Abstract
Background
The natural history of prediabetes in older adults remains unknown.
Objectives
To assess the rate at which prediabetes progresses to diabetes, leads to death or reverts to normoglycaemia in older adults and to identify prognostic factors related to different outcomes of prediabetes.
Methods
In the Swedish National Study on Aging and Care‐Kungsholmen, 2575 diabetes‐free participants aged ≥60 years were examined at baseline and followed for up to 12 years. At each wave, diabetes was diagnosed via medical examination, antidiabetic drug use, medical records or glycated haemoglobin (HbA1c) ≥6.5%. Prediabetes was defined as HbA1c ≥5.7% and normoglycaemia as HbA1c <5.7% in diabetes‐free participants. Data were analysed with multinomial logistic regression.
Results
At baseline, 918 (36%) individuals had prediabetes. Of them, 204 (22%) reverted to normoglycaemia (3.4/100 person‐years, 95% CI 5.6–12.3), 119 (13%) developed diabetes (2.0/100 person‐years, 95% CI 1.7–2.4) and 215 (23%) died (13.0/100 person‐years, 95% CI 11.4–14.9) during the 12‐year follow‐up. The rates of reversion, progression and mortality were higher in the first 6‐year than in the second 6‐year follow‐up, albeit not statistically significant. Lower systolic blood pressure (SBP), absence of heart diseases and weight loss promoted the reversion from prediabetes to normoglycaemia, whilst obesity accelerated its progression to diabetes.
Conclusions
During a 12‐year follow‐up, most of older adults with prediabetes remained stable or reverted to normoglycaemia, whereas only one‐third developed diabetes or died. Lower SBP, no heart diseases and weight management may promote reversion to normoglycaemia, suggesting possible strategies for achieving normoglycaemia in older adults with prediabetes.
Keywords: older adults, prediabetes, prognostic factors, progression, reversion
Introduction
In 2017, 352 (7.3%) million adults were living with prediabetes worldwide, and this number is expected to increase to 587 million (8.3%) by 2045 1. Prediabetes is an asymptomatic condition preceding type 2 diabetes (hereafter, diabetes). It is characterized by hyperglycaemia, which is defined as a blood glucose level that is higher than normal but below the level for a clinical diagnosis of diabetes 2. Prediabetes is more common in older than younger people; about 48% of U.S. adults aged ≥65 years had this condition in 2010 3, 4.
Prediabetes is a high‐risk state for diabetes; about 5–10% of prediabetes may convert to diabetes annually. According to American Diabetes Association expert panel, 70% of individuals with prediabetes may eventually develop diabetes 5. On the other hand, prediabetes may also convert back to normoglycaemia. Several studies have shown that about 3% of adults aged 25–52 years reverted to normoglycaemia annually 6. In addition, prediabetes is linked to high mortality rate in older adults 7, 8. Although the progression of prediabetes is often described in working‐age adults, the evidence on the progression, reversion and mortality of prediabetes in older adults is limited. Given the high prevalence and heterogeneous outcomes of prediabetes in old age, it is pivotal to investigate the natural course of prediabetes amongst older adults.
Glycaemic control can be improved by lifestyle modification in working‐age adults 9, 10. Clinical trials have demonstrated reduction in the risk of developing diabetes amongst people with prediabetes followed by lifestyle interventions (such as weight loss) 11. However, other trials showed that lifestyle interventions are associated with a reduced risk of progression to diabetes only amongst adults aged 45–60 with prediabetes, and suggested that benefit from low glycaemic level may be confined to younger patients 12. Indeed, glycated haemoglobin A1c (HbA1c) values increase with age amongst diabetes‐free subjects, and low glucose level may increase mortality in old age 13, 14. Questions remain on which factors are related to the reversion from prediabetes to normoglycaemia, independently of mortality amongst the older population.
In the present study, we aimed to estimate the rate at which prediabetes reverts to normoglycaemia, progresses to diabetes or leads to death in older adults and to identify prognostic factors related to the reversion of prediabetes using 12‐year follow‐up data from a population‐based longitudinal study of Swedish older adults.
Materials and methods
Study population
Data were derived from the Swedish National Study on Aging and Care in Kungsholmen (SNAC‐K), an ongoing population‐based longitudinal study on ageing 15. A random sample of all registered inhabitants aged ≥60 years living at home or in nursing homes in Kungsholmen in central Stockholm were invited to the baseline assessment (2001–2004). Because health conditions change more rapidly and attrition rates increase as people grow older, sampling was stratified by 11 age cohorts. The younger age cohorts (60, 66 and 72 years) were followed up every 6 years, and the older age cohorts (78, 81, 84, 87, 90, 93, 96, and ≥99 years) were followed up every 3 years. Of the 4590 people who were alive and eligible, 3363 (73.3%) agreed to participate. In the current study, we excluded people with baseline type 1 (n = 21) or type 2 (n = 292) diabetes, resulting in a total of 3050 diabetes‐free participants. Further, 475 participants refused follow‐up examinations or were no longer contactable, leading to the analytical sample of 2575 diabetes‐free participants who were followed for up to 12 years (Fig. 1). Those who dropped out were older and more likely to be female, to have less than an elementary school education, physically inactive, consume less alcohol and have lower Mini‐Mental State Examination (MMSE) score than those included in the study (P < 0.01 for all; Table A1 in Appendix).
Figure 1.
Flow chart of study population.
Written informed consent was obtained from all participants or from a proxy of those with cognitive impairment. SNAC‐K was approved by the Ethics Committee at Karolinska Institutet and by the Regional Ethical Review Board in Stockholm, Sweden.
Data collection
Data on demographic and lifestyle factors, current medication use, and medical history were collected through structured interviews and clinical examinations carried out by trained nurses and physicians (protocol available from http://www.snac-k.se). Peripheral blood samples were taken for laboratory testing.
Education was measured as the number of years of formal schooling and dichotomized into elementary school versus above elementary school. Smoking was dichotomized into current smoking versus never or former smoked. Alcohol consumption was categorized as ‘no or occasional’, ‘light‐to‐moderate’ or ‘heavy’ based on the frequency and amount of drinks 16. Physically active was dichotomized as being physically inactive (never engaged, engaged ≤2–3 times per month in light‐ or/and moderate‐to‐intense exercise) versus active (performing weekly moderate‐to‐vigorous exercise) 17.
Weight and height were measured without shoes and heavy clothes at each wave. Body mass index (BMI) was calculated as body weight divided by the square of height and categorized as underweight (<20), normal weight (20–24.9), overweight (25–29.9) and obesity (≥30 kg m−2). Weight change was calculated as the weight at the visit when glycaemic status (i.e. prediabetes, normoglycaemia) was identified minus baseline weight. To avoid possible reverse causality, for those who developed diabetes, weight change was calculated as the weight at the follow‐up prior to diabetes occurrence minus baseline weight. We divided weight change into tertiles based on its distribution and further interpreted the tertiles as ‘loss’ (−33 to −5), ‘stable’ (−4 to 0) and ‘gain’ (+1 to +17 kg).
Arterial blood pressure was measured on the right arm using sphygmomanometer when participants were seated. It was measured twice with a 5‐min interval, and the average of the two readings was used to determine systolic and diastolic blood pressure (SBP and DBP [mm Hg]). High total cholesterol was defined as nonfasting total cholesterol of ≥6.22 mmol L−1 or use of cholesterol‐lowering agents (Anatomical Therapeutic Chemical [ATC] code C10). MMSE was used to evaluate global cognitive function 18.
Data on chronic medical conditions included self‐reports and medication use, SNAC‐K clinical examination results, and information from the Swedish National Inpatient/Outpatient Registers. International Classification of Disease (ICD 10) was used to identify medical conditions 19. Cerebrovascular diseases and heart diseases (atrial fibrillation, cardiac valve disease, heart failure, bradycardias, conduction diseases, ischaemic heart disease) were identified. Data on the total number of medications used were obtained through visual inspection.
For participants who died during follow‐up, information on the cause of death was extracted from the Swedish Cause of Death Registry (Jun 2001–December 2016). It was also used to obtain information on the vital status of participants who dropped out.
Assessment of prediabetes, diabetes and normoglycaemia
HbA1c was collected at each wave. Until December 2010, HbA1c was assessed with Swedish Mono‐S filament High Performance Liquid Chromatography, and 1.1% was added to the individual's values to render them equal to international values in accordance with National Glycohemoglobin Standardization Program (NGSP; HbA1c in %) 20. Since 1 January 2011, HbA1c has been assessed with the International Federation of Clinical Chemistry (IFCC) reference method. A standard equation (NGSP = [0.9148 * IFCC] + 2.152; available at: http://www.ngsp.org/ifccngsp.asp) was applied to convert IFCC HbA1c (in mmol mol−1) to NGSP value (in %), to render HbA1c results from all waves comparable 21.
At each wave, diabetes was identified by combining information from different sources: medical examination, antidiabetic drug use, diagnoses from the National Patient Register (ICD‐9: code 250; ICD‐10: code E11), from the Swedish Cause of Death Registry (any cause of death = diabetes) or HbA1c ≥6.5% (48 mmol mol−1) 2. In diabetes‐free participants, prediabetes was defined as HbA1c of ≥5.7% to 6.4% (39–46 mmol mol−1), and normoglycaemia was defined as HbA1c <5.7% (39 mmol mol−1) 2.
Statistical analysis
Baseline characteristics in participants with normoglycaemia and prediabetes were compared with the chi‐square test for categorical variables and the t‐test or Mann–Whitney test for continuous variables. Evolution rates from prediabetes to each outcome were calculated as the number of events divided by the follow‐up time (sum of person‐years at risk). For participants who developed diabetes, follow‐up time was calculated as full time during which participants were diabetes‐free plus half of the time during which diabetes developed. If no diabetes occurred, follow‐up time was calculated as the time from baseline until last examination. Similar methods were applied to calculate mortality rate and reversion rate in prediabetes, except that the follow‐up time was the full interval between baseline examination and date of death, or the follow‐up examinations when normoglycaemia was first identified. All estimates from baseline to the 12‐year follow‐up and every 6 years were reported. Furthermore, the corresponding evolution rates of incident prediabetes identified during first 6‐year follow‐up were calculated.
Multinomial logistic regression was used to calculate the odds ratios and 95% confidence intervals of the factors associated with reversion to normoglycaemia, progression to diabetes or death in prediabetes, with stable prediabetes during 12‐year follow‐up as the reference group. The selection procedure followed a forward stepwise approach. All analyses were first adjusted for age, sex, education and follow‐up time (model 1). Alcohol consumption, physical activity, baseline BMI, SBP, MMSE, heart diseases and total number of medications used were included as covariates. We additionally included weight change in the analysis (model 2). Potential interactions between prognostic factors were investigated. Stratified analysis was performed if a significant interaction was present.
Multiple imputation was performed to handle missing data at baseline (<8.5%). P < 0.05 was considered statistically significant, and all analyses were performed using stata 15.0 (StataCorp, College Station, TX, USA).
Results
Baseline characteristics of the study population
At baseline, the mean age of the 2575 participants was 74.4 ± 11.3 years. A total of 1684 (65.4%) were women, 1656 (64.3%) had normoglycaemia, and 918 (35.6%) had prediabetes. Baseline characteristics of participants according to glycaemic status are summarized in Table 1. People with prediabetes were older than those with normoglycaemia. They were also more likely to consume less alcohol, to be overweight or obese, to have higher SBP and MMSE score, to have heart diseases and to use more medications.
Table 1.
Baseline characteristics of the SNAC‐K participants by glycaemic status (n = 2575)
| Characteristics | Normoglycaemia | Prediabetes | P |
|---|---|---|---|
| n = 1656 | n = 918 | ||
| Age group | |||
| 60/66 | 734 (44.3) | 305 (33.2) | <0.001 |
| 72/78 | 438 (26.4) | 276 (30.1) | |
| 81–87 | 256 (15.6) | 198 (21.6) | |
| 90+ | 229 (13.8) | 139 (15.1) | |
| Female sex | 1081 (65.2) | 603 (65.7) | 0.819 |
| Education above elementary school | 1381 (84.2) | 756 (82.4) | 0.225 |
| Current smokera | 208 (13.0) | 146 (16.0) | 0.073 |
| Alcohol consumption | |||
| No/occasional | 507 (31.7) | 357 (39.3) | 0.001 |
| Light‐to‐moderate | 818 (51.1) | 409 (45.0) | |
| Heavy | 277 (17.3) | 142 (15.6) | |
| Physically active, n (%) | 1131 (68.3) | 613 (66.8) | 0.442 |
| Body mass index | |||
| <20 | 110 (7.1) | 61 (6.9) | 0.002 |
| 20–24.9 | 692 (44.8) | 346 (38.9) | |
| 25–29.9 | 594 (38.5) | 357 (40.1) | |
| ≥30 | 149 (9.6) | 126 (14.2) | |
| SBPa, mm Hg | 141 (±20.7) | 143 (±20.2) | 0.018 |
| DBPa, mm Hg | 81 (±11.3) | 80 (±10.9) | 0.194 |
| MMSE score | 27 (±6.5) | 28 (±3.2) | <0.001 |
| High total cholesterol | 718 (48.4) | 475 (52.0) | 0.080 |
| Cerebrovascular diseases | 117 (7.1) | 73 (7.9) | 0.407 |
| Heart diseases | 331 (20.0) | 248 (27.0) | <0.001 |
| Total no. of medication use | 3.7 (±3.2) | 4.0 (±3.4) | 0.003 |
DBP, diastolic blood pressure; MMSE, Mini‐Mental State Examination; SBP, systolic blood pressure.
Data are presented as mean ± standard deviation, number (%).
Missing data: 17 were missing data on education, 69 on smoking, 65 on alcohol consumption, 218 on body mass index and 177 on total cholesterol.
Natural history of prediabetes
The total follow‐up time for reversal to normoglycaemia was 5675 person‐years [median: 8.7 years, interquartile range (IQR): 6.0 years], for incident diabetes, 5339 person‐years (median: 8.6 years, IQR: 8.4 years) and for death, 1627 person‐years (median: 4.3 years, IQR: 4.54 years). The transition of glycaemic status and death over 12 years is shown in Fig. 2.
Figure 2.
Transition of glycaemic status and death over the examination waves.
Of the 918 people with prediabetes, 380 (41.3%) had prediabetes at the 12‐year follow‐up, 204 [22.2%; 3.4/100 person‐years (95% CI: 3.1–4.1)] reverted to normoglycaemia, 119 [11.0%; 2.3/100 person‐years (95% CI: 1.9–2.7)] progressed to diabetes and 215 [23.4%; 13.0/100 person‐years (95% CI: 11.4–14.9)] died. The results remain similar after adjusting for sex [reversion rate = 3.6/100 person‐years (95% CI: 3.1–4.1); progression rate = 2.2/100 person‐years (95% CI: 1.8–2.6); mortality rate = 12.9/100 person‐years (95% CI: 11.2–14.7)]. We observed an age‐related gradient in the reversion (P for trend <0.001) and progression rates (P for trend = 0.63). Highest mortality rate was observed in those ≥90 years (Table 2).
Table 2.
Reversion, progression and mortality rates (100 person‐years, [95% confidence interval]) by age group during 12 years amongst people with prediabetes
| Reversion from prediabetes to normoglycaemia | |||
|---|---|---|---|
| Age group | Prediabetes (n) | Normoglycaemia (n)/person‐years | Reversion rate (95% CI) |
| 60/66 | 305 | 74/2597 | 2.8 (2.3–3.6) |
| 72/78 | 276 | 62/1860 | 3.3 (2.5–4.3) |
| 81–87 | 198 | 43/938 | 4.6 (3.4–6.2) |
| 90+ | 139 | 25/279 | 8.9 (6.1–13.3) |
| Total | 918 | 204/5674 | 3.4 (3.1–4.1) |
| Progression from prediabetes to diabetes | |||
|---|---|---|---|
| Age group | Prediabetes (n) | Diabetes (n)/ person‐years | Progression rate (95% CI) |
| 60/66 | 305 | 49/2457 | 2.0 (1.5–2.6) |
| 72/78 | 276 | 38/1718 | 2.2 (1.6–3.0) |
| 81–87 | 198 | 25/900 | 2.7 (1.9–4.1) |
| 90+ | 139 | 7/263 | 2.7 (1.3–5.6) |
| Total | 918 | 119/5339 | 2.3 (1.9–2.7) |
| Death | |||
|---|---|---|---|
| Age group | Prediabetes (n) | Death (n)/person‐years | Mortality rate (95% CI) |
| 60/66 | 305 | 35/254 | 13.7 (9.8–19.2) |
| 72/78 | 276 | 58/474 | 12.0 (0.3–15.6) |
| 81–87 | 198 | 46/503 | 8.9 (6.7–11.9) |
| 90+ | 139 | 76/394 | 18.9 (15.1–23.8) |
| Total | 918 | 215/1627 | 13.0 (11.4–14.9) |
Table 3 presents factors related to the natural history of prediabetes over the 12 years of follow‐up. After controlling for individual follow‐up time, demographic factors, lifestyle factors, baseline BMI, SBP, MMSE, heart diseases and total number of medications (model 2), weight loss significantly increased the odds of reverting to normoglycaemia [odd ratios (OR) = 2.0, 95% CI: 1.1–3.2]. The association appears to be more evident in participants who were overweight (P for interaction: 0.03) or obese (P for interaction: 0.06) at baseline than in those who were not. In stratified analysis, we combined overweight and obese into one group due to the same direction of association and the small number of people with obesity. The odds ratios were 2.9 (95% CI 1.4–6.2) for weight loss on reversal to normoglycaemia in those who were overweight or obese at baseline. Apart from this, reversion to normoglycaemia was also inversely associated with higher SBP (OR = 0.9, 95% CI: 0.8–0.9) and heart diseases (OR = 0.5, 95% CI: 0.3–0.9; model 2).
Table 3.
Prognostic factors associated with reversion, progression and death in prediabetes over 12 years of follow‐up. Odds ratios and 95% confidence intervals (OR; 95% CI) from multinomial logistic regression with the 380 participants with stable prediabetes as reference group
| Characteristics | Reverted to normoglycaemia (n = 204) | Progressed to diabetes (n = 119) | Death (n = 215) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| n | Model 1 OR (95% CI) | Model 2 OR (95% CI) | n | Model 1 OR (95% CI) | Model 2 OR (95% CI) | n | Model 1 OR (95% CI) | Model 2 OR (95% CI) | |
| Body mass index | |||||||||
| <25 | 89 | Ref. | Ref. | 38 | Ref. | Ref. | 104 | Ref. | Ref. |
| 25–30 | 85 | 1.0 (0.7–1.5) | 1.3 (0.8–2.2) | 53 | 1.6 (0.9–2.8) | 1.5 (0.8–2.9) | 59 | 0.8 (0.5–1.3) | 1.1 (0.5–2.2) |
| ≥30 | 19 | 0.7 (0.4–1.2) | 0.7 (0.3–1.5) | 27 | 2.5 (1.2–5.1) | 2.8 (1.3–6.6) | 22 | 0.9 (0.5–1.8) | 0.9 (0.3–2.0) |
| Weight change over 12 years, kg | |||||||||
| Loss (−33 to −5) | 80 | 1.6 (1.0–2.6)a | 2.0 (1.1–3.2) | 24 | 0.8 (0.4–1.7) | 0.7 (0.4–1.5) | 23 | 1.2 (0.6–2.5) | 1.1 (0.5–2.5) |
| Stable (−4 to 0) | 59 | Ref. | Ref. | 45 | Ref. | Ref. | 19 | Ref. | Ref. |
| Gain (1 to 17) | 51 | 1.2 (0.7–2.1) | 1.1 (0.6–2.0) | 48 | 1.7 (1.0–3.3)a | 1.7 (0.8–3.4) | 10 | 0.8 (0.3–2.0) | 0.8 (0.3–2.0) |
| Physically active | 147 | 1.0 (0.6–1.6) | 0.9 (0.5–1.8) | 83 | 1.1 (0.6–2.0) | 0.9 (0.5–1.8) | 99 | 0.5 (0.3–0.7) | 0.6 (0.3–1.2) |
| Alcohol consumption | |||||||||
| No/occasional | 66 | 0.7 (0.5–1.1) | 0.5 (0.3–1.1) | 39 | 0.8 (0.5–1.4) | 0.6 (0.3–1.1) | 107 | 0.9 (0.6–1.7) | 1.1 (0.7–1.8) |
| Light‐to‐moderate | 97 | Ref. | Ref. | 56 | Ref. | Ref. | 79 | Ref. | Ref. |
| Heavy | 40 | 1.3 (0.8–2.1) | 1.1 (0.6–2.0) | 23 | 1.3 (0.6–2.6) | 1.2 (0.5–3.5) | 24 | 1.4 (0.7–2.6) | 1.5 (0.7–2.8) |
| SBP | 203 | 0.9 (0.9–1.0) | 0.9 (0.8–0.9) | 117 | 1.1 (1.1–1.2) | 1.0 (0.9–1.0) | 213 | 1.0 (0.9–1.0) | 0.9 (0.8–1.0) |
| Heart diseases | 31 | 0.6 (0.4–0.9) | 0.5 (0.3–0.9) | 38 | 1.3 (0.8–2.4) | 1.2 (0.6–2.5) | 99 | 1.6 (0.9–2.3) | 1.2 (0.7–1.9) |
BMI, body mass index; OR, odds ratio; SBP, systolic blood pressure.
Model 1 adjusted for age, sex, education and follow‐up time. Model 2 adjusted for variables in Model 1 and alcohol consumption, physical activity, body mass index, weight change, SBP, heart diseases, Mini‐Mental State Examination score and total medication use.
P = 0.049.
Participants with baseline obesity had a higher probability of progressing to diabetes than those who were underweight or of normal weight at baseline (OR = 2.8, 95% CI: 1.3–6.6). After controlling for demographic factors and follow‐up time (model 1), higher SBP increased the odds of progressing to diabetes, but the association was not statistically significant after full adjustment (model 2).
Participants who were physically active were associated with reduced mortality (OR = 0.6, 95% CI: 0.3–0.8) adjusting for demographic factors, alcohol consumption, MMSE score, baseline BMI, SBP, heart diseases and total number of medication use (table not shown), but was not statistically significant after taking into account weight change (model 2).
Supplementary analysis
Because of the higher attrition rate in those ≥78 years and the frequent transitions between normoglycaemia and prediabetes, we also estimated occurrence of each outcome during the two follow‐up periods: from baseline (2001–2004) to the 6‐year follow‐up (2007–2010) and from the 6‐year follow‐up (2007–2010) to the 12‐year follow‐up (2013–2016). In general, the evolution rates of prediabetes slowed down: the reversion rate to normoglycaemia decreased from 16.0% (reversion rate = 3.7/100 person‐years, 95% CI: 3.1–4.3) to 13.1% (3.3/100 person‐years, 95% CI 2.5–4.3), the progression rate to diabetes decreased from 10.1% (progression rate = 2.5/100 person‐years, 95% CI: 2.0–3.0) to 6.0% (1.6/100 person‐years, 95% CI: 1.0–2.3), and the mortality rate declined from 17.6% (mortality rate = 14.2/100 person‐years, 95% CI: 12.2–16.6) to 12.2% (8.8/100 person‐years, 95% CI: 6.7–27.3; Table A2 in Appendix). During the first 6‐year follow‐up, 363 incident prediabetes was identified, and corresponding evolution showed generally similar patterns to those from the initial analysis (Table A3 in Appendix).
Discussion
In this population‐based cohort study that followed Swedish older adults aged ≥60 for 12 years, the overall reversion rate from prediabetes to normoglycaemia was 3.4/100 person‐years (about 22%), progression rate to diabetes was 2.0/100 person‐years (13%), and mortality rate was 13.0/100 person‐years (23%). Lower SBP, no heart diseases and weight loss were associated with reversion to normoglycaemia, and obesity anticipated the progression from prediabetes to diabetes. Physical activity also reduced mortality rate related to prediabetes.
Previous studies focused either on the reversion from prediabetes to normoglycaemia, the progression from prediabetes to diabetes, or the mortality rate related to prediabetes; however, these three outcomes have not been assessed simultaneously in older population. A study on the evolution of prediabetes in middle‐aged adults showed a reversion rate of normoglycaemia of around 23% and a progression rate to diabetes of about 30% 22. In a Swedish middle‐aged population, the reversion rate was 36% during 8–10 years 23. We observed an approximately 22% reversion rate in older adults over 12 years. The only other longitudinal population‐based study of older adults to examine the reversion rate to normoglycaemia, the KORA S4/F4 study of 55‐ to 74‐year‐olds in Germany, found a reversion rate of 16.3% over 7 years of follow‐up, using oral glucose tolerance test (OGTT) as diagnostic criterion 24. We reported a slightly higher reversion rate of 16% during the first 6‐year follow‐up, which may be explained by a larger proportion of prediabetes diagnosed by HbA1c, as the sensitivity of HbA1c test in diagnosing prediabetes is inferior to the OGTT 25.
In our study, the incidence of diabetes (2.0/100 person‐years) was slightly higher than that found in studies of people of the same age range carried out in other regions of Europe, where it varied from 0.7 to 1.2/100 person‐years in people with normoglycaemia or prediabetes 26. One study in a Dutch population aged 55 to 75 years reported much higher rates of progression from prediabetes to diabetes; specifically, 5.2/100 person‐years (33%) from impaired fasting glucose (IFG) to diabetes and 5.8/100 person‐years (34%) from impaired glucose tolerance (IGT) to diabetes 27. The higher estimates from the Dutch study are probably due to different methods used to assess prediabetes and the younger population with a higher prevalence of prediabetes (64.5% in the Dutch study and 35.3% in SNAC‐K).
The mortality rate of prediabetes in our study was slightly higher than in a few previous studies that used HbA1c as prediabetes assessment. In participants with mean age of 52 years without diabetes treatment in the NIPPON DATA90, the mortality rate was 23.7% during 10 years of observation 8. A Danish study reported a 10‐year mortality rate of around 15.7% in those with mean age of 56 years with prediabetes 28. The higher rates observed in our study could be explained by older age, higher level of average HbA1c, blood pressure, more preexisting diseases and longer follow‐up compared to other studies.
To our knowledge, only one observational study in older adults has investigated the association between weight change and normalization of glycaemia 24. This study showed that weight loss, but not initial BMI, strongly increased the likelihood of reverting from prediabetes to normoglycaemia. Previous studies also show that lower SBP could decrease insulin resistance 29, which is crucial in the aetiology and progression of prediabetes. Our results were consistent with these findings. Additionally, in the stratified analysis, the association between weight loss and reversion to normoglycaemia was present only amongst individuals who were overweight or obese at baseline, suggesting that lifestyle modifications, promoting weight loss, can help restore the normoglycaemia in adults with adiposity 30. Currently, the guidelines for prevention of diabetes emphasize the management of blood pressure, through lowering SBP, as a potential strategy to improve the glycaemic control in diabetes 31. Based on this, we believe that even at the prediabetic stage, lowering the SBP levels might improve glycaemia, thus promoting normoglycaemia restoration. However, reduction of blood pressure should not be excessively strict in very old people in order to avoid, for example, possible falls due to hypotension. Furthermore, preexisting heart diseases might hamper the reversion to normoglycaemia, as impaired vascular endothelial function and oxidative stress generated by atherosclerosis may further worsen insulin resistance 32.
Apart from overweight and obesity, weight gain may also increase the risk of progression from prediabetes to diabetes through increased insulin resistance 12, 33. Weight gain, particularly accumulation of visceral fat, could increase inflammation around liver and impair insulin signalling, which in turn worsens insulin sensitivity 34. Indeed, we found this association was independent of baseline BMI, albeit statistically insignificant.
Physical activity was found to be associated with reduced mortality, which was consistent with previous studies 35. However, the significance was diminished after additional adjustment for weight change, indicating that weight change might mediate the physical activity–mortality association. We further performed analysis in people without heart diseases to rule out potential reverse causality and the results remained similar.
Although the factors that promoted normoglycaemia differed to some extent from those that predicted progression to diabetes, the impact of weight change appears to be consistent. This result suggests that weight management may be an effective strategy for preventing prediabetes and its progression to diabetes, possibly by improving insulin sensitivity. For example, weight loss, which can be achieved through regular physical activity, could be recommended to older adults with overweight or obesity, as possible effective strategy to restore normoglycaemia. Another reason to emphasize weight management in people with prediabetes is that more cardiovascular benefits can be obtained if an effective weight management is achieved during the prediabetic stage 36.
The main strengths of our study included the population‐based longitudinal study design, identifying diabetes from multiple sources, the repeated measurements of health‐related conditions, repeated blood sampling and the long follow‐up time. However, some limitations need to be acknowledged. First, we could not differentiate prediabetes phenotypes. IFG, IGT and the combination of IFG and IGT represent multiple pathophysiological abnormalities, which are likely to differ in their rates of evolution and clinical relevance 5. Moreover, the recommended HbA1c cut‐offs to identify diabetes (≥6.5%) might have lower sensitivity compared to OGTT 25. Therefore, the number of people with diabetes and the magnitude of observed associations may have been underestimated. Nevertheless, ascertaining prediabetes and diabetes with the HbA1c test has substantial advantages particularly in older populations: it does not require fasting status and can be measured regardless of the time of day, and it reflects hyperglycaemia over the past three months. As HbA1c was likely to be influenced by the presence of anaemia, we further conducted sensitivity analysis to exclude those with anaemia and the results remain similar. Secondly, selection bias may be present, as those who completed the follow‐ups were generally younger and more educated, more physically active and have higher MMSE score at baseline than those lost to follow‐up. Hence, this may have led to overestimation of the reversion rate to normoglycaemia and an underestimation of incident diabetes. However, this should have relatively small impact of the results, as the number of those dropped out was relatively small and the probability of participation was independently of HbA1c level. The limitations of a potential selection bias and small sample size may still account for the lack of association between some lifestyle factors and evolution of prediabetes. For the missing information of covariates at baseline, analyses using multiple imputation yields estimates with similar magnitude and direction for most prognostic factors. Finally, since there is no conventional definition of weight loss, stable or gain in older adults, we conducted the analysis based on data distribution. As the range of values in weight change is wide and our results might be potentially driven by the extreme values, we excluded outliers and the association between weight loss and normoglycaemia remained similar. The major findings of this study can be generalized to populations with characteristics similar to those of our study population.
In conclusion, our findings provide evidence on the natural history of prediabetes: during a 12‐year follow‐up, 42% older adults affected by prediabetes remain stable, 22% reverted to normoglycaemia, whereas 13% progressed to diabetes or 23% died. We found that baseline BMI, weight changes, SBP and preexisting heart diseases could influence the natural history of prediabetes. The latter finding may help to identify people at high risk of progressing to diabetes and suggests possible strategies for achieving normoglycaemia in older adults with prediabetes.
Conflict of interest statement
No potential conflicts of interest relevant to this article were reported.
Funding
The SNAC‐K (http://www.snac.org) is financially supported by the Swedish Ministry of Health and Social Affairs; the participating county councils and municipalities; the Swedish Research Council; and the Swedish Research Council for Health, Working Life and Welfare. In addition, W.X. also received grants from the Swedish Research Council (No 2017‐00981), the National Natural Science Foundation of China (No. 81771519), the Konung Gustaf V:s och Drottning Victorias Frimurare Foundation (No. 2016–2017) and Eva och Wera Cornell Stiftelsen. Y.S. also received grants from the China Scholarship Council (No. 201600160093) and the Swedish National Graduate School on Aging and Health. Finally, this project is part of the CoSTREAM (http://www.costream.eu/) from the European Union's Horizon 2020 research and innovation programme (No. 667375).
Availability of data and material
The data that support the findings of this study are available from the SNAC‐K but restrictions apply to the availability of these data, which were used under licence for the current study, and so are not publicly available. The data are, however, available from the authors upon reasonable request and with permission of the SNAC‐K study.
Ethics approval and consent to participate
Written informed consent was obtained from all participants or from a proxy of those with cognitive impairment. SNAC‐K was approved by the Ethics Committee at Karolinska Institutet and by the Regional Ethical Review Board in Stockholm, Sweden (Dnrs: KI 01‐114, 04‐929/3, Ö26‐2007, 2009/595‐32, 2010/447‐31/2, 2013/828‐31/3 and 2016/730‐31/1).
Authors’ contributions
Y.S. and W.X. conceptualized and designed the study. Y.S. performed the literature search; analysed the data; interpreted the findings; and wrote, reviewed and edited the manuscript. L.F. designed, initiated and directed SNAC‐K. All authors contributed to the data interpretation and revision of the manuscript and approved the final draft. Y.S. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Consent for publication
Not applicable.
Acknowledgements
We thank all participants and staff involved in data collection and management in Swedish National Study on Aging and Care in Kungsholmen (SNAC‐K). In addition, we are grateful to Kimberly Kane at the Aging Research Center, Karolinska Institutet, for useful comments on the text.
Table A1.
Baseline characteristics of those included in the study sample and those not
| Characteristics | Study sample (n = 2575) | Excluded from study sample (n = 475) | P value |
|---|---|---|---|
| Age group | |||
| 60/66 | 1039 (40.3) | 167 (35.2) | 0.003 |
| 72/78 | 714 (27.7) | 115 (24.2) | |
| 81–87 | 454 (17.6) | 112 (23.6) | |
| 90+ | 368 (14.3) | 81 (17.1) | |
| Female sex | 1684 (65.4) | 341 (71.8) | 0.007 |
| Education above elementary school | 2137 (83.5) | 363 (78.9) | 0.015 |
| Current smoker | 354 (14.1) | 74 (16.4) | 0.198 |
| Alcohol consumption | |||
| No/occasional | 864 (34.4) | 216 (48.0) | <0.001 |
| Light‐to‐moderate | 1227 (48.9) | 172 (38.2) | |
| Heavy | 419 (16.7) | 62 (13.8) | |
| Physically active | 1744 (67.7) | 282 (60.0) | <0.001 |
| Body mass index, kg m−2 | |||
| <20 | 153 (6.5) | 40 (10.1) | 0.065 |
| 20–24.9 | 1005 (42.6) | 159 (40.3) | |
| 25–29.9 | 933 (39.6) | 149 (37.7) | |
| ≥30 | 266 (11.3) | 47 (11.9) | |
| HbA1C, % | 5.5 (±0.3) | 5.5 (±0.4) | 0.376 |
| SBP, mm Hg | 142 (±20.5) | 144 (±20.6) | 0.099 |
| DBP, mm Hg | 81 (±11.1) | 81 (±11.4) | 0.844 |
| High total cholesterol | 1193 (49.8) | 191 (47.5) | 0.406 |
| MMSE score | 27.4 (±5.8) | 26.2 (±7.0) | <0.001 |
| Cerebrovascular diseases | 190 (7.4) | 40 (8.5) | 0.429 |
| Heart diseases | 579 (22.5) | 112 (23.6) | 0.601 |
| Total no. of medications use | 3.8 (±3.2) | 4.0 (±3.2) | 0.074 |
DBP, diastolic blood pressure; HbA1C, glycated haemoglobin A1C; MMSE, Mini‐Mental State Examination; SBP, systolic blood pressure.
Data are presented as mean ± standard deviation, number (%).
Table A2.
Evolution rate (100 person‐years [95% confidence interval]) of baseline prediabetes during two periods of follow‐up
| First 6‐year follow‐up (2001–2004 to 2007–2010) | Second 6‐year follow‐up (2007–2010 to 2013–2016) | |||||
|---|---|---|---|---|---|---|
| Reversion from prediabetes to normoglycaemia | ||||||
| Age group | Prediabetes (n) | Normoglycaemia (n)/person‐years | Reversion rate (95% CI) | Prediabetes (n) | Normoglycaemia (n)/person‐years | Reversion rate (95% CI) |
| 60/66 | 305 | 46/1627 | 2.9 (2.1–3.8) | 197 | 28/899 | 3.1 (2.1–4.5) |
| 72/78 | 276 | 38/1305 | 2.9 (2.1–4.0) | 158 | 24/600 | 3.9 (2.6–5.9) |
| 81–87 | 198 | 39/796 | 4.9 (3.6–6.7) | 66 | 4/179 | 2.2 (0.8–5.9) |
| 90+ | 139 | 24/264 | 9.1 (6.1–13.5) | 13 | 1/23 | 4.2 (0.6–30.3) |
| Total | 918 | 147/3994 | 3.7 (3.1–4.3) | 434 | 57/1702 | 3.3 (2.5–4.3) |
| Progression from prediabetes to diabetes | ||||||
|---|---|---|---|---|---|---|
| Age group | Prediabetes (n) | Diabetes (n)/ person‐years | Progression rate (95% CI) | Prediabetes (n) | Diabetes (n)/ person‐years | Progression rate (95% CI) |
| 60/66 | 305 | 41/1513 | 2.7 (2.0–3.7) | 197 | 8/859 | 0.9 (0.2–1.6) |
| 72/78 | 276 | 24/1249 | 1.9 (1.3–2.9) | 158 | 14/576 | 2.6 (1.2–3.9) |
| 81–87 | 198 | 22/750 | 2.9 (1.9–4.5) | 66 | 3/173 | 2.1 (0.6–5.4) |
| 90+ | 139 | 6/255 | 2.3 (1.1–5.2) | 13 | 1/21 | 4.6 (0.6–32.3) |
| Total | 918 | 93/3769 | 2.5 (2.0–3.0) | 434 | 26/1649 | 1.7 (1.1–2.4) |
| Death | ||||||
|---|---|---|---|---|---|---|
| Age group | Prediabetes (n) | Death (n)/ person‐years | Mortality rate (95% CI) | Prediabetes (n) | Death (n)/ person‐years | Mortality rate (95% CI) |
| 60/66 | 305 | 21/115 | 18.1 (11.8–27.9) | 197 | 14/74 | 18.9 (11.2–32.0) |
| 72/78 | 276 | 36/258 | 13.5 (9.7–18.8) | 158 | 22/310 | 7.1 (4.6–10.7) |
| 81–87 | 198 | 32/377 | 8.2 (5.8–11.7) | 66 | 14/182 | 7.6 (4.5–12.9) |
| 90+ | 139 | 73/365 | 19.7 (15.6–24.8) | 13 | 3/34 | 10.0 (2.8–27.4) |
| Total | 918 | 162/1117 | 14.2 (12.2–16.6) | 434 | 53/600 | 9.0 (6.6–11.5) |
Table A3.
Evolution rate (100 person‐years, [95% confidence interval]) of incident prediabetes at first 6‐year follow‐up (2007–2010)
| Reversion from prediabetes to normoglycaemia | |||
|---|---|---|---|
| Age group | Prediabetes (n) | Normoglycaemia (n)/person‐years | Reversion rate (95% CI) |
| 60/66 | 216 | 79/1015 | 7.8 (6.2–9.7) |
| 72/78 | 95 | 20/410 | 5.3 (2.9–7.6) |
| 81–87 | 41 | 4/100 | 4.5 (1.5–10.7) |
| 90+ | 11 | 1/18 | 5.6 (0.8–39.9) |
| Total | 363 | 104/1543 | 6.7 (5.6–8.2) |
| Progression from prediabetes to diabetes | |||
|---|---|---|---|
| Age group | Prediabetes (n) | Diabetes (n)/ person‐years | Progression rate (95% CI) |
| 60/66 | 216 | 10/986 | 1.0 (0.3–1.5) |
| 72/78 | 95 | 4/402 | 1.4 (0.01–2.7) |
| 81–87 | 41 | 0/100 | – |
| 90+ | 11 | 0/17 | – |
| Total | 363 | 14/1506 | 0.9 (0.4–1.5) |
| Death | |||
|---|---|---|---|
| Age group | Prediabetes (n) | Death (n)/ person‐years | Mortality rate (95% CI) |
| 60/66 | 216 | 10/113 | 8.8 (4.7–16.4) |
| 72/78 | 95 | 9/169 | 5.3 (2.8–10.2) |
| 81–87 | 41 | 8/143 | 5.6 (2.7–11.2) |
| 90+ | 11 | 1/39 | 2.6 (0.4–18.1) |
| Total | 363 | 28/465 | 6.0 (4.2–8.7) |
Table A4.
Baseline characteristics of participants with prediabetes (n = 918) by glycaemic status and death at 12‐year follow‐up
| Characteristics | Reverted to normoglycaemia (n = 204) | Remained as prediabetes (n = 380) | Progressed to diabetes (n = 119) | Death (n = 215) | P value |
|---|---|---|---|---|---|
| Age group | |||||
| 60/66 | 74 (36.3) | 147 (38.7) | 49 (41.2) | 35 (16.3) | <0.001 |
| 72/78 | 63 (30.4) | 118 (31.1) | 38 (31.9) | 58 (27.9) | |
| 81–87 | 43 (21.1) | 84 (22.1) | 25 (21.0) | 46 (21.4) | |
| 90+ | 25 (12.3) | 31 (8.2) | 7 (5.9) | 76 (35.4) | |
| Female sex | 147 (72.1) | 266 (70.0) | 66 (55.5) | 124 (57.7) | <0.001 |
| Education above elementary school | 175 (85.8) | 316 (83.2) | 101 (84.9) | 164 (76.3) | 0.050 |
| Current smoker | 32 (15.8) | 61 (16.1) | 12 (10.2) | 41 (19.3) | 0.192 |
| Alcohol consumption | |||||
| Occasional or no | 66 (32.5) | 145 (38.5) | 39 (33.1) | 107 (50.9) | 0.003 |
| Light‐to‐moderate | 97 (47.8) | 177 (46.9) | 56 (47.5) | 79 (37.6) | |
| Heavy | 40 (19.7) | 55 (14.6) | 23 (19.5) | 24 (11.4) | |
| Physically active | 147 (72.1) | 284 (74.7) | 83 (69.7) | 99 (46.1) | <0.001 |
| Body mass index, kg m−2 | |||||
| <20 | 7 (3.6) | 11 (2.9) | 5 (4.2) | 33 (17.8) | <0.001 |
| 20–24.9 | 82 (42.3) | 151 (40.9) | 33 (27.9) | 71 (38.4) | |
| 25–29.9 | 85 (44.3) | 153 (41.5) | 53 (44.9) | 59 (31.9) | |
| ≥30 | 19 (9.8) | 54 (14.6) | 27 (22.9) | 22 (11.9) | |
| SBP, mm Hg | 142 (±18.1) | 144 (±18.7) | 149 (±20.5) | 140 (±22.7) | <0.001 |
| DBP, mm Hg | 81 (±10.4) | 81 (±9.9) | 83 (±12.1) | 77 (±12.4) | <0.001 |
| HbA1c (%) | 5.8 (±0.1) | 5.9 (± 0.2) | 6.0 (±0.2) | 5.9 (±0.2) | <0.001 |
| High total cholesterol | 113 (55.6) | 209 (55.0) | 65 (55.6) | 88 (41.5) | 0.007 |
| MMSE score | 28 (±2.2) | 27 (± 2.1) | 28 (±1.8) | 27 (±5.0) | <0.001 |
| Cerebrovascular diseases | 9 (4.4) | 25 (6.6) | 10 (8.4) | 29 (13.5) | 0.004 |
| Heart diseases | 31 (15.2) | 80 (21.1) | 38 (31.9) | 99 (46.1) | <0.001 |
| No. of medications use | 3.6 (±3.2) | 3.6 (±3.4) | 3.9 (±2.9) | 5.1 (±3.6) | <0.001 |
DBP, diastolic blood pressure; HbA1c, glycated haemoglobin; MMSE, Mini‐Mental State Examination; SBP, systolic blood pressure.
Data are presented as mean ± standard deviation, number (%).
Pairwise comparison using Bonferroni correction; P‐value < 0.05; reference group: participants who still had prediabetes during follow‐up.
Shang Y, Marseglia A, Fratiglioni L, Welmer A‐K, Wang R, Wang H‐X, Xu W (Karolinska Institutet; Stockholm Gerontology Research Center; Karolinska University Hospital; Stockholm University; Stockholm, Sweden; and Tianjin Medical University, Tianjin, China) Natural history of prediabetes in older adults from a population‐based longitudinal study. J Intern Med 2019;286:326–340.
Contributor Information
Y. Shang, Email: ying.shang@ki.se.
W. Xu, Email: weili.xu@ki.se.
References
- 1. International Diabetes Federation . IDF Diabetes Atlas, 8th edn Brussels, Belgium: International Diabetes Federation, 2017; http://www.diabetesatlas.org. [Google Scholar]
- 2. American Diabetes Association . Classification and diagnosis of diabetes. Sec. 2. In Standards of Medical Care in Diabetes 2017. Diabetes Care 2017; 40(Suppl. 1): S11–24.27979889 [Google Scholar]
- 3. Ferrannini E. Definition of intervention points in prediabetes. Lancet Diabetes Endocrinol 2014; 2: 667–75. [DOI] [PubMed] [Google Scholar]
- 4. Bullard KM, Saydah SH, Imperatore G et al Secular changes in U.S. Prediabetes prevalence defined by hemoglobin A1c and fasting plasma glucose: National Health and Nutrition Examination Surveys, 1999–2010. Diabetes Care 2013; 36: 2286–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Tabák AG, Herder C, Rathmann W, Brunner EJ, Kivimäki M. Prediabetes: a high‐risk state for diabetes development. Lancet 2012; 379: 2279–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. 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: 305–12. [DOI] [PubMed] [Google Scholar]
- 7. Kowall B, Rathmann W, Heier M et al Categories of glucose tolerance and continuous glycemic measures and mortality. Eur J Epidemiol 2011; 26: 637–45. [DOI] [PubMed] [Google Scholar]
- 8. Sakurai M, Saitoh S, Miura K et al HbA1c and the risks for all‐cause and cardiovascular mortality in the general Japanese population: NIPPON DATA90. Diabetes Care 2013; 36: 3759–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Feldman AL, Griffin SJ, Ahern AL et al Impact of weight maintenance and loss on diabetes risk and burden: a population‐based study in 33,184 participants. BMC Public Health 2017; 17: 170. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Mclean DL, Simpson SH, Mcalister FA, Tsuyuki RT. Treatment and blood pressure control in 47,964 people with diabetes and hypertension: A systematic review of observational studies. Can J Cardiol 2006; 22: 855–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Herman WH, Pan Q, Edelstein SL et al Impact of lifestyle and metformin interventions on the risk of progression to diabetes and regression to normal glucose regulation in overweight or obese people with impaired glucose regulation. Diabetes Care 2017; 40: 1668–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Knowler WC, Diabetes Prevention Program Research Group . Reduction in the incidence of type 2 diabetes with intervention or metformin. N Engl J Med 2002; 346: 393–403. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Pani L, Korenda L, Meige J et al Effect of aging on A1C levels in individuals without diabetes evidence from the Framingham Offspring Study and the National Health and Nutrition Examination Survey 2001–2004. Diabetes Care 2008; 31: 1991–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Selvin E, Steffes M, Zhu H, Matsushita K, Wagenknecht L et al Glycated hemoglobin, diabetes, and cardiovascular risk in nondiabetic adults. N Engl J Med 2010; 362: 800–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Lagergren M, Fratiglioni L, Hallberg IR et al A longitudinal study integrating population, care and social services data. The Swedish National study on Aging and Care (SNAC). Aging Clin Exp Res 2004; 16: 158–68. [DOI] [PubMed] [Google Scholar]
- 16. Breslow RA, Chen CM, Graubard BI, Jacobovits T, Kant AK. Diets of drinkers on drinking and nondrinking days: NHANES 2003–2008. Am J Clin Nutr 2013; 97: 1068–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Rydwik E, Welmer AK, Kåreholt I, Angleman S, Fratiglioni L, Wang HX. Adherence to physical exercise recommendations in people over 65—The SNAC‐Kungsholmen study. Eur J Public Health 2013; 23: 799–804. [DOI] [PubMed] [Google Scholar]
- 18. Folstein MF, Folstein SE, Mchugh PR. “Mini‐Mental State” A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: 189–98. [DOI] [PubMed] [Google Scholar]
- 19. Calderón‐Larrañaga A, Vetrano DL, Onder G et al Assessing and measuring chronic multimorbidity in the older population: A proposal for its operationalization. J Gerontol A Biol Sci Med Sci 2017; 72: 1417–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Goodall I. HbA1c standardisation destination‐global IFCC Standardisation. How, why, where an when‐a tortuous pathway from kit manufacturers, via inter‐laboratory lyophilized and whole blood comparisons to designated national comparison schemes. Clin Biochem Rev 2005; 26: 5–19. [PMC free article] [PubMed] [Google Scholar]
- 21. Hoelzel W, Weykamp C, Jeppsson J‐O et al IFCC reference system for measurement of hemoglobin A1c in human blood and the national standardization schemes in the United States, Japan, and Sweden: a method‐comparison study. Clin Chem 2004; 50: 166–74. [DOI] [PubMed] [Google Scholar]
- 22. Song X, Qiu M, Zhang X et al Gender‐related affecting factors of prediabetes on its 10‐year outcome. BMJ Open Diabetes Research and Care 2016; 4: e000169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Alvarsson M, Hilding A, Östenson CG. Factors determining normalization of glucose intolerance in middle‐aged Swedish men and women: A 8–10‐year follow‐up. Diabet Med 2009; 26: 345–53. [DOI] [PubMed] [Google Scholar]
- 24. Kowall B, Rathmann W, Heier M et al Impact of weight and weight change on normalization of prediabetes and on persistence of normal glucose tolerance in an older population: the KORA S4/F4 study. Int J Obes 2012; 36: 826–33. [DOI] [PubMed] [Google Scholar]
- 25. Bonora E, Tuomilehto J. The pros and cons of diagnosing diabetes with A1c. Diabetes Care 2011; 34(Suppl 2): S185–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Bourdel‐Marchasson I, Dubroca B, Letenneur L et al Incidence and predictors of drug‐treated diabetes in elderly French subjects. The PAQUID Epidemiological Survey. Diabet Med 2000; 17: 675–81. [DOI] [PubMed] [Google Scholar]
- 27. de Vegt F, Dekker JM, Jager A et al relation of impaired fasting and postload glucose with incident type 2 diabetes in a Dutch population: The Hoorn Study. JAMA 2001; 285: 2109–13. [DOI] [PubMed] [Google Scholar]
- 28. Eskesen K, Jensen MT, Galatius S et al Glycated haemoglobin and the risk of cardiovascular disease, diabetes and all‐cause mortality in the Copenhagen City Heart Study. J Intern Med 2013; 273: 94–101. [DOI] [PubMed] [Google Scholar]
- 29. Ferrannini E, Natali A, Capaldo B, Lehtovirta M, Jacob S. Yki‐Järvinen. Insulin resistance, hyperinsulinemia, and blood pressure: role of age and obesity. European Group for the Study of Insulin Resistance (EGIR). Hypertension 1997; 30: 1144–9. [DOI] [PubMed] [Google Scholar]
- 30. Shimizu S, Kawata Y, Kawakami N, Aoyama H. Effects of changes in obesity and exercise on the development of diabetes and return to normal fasting plasma glucose levels at one‐year follow‐up in middle‐aged subjects with impaired fasting glucose. Environ Health Prev Med 2001; 6: 127–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. de Boer IH, Bangalore S, Benetos A et al Diabetes and hypertension: a position statement by the American Diabetes Association. Diabetes Care 2017; 40: 1273–84. [DOI] [PubMed] [Google Scholar]
- 32. Nagoshi T, Yoshimura M, Rosano GMC, Lopaschuk GD, Mochizuki S. Optimization of cardiac metabolism in heart failure. Curr Pharm Des 2011; 17: 3846–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Nakasone Y, Miyakoshi T, Sato Y et al Impact of weight gain on the evolution and regression of prediabetes: a quantitative analysis. Eur J Clin Nutr 2017; 71: 206–11. [DOI] [PubMed] [Google Scholar]
- 34. Hardy OT, Czech MP, Corvera S. What causes the insulin resistance underlying obesity? Curr Opin Endocrinol Diabetes Obes 2012; 19: 81–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Arem H, Moore SC, Patel A et al Leisure time physical activity and mortality: A detailed pooled analysis of the dose‐response relationship. JAMA Intern Med 2015; 175: 959–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Perreault L, Temprosa M, Mather KJ et al Regression from prediabetes to normal glucose regulation is associated with reduction in cardiovascular risk: results from the Diabetes Prevention Program outcomes study. Diabetes Care 2014; 37: 2622–31. [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.
Data Availability Statement
The data that support the findings of this study are available from the SNAC‐K but restrictions apply to the availability of these data, which were used under licence for the current study, and so are not publicly available. The data are, however, available from the authors upon reasonable request and with permission of the SNAC‐K study.