Strandberg TE, Pienimäki T, Strandberg AY, Salomaa VV, Pitkälä KH, Tilvis RS, Miettinen TA. One-Hour Glucose, Mortality, and Risk of Diabetes: A 44-Year Prospective Study in Men. Arch Intern Med. 2011;171(10):941-954. doi:10.1001/archinternmed.2011.183
Copyright 2011 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2011
Two-hour postload glucose measurement is currently recommended for early detection of diabetes.1 The less time-consuming 1-hour postload glucose measurement (1-hour glucose) has been reported to predict cardiovascular mortality2 and Medicare expenditure3 and to be associated with carotid intima-media thickness.4 Recently, 1-hour glucose has also aroused interest in the prediction of diabetes both in United States and Finnish cohorts5- 7 and in the assessment of risk for chronic kidney disease.8 However, the status of 1-hour glucose as a risk-predicting tool is still far from established. Because blood 1-hour glucose was used to assess cardiovascular risk in early midlife in the Helsinki Businessmen Study during the 1960s,9 we had a unique opportunity to evaluate how this relatively simple test in conjunction with body mass index (BMI) would predict mortality and development of diabetes over 44 years of follow-up.
This study was approved by the Ethics Committee of the Helsinki University Hospital Department of Medicine. The characteristics of this cohort of mainly business executives, born between 1919 and1934, have been described previously.9 Baseline examinations between 1964 and 1973 included measurements of blood 1-hour glucose level, BMI (calculated as weight in kilograms divided by height in meters squared; normal weight <25, overweight 25-29, and obese ≥30), and cardiovascular risk factors. Follow-up of the cohort for this report is through December 2007. Mortality and causes of deaths were ascertained from national registers and development of diabetes from mailed questionnaire surveys performed between the years 1985 and 2007 and from the national drug reimbursement register. Because results were similar when self-reported cases were added to the register-verified cases of diabetes, all cases were used in the analyses to increase statistical power. Statistical analyses were performed with NCSS 2004 (NCSS, Kaysville, Utah). Statistical significance was defined as a 2-sided P value lower than .05.
A total of 2756 healthy men with 1-hour glucose and without diagnosed diabetes or cardiovascular disease at baseline were included in the analyses, giving a median follow-up time of 37 years (interquartile range [IQR], 29-40 years; maximum 44 years) and 91 066 person-years. Median age at baseline was 42 years (IQR, 39-46 years), median age of survivors by 2007 was 79 years (IQR, 76-83 years). Median blood 1-hour glucose and BMI were 108 mg/dL (to convert glucose to millimoles per liter, multiply by 0.0555) (IQR, 88.2-133 mg/dL), and 25.7 (IQR, 24.1-27.5), respectively. At baseline, BMI (P < .001), systolic and diastolic blood pressure (P < .001), and cholesterol level (P = .002) increased significantly with increasing 1-hour glucose quartiles, while there was no significant difference in ever smokers between quartiles (P = .30).
During follow-up, a total of 1287 men (46.7%) died, 509 men died of cardiovascular disease, and 357 men (13.0%) developed diabetes (68 cases were from questionnaire data only). There was a strong and graded relationship between 1-hour glucose quartiles and both total and cardiovascular mortality (P < .001) and development of diabetes (P < .001) during follow-up. One-hour glucose was not related to noncardiovascular mortality in this cohort. Both BMI and 1-hour glucose (both available for 2709 men [1263 deaths and 355 incident cases of diabetes]) in the same model independently predicted diabetes risk (P < .001). The impact of their combinations on total mortality and diabetes risk during follow-up are given in the Table. Especially, the combination of blood 1-hour glucose level higher than 144 mg/dL (plasma value, >161 mg/dL: the cut point used in the study by Manco et al7) and BMI of 30 or higher was associated with a 10.1-fold increase of diabetes risk independently of cardiovascular risk factors (Table). Only 3.2% (86 of 2705) of the cohort belonged to this subgroup of the highest risk (BMI ≥30 and 1-hour glucose level >161 mg/dL); 84.2% of all incident diabetes (299 of 355) occurred among those who were either overweight (BMI ≥25, 48.2% of cases), had 1-hour glucose level higher than 161 mg/dL (26.8% of cases), or both (9.3% of cases) at baseline. Finally, 43.9% and 81.6% of patients in the lowest and highest baseline risk subgroup, respectively (P < .001), had either died or developed diabetes during follow-up.
Our exceptionally long follow-up demonstrates the very high diabetes and mortality risk associated with elevated 1-hour glucose level measured in early midlife. The risk was independent of traditional cardiovascular risk factors, cholesterol level, blood pressure, and smoking, and it was accentuated when combined with BMI. Our results suggest that diabetes risk could be assessed with a less time and resource-consuming method. Closer examination of high-risk individuals who survive to old age without diabetes and low-risk men who develop diabetes would give further information about pathways to diabetes. The obvious need is to compare 1-hour glucose with fasting and 2-hour postload glucose levels as well as hemoglobin A1c1 in a similar setting and in various populations. Because 1-hour glucose predicted diabetes more strongly than mortality, it is also important to clarify the impact of elevated 1-hour glucose on quality of life during one's life course.
Correspondence: Dr T. E. Strandberg, Institute of Health Sciences/Geriatrics, University of Oulu and University Hospital, PO Box 5000, FIN-90014 Oulu, Finland (email@example.com).
Author Contributions: All authors had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: T. E. Strandberg, Pienimäki, and A. Y. Strandberg. Acquisition of data: T. E. Strandberg and Miettinen. Analysis and interpretation of data: T. E. Strandberg, Salomaa, Pitkälä, and Tilvis. Drafting of the manuscript: T. E. Strandberg, Pienimäki, and A. Y. Strandberg. Critical revision of the manuscript for important intellectual content: Pienimäki, Salomaa, Pitkälä, Tilvis, and Miettinen. Statistical analysis: T. E. Strandberg, A. Y. Strandberg, and Salomaa. Obtained funding: Tilvis and Miettinen. Administrative, technical, and material support: Pitkälä and Tilvis.
Financial Disclosure: Although not directly related to this epidemiological study, some of the authors have had various cooperation with companies also marketing antidiabetic products. Dr T. E. Strandberg has had consultancies and/or received speaker honoraria for Abbott, AstraZeneca, Boehringer, Leiras, Merck, Novartis, and Pfizer; developed educational presentations for Leiras and Novartis; has stock (small amount) in Orion Pharma; and received meeting expenses from Bayer, Leiras, and Merck. Dr Strandberg has developed educational presentations for Leiras and Novartis and owns stock (small amount) in Orion Pharma. Dr Pitkälä has received speaker honoraria from Janssen, Leiras, Lundbeck, Novartis, and Pfizer. Dr Tilvis has received speaker honoraria from Novartis and Pfizer and meeting expenses from Boehringer.
Funding/Support: The study was supported by the Jahnsson Foundation (Dr T. E. Strandberg), the University Central Hospital of Oulu, and the University Central Hospital of Helsinki.
Role of the Sponsors: The funding sources had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.