabetes who are at increasing risks of developing diabetes, neither definition of prediabetes was robust for predicting future diabetes progression. First, no matter which diagnostic definition the diagnosis was based on, only a small proportion of prediabetic individuals progressed to diabetes. This finding was consistent with prior studies.7,26,27 Besides, among the individuals with normoglycemia, around 10% of them progressed into diabetes in 5 years. Such result also indicated that prediabetes based on the ADA and WHO definitions may ignore part of the individuals who have the probability of progressing to diabetes in the future, and prediabetes according to both definitions might not be the robust criteria. Moreover, according to the ADA definition, the number of prediabetic individuals was approximately three times those defined by WHO at baseline, but the number of individuals who progressed to diabetes in 5 years was only around twice as much as those diagnosed by the WHO definition. Thus, it may not be necessary to enlarge the FPG threshold by 10 mg/dL according to the ADA definition. Furthermore, although participants with a baseline FPG level of 100–110 mg/dL had a higher risk of developing incident total diabetes than those with a baseline FPG level < 100 mg/dL, the risks of progressing to incident diagnosed diabetes or death showed no difference between the two groups. Considering that enlarging the FPG threshold by 10 mg/dL may place an unnecessary burden on the individuals themselves and the public health care system, such enlargement might have little clinical and public health benefit.
Second, in terms of the ability of the two prediabetes definitions for predicting diabetes development, both the specificity and sensitivity of the two definitions did not perform well (34–84%), and the positive predictive values based on the two definitions were even lower than 25%, indicating that both definitions were not suitable for predicting the future development of diabetes. Moreover, another study found that individuals suffering prediabetes were more inclined to reverse to normoglycemia or stay the same rather than progressed into diabetes which also indicated that prediabetes based on the WHO and ADA definition may not be the robust criteria for predicting the development of diabetes in the older population.7 Accounting for this result, it is non-essential to take aggressive interventions like pharmacologic interventions to prevent type 2 diabetes in middle-aged and older prediabetic populations which might cause an unnecessary health burden.28 For such individuals, self-management like lifestyle change would be better.
Third, compared with populations in Western countries, Chinese individuals have specific metabolic characteristics. One is that the Chinese population can develop type 2 diabetes at a lower BMI level8,9,29,30 than the European and American populations, and have a tendency for visceral adiposity and a higher proportion of body fat at the same BMI compared with other populations.29,31 Additionally, Chinese individuals have a high prevalence of isolated postprandial hyperglycemia which accounted for 50% of the population with diabetes.32,33 Moreover, Asians are more inclined to have pancreatic β-cell secretory defects and finally progress to β-cell dysfunction.8,34 These variations indicated that Chinese diabetic individuals might have different glucose metabolism patterns compared with Western populations, and such disparities might have significant implications in the selection of diagnostic criteria and cut-off points of prediabetes. Considering that studies conducted in Asian populations have limited sample sizes, the definitions of prediabetes based on the ADA or WHO guidelines may not be suitable for Chinese individuals. This finding also suggested that further studies should develop better predictive models of prediabetes in Chinese middle-aged and older individuals.
This longitudinal study utilized the nationally representative data, which can provide enough statistical power to explore the risk of progression from prediabetes to diabetes and compared the two most commonly used definitions in China. But some limitations of our study merit consideration. First, an oral glucose-tolerance test was not conducted in CHARLS. But according to another high-quality study which used the ADA definition and conducted the oral glucose-tolerance test, the prevalence of prediabetes in Chinese individuals aged 40 and older was 52.6%,14 which was consistent with this study. Therefore, the effect of not including the oral glucose-tolerance test was mild. Besides, some high-quality studies based on CHARLS also used FPG and HbA1c level as the definition of diabetes and prediabetes which also indicated the reliability of these data.35–37 Second, part of the individuals with diabetes was identified by self-reported diagnosis in this study. However, according to previous validation studies, the self-reports of common chronic diseases were reliable and well-accepted.38,39 Also, a large number of high-quality studies based on CHARLS also utilized self-reported diagnosis.35,40 Third, since the blood sample data were only collected at visit 1 and visit 3 but not at visit 2, therefore, in terms of the incident total diabetes, the “survival time” considered in the Cox regression may be relatively rough. But our main purpose was to detect the risk signal but not the absolute risk.
CONCLUSION
This study demonstrated that the prevalence of prediabetes in Chinese middle-aged and older populations was high. But neither the ADA nor the WHO definition has good predictive ability for diabetes, suggesting that both definitions are not robust for prediabetes in the Chinese population. Further studies should explore the better prediabetes definition in Chinese middle-aged and older individuals.
Supplementary information
(DOCX 229 kb)
Acknowledgements
This work was supported by the National Natural Science Foundation (Grant number 81922016 and 81870607), Shandong Provincial Natural Science Foundation (Grant number ZR2019JQ25), National Key R&D Program of China (Grant number 2017YFC0908900), and Innovation Fund for Outstanding Doctoral Candidates of Peking University Health Science Center (China) (grant number not applicable). The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Author Contribution
S.F.W., S.Y.Z., and Y.F.S. took responsibility for the study conception and design. L.X., S.F.W., and L.L.L. contributed to data collection and assembly. L.X., S.F.W., and L.L.L. contributed to data verification. L.X., H.S., S.F.W., S.Y.Z., and Y.F.S. contributed to data interpretation. L.X. contributed to statistical analysis. H.S. and L.X. were responsible for the manuscript draft. All of the authors contributed to the manuscript review and revision, and approval of submission.
Declarations
Conflict of Interest
The authors declare that they do not have a conflict of interest.
Footnotes
Lu Xu and Hang Sun are joint first authors.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Siyan Zhan, Email: siyan-zhan@bjmu.edu.cn.
Shengfeng Wang, Email: shengfeng1984@126.com.
Yongfeng Song, Email: syf198506@163.com.
References
- 1.Kaul K, Tarr JM, Ahmad SI, Kohner EM, Chibber R. Introduction to diabetes mellitus. Adv Exp Med Biol. 2012;771:1–11. doi: 10.1007/978-1-4614-5441-0_1. [DOI] [PubMed] [Google Scholar]
- 2.American Diabetes Association Diagnosis and classification of diabetes mellitus. Diabetes Care. 2013;36(Supplement_1):S67–S74. doi: 10.2337/dc13-S067. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Vos T, Abajobir AA, Abate KH, et al. Global, regional, and national incidence, prevalence, and years lived with disability for 328 diseases and injuries for 195 countries, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. The Lancet. 2017;390(10100):1211–1259. doi: 10.1016/S0140-6736(17)32154-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tabák AG, Herder C, Rathmann W, Brunner EJ, Kivimäki M. Prediabetes: a high-risk state for diabetes development. The Lancet. 2012;379(9833):2279–2290. doi: 10.1016/S0140-6736(12)60283-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Nathan DM, Davidson MB, DeFronzo RA, et al. Impaired fasting glucose and impaired glucose tolerance: implications for care. Diabetes Care. 2007;30(3):753–759. doi: 10.2337/dc07-9920. [DOI] [PubMed] [Google Scholar]
- 6.Forouhi NG, Luan J, Hennings S, Wareham NJ. Incidence of type 2 diabetes in England and its association with baseline impaired fasting glucose: the Ely study 1990-2000. Diabet Med. 2007;24(2):200–207. doi: 10.1111/j.1464-5491.2007.02068.x. [DOI] [PubMed] [Google Scholar]
- 7.Rooney MR, Rawlings AM, Pankow JS, et al. Risk of progression to diabetes among older adults with prediabetes. JAMA Intern Med. 2021;181(4):511. doi: 10.1001/jamainternmed.2020.8774. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Yoon KH, Lee JH, Kim JW, et al. Epidemic obesity and type 2 diabetes in Asia. The Lancet. 2006;368(9548):1681–1688. doi: 10.1016/S0140-6736(06)69703-1. [DOI] [PubMed] [Google Scholar]
- 9.Hu C, Jia W. Diabetes in China: epidemiology and genetic risk factors and their clinical utility in personalized medication. Diabetes. 2018;67(1):3–11. doi: 10.2337/dbi17-0013. [DOI] [PubMed] [Google Scholar]
- 10.Chi ZS, Lee ET, Lu M, Keen H, Bennett PH. Vascular disease prevalence in diabetic patients in China: standardised comparison with the 14 centres in the WHO multinational study of vascular disease in diabetes. Diabetologia. 2001;44(S2):S82–S86. doi: 10.1007/PL00002944. [DOI] [PubMed] [Google Scholar]
- 11.Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China Health and Retirement Longitudinal Study (CHARLS) Int J Epidemiol. 2012;43(1):61–68. doi: 10.1093/ije/dys203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.World Health Organization. Use of glycated haemoglobin (HbA1c) in the diagnosis of diabetes mellitus: abbreviated report of a WHO Consultation. 2011. [PubMed]
- 13.Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94(446):496–509. doi: 10.1080/01621459.1999.10474144. [DOI] [Google Scholar]
- 14.Xu Y. Prevalence and control of diabetes in Chinese adults. JAMA. 2013;310(9):948. doi: 10.1001/jama.2013.168118. [DOI] [PubMed] [Google Scholar]
- 15.Ali MK, Bullard KM, Saydah S, Imperatore G, Gregg EW. Cardiovascular and renal burdens of prediabetes in the USA: analysis of data from serial cross-sectional surveys, 1988–2014. Lancet Diabetes Endocrinol. 2018;6(5):392–403. doi: 10.1016/S2213-8587(18)30027-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bahijri SM, Jambi HA, Al Raddadi RM, Ferns G, Tuomilehto J. The prevalence of diabetes and prediabetes in the adult population of Jeddah, Saudi Arabia- a community-based survey. Barengo NC, ed. PLOS ONE. 2016;11(4):e0152559. doi: 10.1371/journal.pone.0152559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Buckley CM, Madden J, Balanda K, et al. Pre-diabetes in adults 45 years and over in Ireland: the Survey of Lifestyle, Attitudes and Nutrition in Ireland 2007. Diabet Med. 2013;30(10):1198–1203. doi: 10.1111/dme.12226. [DOI] [PubMed] [Google Scholar]
- 18.World Health Organization. Definition and diagnosis of diabetes mellitus and intermediate hyperglycemia; 2006. https://www.who.int/publications/i/item/definition-and-diagnosis-of-diabetes-mellitus-and-intermediate-hyperglycaemia.
- 19.Anjana RM, Shanthi Rani CS, Deepa M, et al. Incidence of diabetes and prediabetes and predictors of progression among Asian Indians: 10-year follow-up of the Chennai Urban Rural Epidemiology Study (CURES) Diabetes Care. 2015;38(8):1441–1448. doi: 10.2337/dc14-2814. [DOI] [PubMed] [Google Scholar]
- 20.Gerstein HC, Santaguida P, Raina P, et al. Annual incidence and relative risk of diabetes in people with various categories of dysglycemia: a systematic overview and meta-analysis of prospective studies. Diabetes Res Clin Pract. 2007;78(3):305–312. doi: 10.1016/j.diabres.2007.05.004. [DOI] [PubMed] [Google Scholar]
- 21.Rett K, Gottwald-Hostalek U. Understanding prediabetes: definition, prevalence, burden and treatment options for an emerging disease. Curr Med Res Opin. 2019;35(9):1529–1534. doi: 10.1080/03007995.2019.1601455. [DOI] [PubMed] [Google Scholar]
- 22.Fonseca VA. Defining and characterizing the progression of type 2 diabetes. Diabetes Care. 2009;32(suppl_2):S151–S156. doi: 10.2337/dc09-S301. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Tabák AG, Jokela M, Akbaraly TN, Brunner EJ, Kivimäki M, Witte DR. Trajectories of glycaemia, insulin sensitivity, and insulin secretion before diagnosis of type 2 diabetes: an analysis from the Whitehall II study. The Lancet. 2009;373(9682):2215–2221. doi: 10.1016/S0140-6736(09)60619-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Mahat RK, Singh N, Arora M, Rathore V. Health risks and interventions in prediabetes: a review. Diabetes Metab Syndr Clin Res Rev. 2019;13(4):2803–2811. doi: 10.1016/j.dsx.2019.07.041. [DOI] [PubMed] [Google Scholar]
- 25.Zand A. Prediabetes: why should we care? Methodist Debakey Cardiovasc J. 2018;14(4):289–297. doi: 10.14797/mdcj-14-4-289. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Song X, Qiu M, Zhang X, et al. Gender-related affecting factors of prediabetes on its 10-year outcome. BMJ Open Diabetes Res Care. 2016;4(1):e000169. doi: 10.1136/bmjdrc-2015-000169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Shang Y, Marseglia A, Fratiglioni L, et al. Natural history of prediabetes in older adults from a population-based longitudinal study. J Intern Med. 2019;286(3):326–340. doi: 10.1111/joim.12920. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Davidson MB. Metformin should not be used to treat prediabetes. Diabetes Care. 2020;43(9):1983–1987. doi: 10.2337/dc19-2221. [DOI] [PubMed] [Google Scholar]
- 29.Ma RCW, Lin X, Jia W. Causes of type 2 diabetes in China. Lancet Diabetes Endocrinol. 2014;2(12):980–991. doi: 10.1016/S2213-8587(14)70145-7. [DOI] [PubMed] [Google Scholar]
- 30.Chan JCN, Zhang Y, Ning G. Diabetes in China: a societal solution for a personal challenge. Lancet Diabetes Endocrinol. 2014;2(12):969–979. doi: 10.1016/S2213-8587(14)70144-5. [DOI] [PubMed] [Google Scholar]
- 31.Ma RCW, Chan JCN. Type 2 diabetes in East Asians: similarities and differences with populations in Europe and the United States. Ann N Y Acad Sci. 2013;1281(1):64–91. doi: 10.1111/nyas.12098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Wenying Y, Juming L, Jianping W, et al. Prevalence of diabetes among men and women in China. N Engl J Med. 2010:12. [DOI] [PubMed]
- 33.Ma RCW. Epidemiology of diabetes and diabetic complications in China. Diabetologia. 2018;61(6):1249–1260. doi: 10.1007/s00125-018-4557-7. [DOI] [PubMed] [Google Scholar]
- 34.Unger RH. Lipotoxicity in the pathogenesis of obesity-dependent NIDDM. 1995;44:8. [DOI] [PubMed]
- 35.Cai X, Qiu S, Liu S, et al. Body-weight fluctuation and risk of diabetes in older adults: the China Health and Retirement Longitudinal Study (CHARLS) Diabetes Res Clin Pract. 2020;169:108419. doi: 10.1016/j.diabres.2020.108419. [DOI] [PubMed] [Google Scholar]
- 36.Zhang L, Yang J, Liao Z, et al. Association between diabetes and cognitive function among people over 45 years old in China: a cross-sectional study. Int J Environ Res Public Health. 2019;16(7):1294. doi: 10.3390/ijerph16071294. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Sun Y, Zhang L, Duan W, Meng X, Jia C. Association between famine exposure in early life and type 2 diabetes mellitus and hyperglycemia in adulthood: results from the China Health And Retirement Longitudinal Study (CHARLS) J Diabetes. 2018;10(9):724–733. doi: 10.1111/1753-0407.12653. [DOI] [PubMed] [Google Scholar]
- 38.Martin LM, Leff M, Calonge N, Garrett C, Nelson DE. Validation of self-reported chronic conditions and health services in a managed care population. Am J Prev Med. 2000;18(3):215–218. doi: 10.1016/s0749-3797(99)00158-0. [DOI] [PubMed] [Google Scholar]
- 39.Yuan X, Liu T, Wu L, Zou ZY, Li C. Validity of self-reported diabetes among middle-aged and older Chinese adults: the China Health and Retirement Longitudinal Study. BMJ Open. 2015;5(4):e006633–e006633. doi: 10.1136/bmjopen-2014-006633. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Silverstein M, Gong CH, Kendig H. Perceived availability of future care and depressive symptoms among older adults in China: evidence from CHARLS. BMC Geriatr. 2020;20(1):31. doi: 10.1186/s12877-020-1435. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(DOCX 229 kb)