Elevated Serum Fructosamine Concentration as an Index of Risk State of Abnormal Glucose Metabolism for Coronary Atherosclerosis

To examine the usefulness of hyperfructosaminemia as an index of risk state of abnormal glucose metabolism for coronary atheroscrelotic disease, serum fructosamine concentration was compared between 130 male cases with coronary stenosis, aged 60 years or younger, and 260 age-matched male controls. Past history of diabetes mellitus (DM) and information on confounders such as serum lipid concentration, obesity, smoking habits, and family history of ischemic heart disease were obtained from all the subjects. There was no difference in frequency of the past history of DM between the cases (8.5%) and the controls (6.9%), but that of hyperfructosaminemia (cases : 36.2%, controls : 15.0%) was significantly higher in the cases than in the controls (p<0.01). When confounding effects by other risk factors were adjusted by multiple logistic regression analysis, the odds ratio of the past history of DM (1.19, 95% confidence interval : 0.47-3.03) was not significant, but that of hyperfructosaminemia (3.59, 95% confidence interval : 1.94.6.65) was significant after the past history of DM was replaced by it. Thus, abnormal glucose metabolism evaluated by hyperfructosaminemia was an independent coronary risk factor, although that evaluated by the past history of DM was not. These results suggest that serum fructosamine is a useful index of coronary risk state related to abnormal glucose metabolism. J Epidemiol, 1992; 2 : 105-110.

To examine the usefulness of hyperfructosaminemia as an index of risk state of abnormal glucose metabolism for coronary atheroscrelotic disease, serum fructosamine concentration was compared between 130 male cases with coronary stenosis, aged 60 years or younger, and 260 age-matched male controls.Past history of diabetes mellitus (DM) and information on confounders such as serum lipid concentration, obesity, smoking habits, and family history of ischemic heart disease were obtained from all the subjects.
There was no difference in frequency of the past history of DM between the cases (8.5%) and the controls (6.9%), but that of hyperfructosaminemia (cases : 36.2%,controls : 15.0%) was significantly higher in the cases than in the controls (p<0.01).
When confounding effects by other risk factors were adjusted by multiple logistic regression analysis, the odds ratio of the past history of DM (1.19, 95% confidence interval : 0.47-3.03)was not significant, but that of hyperfructosaminemia (3.59, 95% confidence interval : 1.94.6.65) was significant after the past history of DM was replaced by it.Thus, abnormal glucose metabolism evaluated by hyperfructosaminemia was an independent coronary risk factor, although that evaluated by the past history of DM was not.These results suggest that serum fructosamine is a useful index of coronary risk state related to abnormal glucose metabolism.J Epidemiol, 1992; 2 : 105-110.
coronary artery disease, risk factor, serum fructosamine, abnormal glucose metabolism, diabetes mellitus Abnormality of glucose metabolism such as impaired glucose tolerance and diabetes mellitus (DM) has been known to be an important risk factor for coronary artery disease1-3).In our previous study4), DM was also determined to be a coronary risk factor in the young men aged 40 years or younger, but was not in the elderly men aged 41 to 60 years5).In that study, however, information concerning an abnormality in the glucose metabolism had been given by the memory of participants as to the past history of DM, and by either the fasting plasma glucose level or the result of oral glucose tolerance test which were not performed in every subject.In addition, the past history of DM might have been more accurately screened in the controls than in the cases because majority of the former is thought to have received health check-up annually.This might be the reason that DM did not come out as the risk factor in the elderly in the previous study4).
Thus in the present study, we tested reliability of the serum fructosamine concentration, well known as a good indicator of mean blood glucose level during I to 2 weeks before determination6-8), as the objective index for evaluating abnormal glucose metabolism.
Then, the importance of the abnormal glucose metabolism, estimated by serum fructosamine level, as a coronary risk factor was investigated, and the usefulness of Five cases and four controls were excluded for comparison of body mass index because of a lack of information about their height and/or weight.
(case : 48.4 years, control : 48.2 years) and serum total cholesterol (case : 194.0 mg/dl, control : 192.8 mg/dl) were not different between the cases and the controls.Serum fructosamine well showed an approximate normal distribution both in the cases and the controls (Fig. 1) Profiles of coronary risk factors in the cases and the controls are shown in Table 2.Although the rate of those who had a past history of DM was not different between the cases (8.5%) and the controls (6.9%), that of hyperfructosaminemics was significantly higher in the cases (36.2%) than in the controls (15.0%,p<0.01).
Confounders such as hypo-HDL-cholesterolemia (p<0.01),smoking habit (p<0.01), a past history of hypertension (p<0.01) and the family history of ischemic heart disease (p<0.05) were present significantly more often in the cases than in the controls.The frequency of obesity also tended to be higher in the  Five cases and four controls were excluned for comparison of the frequency of obesity because of a lack of information about their height and/or weight.
cases (p<0.10)but that of hypercholesterolemics was not different between the two groups.When frequency of hyperfructosaminemics was compared between subgroups with and without the past history of DM, they occupied the major part of the subgroup with the past history both in the cases (81.8%) and in the controls (61.1%).These percentages were significantly higher (p<0.01)than the respective ones observed in the cases (31.9%) and in the controls (11.6%) of the subgroup without a past history (Table 3).
In a multiple logistic regression analysis in which The number of subjects was 125 in the cases and 256 in the controls One control of young men and five cases and three controls of elderly men were excluded for comparison of the frequency of obesity because of the same reason as in Table 1. 1) By Fisher's exact probability test.The number of subjects was 43 in the cases and 85 in the controls among the young men, and 82 in the cases and 171 in the controls among the enderly men atory variables as used in the whole group were entered.

DISCUSSION
Serum fructosamine concentration has been used as a short term indicator of blood glucose level for the control of DM6-8).In addition, serum fructosamine concentration has been proven to be significantly higher in diabetics than in normoglycemics6).Therefore, it would be useful to employ serum fructosamine concentration to estimate the glucose metabolism abnormality.
In previous studies2-3) impaired glucose tolerance, a milder form of hyperglycemia than DM, was shown to be a coronary risk factor as well.Hence we set the cut off level of hyperfructosaminemia at a relatively low point, i.e., mean plus one standard deviation of the control subjects.Hyperfructosaminemics screened by this method in those with a past history of DM accounted for more than 80% of the cases and 60% of the controls.
Although these percentages were significantly higher than the respective values of 'those who had no past history of DM, hyperfructosaminemics occupied 32% of the cases and 12% of the controls even in the latter group.This result supports that not only DM but also a milder form or a latent state of glucose metabolism abnormality could be screened as hyperfructosaminemia by our criteria.
When we obtained the odds ratio, hyperfructosaminemia was approved as an independent risk factor for coronary artery disease, whereas a past history of DM was not.In our previous study5), DM did not show a significant odds ratio as a coronary risk factor in the elderly men.In that study, the presence of DM was judged not only by a past history of DM but also by a fasting blood glucose level or by the result of an oral glucose tolerance test.However, not all the cases and none of the controls were subjected to the latter two tests.Despite such a situation which would have caused an overestimation of risk of DM, the odds ratio was insignificant.
In this study, however, hyperfructosaminemia was approved as a coronary risk factor in the elderly group, too.These findings would indicate that in assessment of coronary risk status, hyperfructosaminemia is a better index than the information based on a past history and either fasting blood glucose level or an oral glucose tolerance test performed non-uniformly.
The oral glucose tolerance test has been an established method to diagnose the abnormal glucose metabolism, and the abnormal result in this test has been proven to be a potent risk factor for coronary artery disease2,3).This test is also useful in evaluating resistivity of insulin response which has also shown to be a risk factor for coronary artery diseases9,10) However, this method requires a fairly long time of fasting before application and needs two hours until completion of the test.Therefore, it is difficult to be utilized in a health check-up for a large population.On the other hand, concentration of fructosamine in the serum and glycosylated hemoglobin components of total hemoglobin 20), which are indicators of glycosylation of protein, can be determined in the blood sample obtained from non fasting subjects.Furthermore, concentration of fructosamine can be determined in the stored serum.Therefore, it is easily applicable to a health check-up of a large population.
Recently, glycosylation or glycation of various proteins is considered to play an important role in the genesis of atherosclerosis1-19).
Since serum fructosamine concentration reflects the total concentration of all types of glycosylated protein in the serum, it is possible that serum fructosamine concentration is related to the amount of such atherogenic glycosylated proteins.If so, hyperfructosaminemia may become an index to assess atherogenecity itself apart from an index for hyperglycemia.
Figure 1.Distributions of serum fructosamine levels.Dotted line indicates 285 ƒÊmol/1 which is the value of mean plus one standard deviation of the controls.

Table 1 .
Mean values and standard deviations of coronary risk factors in the cases (n = 130) and the controls (n=260).

Table 2 .
Profile of coronary risk factors in the cases (n= 130) and the controls (n=260).

Table 3 .
Frequency of hyperfructosaminemia (%) in the subgroup with or without a past history of diabetes mellitus.

Table 6 .
Results of multiple logistic regression analysis (model 2) among each subgroup classified by age .