Current issues of ACP Journal Club are published in Annals of Internal Medicine


Nondiabetic men with high blood glucose levels were at increased risk for death

ACP J Club. 1998 Jul-Aug;129:18. doi:10.7326/ACPJC-1998-129-1-018

Related Content in the Archives
Men with type 2 diabetes mellitus or impaired glucose tolerance had an elevated risk for early death

Source Citation

Balkau B, Shipley M, Jarrett RJ, et al. High blood glucose concentration is a risk factor for mortality in middle-aged nondiabetic men. Diabetes Care. 1998 Mar;21:360-7.



Are men with high but nondiabetic blood glucose levels at increased risk for death?


Joint analysis of 3 cohort studies with 20-year follow-up.


Occupation-based studies in London, Paris, and Helsinki.


Working men who were 44 to 55 years of age at baseline (1967 to 1974) (mean age 49 y) and were participants in the Whitehall Study (n = 10 025), the Paris Prospective Study (n = 6629), or the Helsinki Policeman Study (n = 631). Men were classified as nondiabetic if, on a baseline oral glucose tolerance test, the 2-hour blood glucose level was < 11.1 mmol/L in the Paris and Whitehall studies and < 10 mmol/L in the Helsinki study.

Assessment of risk factors

Men with 2-hour blood glucose levels > the 80 percentile were compared with men with levels ≤ the 80 percentile. Hazard ratios (HRs) were adjusted for age, body mass index, systolic blood pressure, hypertension treatment, baseline cardiovascular disease, total cholesterol levels, smoking status, and administrative grade.

Main outcome measures

All-cause mortality and death from cardiovascular disease (CVD), coronary heart disease (CHD), cerebrovascular disease, and neoplasms according to blood glucose levels.

Main results

The combined age-adjusted HR for all-cause mortality was higher in men with 2-hour blood glucose levels > the 80 percentile than in men with glucose levels ≤ the 80 percentile. The HR was 1.15 (95% CI 1.03 to 1.27) for glucose levels in the 80 to 90 percentiles and increased to 1.63 (CI 1.38 to 1.94) for glucose levels ≥ the 97.5 percentile. The association with all-cause mortality remained after adjustment for other risk factors (HR 1.11, CI 1.00 to 1.23 for the 80 to 90 percentiles). The increased risk for death from CVD, CHD, cerebrovascular disease, and neoplasms consistently reached statistical significance at the 97.5 percentile (age-adjusted HRs for CVD 1.77, CI 1.38 to 2.26; for CHD 1.81, CI 1.36 to 2.41; for cerebrovascular disease 2.16, CI 1.14 to 4.11; and for neoplasms 1.43, CI 1.05 to 1.94). After adjustment for other risk factors, the association remained for death from CVD (HR 1.30, CI 1.01 to 1.67) and CHD (HR 1.34, CI 1.00 to 1.80). When HRs were calculated on the basis of fasting blood glucose levels (available for the Helsinki and Paris studies), the risk for all-cause mortality was evident in the upper 2.5 percentile of the glucose distributions.


Nondiabetic middle-aged men with 2-hour blood glucose levels > the 80 percentile were at increased risk for all-cause mortality over a 20-year follow-up period. Risk for death from coronary heart disease was moderately increased when fasting or 2-hour blood glucose levels were > the 97.5 percentile.

Source of funding: Not stated.

For correspondence: Dr. B. Balkau, INSERM U21, 16 avenue Paul Vaillant Couturier, F-94807 Villejuif Cedex, France. FAX 33-1-47-26-94-5. E-mail


The study by Balkau and colleagues concerns the difficult issue of prevention among patients who do not have diabetes according to standard criteria but whose fasting or 2-hour glucose levels approach diabetic diagnostic levels. If nondiabetic men in this study have an increased risk for mortality, especially if it is caused by a single factor, such as CHD, preventive strategies might be used to decrease this risk.

Unfortunately, these combined data do not really answer this preventive question. Some inconsistency exists among the 3 cohort studies in the relative contributions of neoplastic (Paris) and cardiovascular (Helsinki and Whitehall) causes to overall mortality. CHD mortality is not increased until glucose levels are > the 97.5 percentile, thus showing no graded response as glucose levels increase. In addition, the graded increase in overall mortality as one ascends the percentiles of glucose levels would most likely not be statistically significant without the neoplastic deaths. It is hard to focus preventive strategies to improve survival when one is dealing with both neoplastic and CHD deaths.

With respect to CHD risk, an increased glucose level is probably a marker for early insulin resistance with the associated CHD risk factors, including increased triglyceride levels; low high-density lipoprotein (HDL) cholesterol levels; small, dense, low-density lipoprotein particles; and increased coagulability (increased plasminogen activator inhibitor and factor VIIc levels).

Because these increased glucose levels suggest insulin resistance, weight loss and exercise may reduce the long-term risk for CHD by ameliorating the insulin resistance. High-risk men with high-normal glucose levels may be differentiated from low-risk men by identification of other known factors associated with insulin resistance (obesity or family history of diabetes) or by the presence of other markers of insulin resistance (increased blood pressure, increased triglyceride levels, or decreased HDL cholesterol levels). Whether reducing insulin resistance in nondiabetic men will actually decrease mortality awaits long-term interventional studies.

Donald A. Smith, MD
Mount Sinai Medical CenterNew York, New York, USA