In the treatment of diabetes mellitus which often results in devastating vascular complications and premature death, early intervention is strongly recommended [1]. We here examined the legitimacy of this recommendation, because there has been no clinical study in which the relationship between untreated duration of diabetes and the treatment outcome was critically evaluated.

Stable outpatients with type 2 diabetes [2] in whom drug therapy had not been performed were studied. Those with glycohaemoglobin A1c (HbA1c) >7.0% after 2 months of a run-in period with recommendations on exercise (a minimum of 30 min walking day–1 ≅ 160 kcal day–1) and diet (25–30 kcal kg–1 ideal body weight/day) were selected. All patients were treated under the same treatment protocol (see below) with a target HbA1c <6.5%. Patients older than 75 years and those with advanced vascular complications such as proliferative retinopathy, nephrotic syndrome and/or overt azotaemia, or severe nondiabetic diseases were excluded. A total of 111 patients who fulfilled the entry criteria were studied.

A 75-g oral glucose tolerance test (OGTT) was performed after an overnight fast in the morning in all patients before institution of therapy. In the patients whose HbA1c was successfully lowered to the target value, <6.5%, the OGTT was repeated upon follow-up. At the second OGTT, glibenclamide was withdrawn only for that morning.

In each patient, time of diagnosis of diabetes was asked, and ‘untreated duration of diabetes’ was calculated. Regarding family history of diabetes, presence of it within the third degree relatives was questioned. Other pretreatment clinical variables such as age, HbA1c, the BMI, glucose-induced insulin secretion indexed by the ratio of change in insulin to change in glucose from 0 to 30 min on 75 g OGTT (ΔIRI30/ΔPG30) [3], insulin resistance determined by the homeostasis model assessment (HOMA-R) [4], and fasting serum insulin/C-peptide ratio were also incorporated into the analysis. Glibenclamide monotherapy was performed for 10.1 ± 5.3 months (mean ± SD, the minimum of 4.5 months). The initial dose of glibenclamide was 1.25 mg day–1 and it was increased up to 10 mg day–1. Failure to attain the target HbA1c value was judged after at least 4 months of treatment with 10 mg day–1 glibenclamide.

First, the patients were classified into two groups based on the attainment of target HbA1c value, <6.5%. The patients in whom the target HbA1c (<6.5%) was attained and not attained were classified as good responders (group I) and poor responders (group II), respectively (Table 1). Next, we stratified the good responders (patients in group I, n=69) based on the post-treatment metabolic profiles. To this end, the patients were put in order based on the post-treatment ΔIRI30/ΔPG30 and HOMA-R. Then, those belonging to the upper and lower halves in each determination were separated: group IA, those belonging to the upper half ΔIRI30/ΔPG30 and lower half HOMA-R; group IB, those belonging to the upper half ΔIRI30/ΔPG30 and upper half HOMA-R, or lower half ΔIRI30/ΔPG30 and lower half HOMA-R; group IC, those belonging to the lower half ΔIRI30/ΔPG30 and upper half HOMA-R. By this grouping, it is obvious that metabolic profile is best, intermediate and worst in groups IA, IB, and IC, respectively.

Table 1.   Initial clinical data of the good responders (group I) and the poor responders (group II) Thumbnail image of

Statistical analysis was performed by StatView (SAS®, Cary, NC, USA) and < 0.05 was considered significant. Data are presented as mean ± SD.

By the binominal multiple logistic regression analysis taking the pretreatment clinical characteristics as independent variables and the attainment of HbA1c <6.5% as a dependent phenomenon, only longer untreated duration of diabetes was a significant risk for preclusion of attainment of the target HbA1c, <6.5% (Table 1). Regression coefficient was 0.327 implying that with 1-year increase in untreated duration, the risk for not attaining HbA1c <6.5% increased 39%. None of other pretreatment variables was a significant risk for preclusion of attainment of the target HbA1c. There is a highly significant positive correlation between the untreated duration and the HbA1c attained (< 0.0001, n=111, r2=0.210). The majority (76%, 31/41) of the patients with the untreated duration ≤3 years attained HbA1c < 6.5%, whereas only 34% (24/70) of those with the untreated duration ≥4 years could attain this target value.

The untreated duration was shortest, intermediate, and longest in groups IA, IB and IC, respectively, and by the multinominal multiple logistic regression analysis, the untreated duration was a significant risk for unfavourable metabolic features. The pretreatment HOMA-R, HbA1c and fasting serum immunoreactive insulin/C-peptide immunoreactivity (IRI/CPR) ratio were also significantly related to the classification of the good responders into the three groups. HOMA-R was lower in group IA than in groups IB and IC. Rather unexpectedly, the pretreatment HbA1c was lower in group IB than in groups IA and IC: the values were comparable in the latter two groups. IRI/CPR ratio was highest in group IA, intermediate in group IB, and lowest in group IC. Other initial clinical variables such as age, the BMI, and ΔIRI30/ΔPG30 were not correlated with disposition of the good responders into the three groups. Glucose-induced insulin secretion (indexed by ΔIRI30/ΔPG30) was greater after the treatment than before the treatment in 84% of the patients in group I, however, the value was unchanged or lower after the treatment in 16% of them. The untreated duration was shorter in those with improved β-cell function and the difference was marginally significant (P=0.046). There was a weak but significant negative correlation between untreated duration and the post-treatment ΔIRI30/ΔPG30 P=0.024, r2=0.207). On the other hand, HOMA-R-values were lower (improved) after the treatment in 83% of the patients but unchanged or higher (worsened) in 17% of them. Untreated duration was not significantly different between the two groups and there was no correlation between untreated duration and the post-treatment HOMA-R-values.

Clearly, longer untreated duration was a risk for not attaining good glycaemic control by glibemclamide monotherapy. In other words, glycaemic control was easier in those with shorter untreated duration of diabetes. For poorer restoration of insulin secretion and insulin sensitivity amongst the good responders, again, longer untreated duration was a strong risk. It is likely that, irreversibility in β-cell dysfunction, and less importantly that in insulin resistance, may be built up with chronic hyperglycaemia as time goes by. This process appears more strongly dependent upon the duration than the degree of hyperglycaemia within the respective ranges observed in this study.

Serum insulin concentration was clearly increased after glibenclamide therapy in many patients as reported [3]. However, in 11 patients with longer untreated duration, glucose tolerance became near-normal (HbA1c < 6.5%) after glibenclamide monotherapy in the face of no increase or even significant decrease of serum insulin concentration. This phenomenon, improvement of glycaemia by sulphonylurea without increase in serum insulin, has also been reported by other investigators [5]. Glibenclamide might have increased insulin secretion shortly after the initiation of treatment, and as a consequence, plasma glucose (PG) was lowered. Lowering of PG in turn resulted in lesser stimulation of pancreatic β-cell. Thus, in these patients, hyperglycaemia might have been priming the β cell. Alternatively, extrapancreatic action(s) of sulphonylurea [6] may have produced lowering of PG without increase in insulin secretion.

There is a possibility that β cell failure is progressive irrespective of treatment in type 2 diabetes [7]. However, the current data strongly suggest that the control of hyperglycaemia is easier if the treatment is started earlier. We conclude that earlier intervention of type 2 diabetes is of prime importance and should be listed as one of the most important strategies in ‘the war against diabetes’ [8].


  1. Top of page
  2. References
  • 1
    Prevention of Diabetes Mellitus. Report of a WHO Study Group. Technical Report Series, 844. Geneva: WHO, 1994.
  • 2
    Alberti KGMM & Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus. Provisional report of a WHO consultation. Diabet Med 1998; 15: 53953.
  • 3
    Kosaka K, Kuzuya T, Akanuma Y, Hagura R. Increase in insulin response after treatment of overt maturity-onset diabetes is independent of the mode of treatment. Diabetologia 1980; 18: 238.
  • 4
    Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and β-cell function from fasting plasma glucose and insulin concentration in man. Diabetologia 1985; 28: 4129.
  • 5
    Reaven G & Dray J. Effect of chlorpropamide on serum glucose and immunoreactive insulin concentrations in patients with maturity-onset diabetes mellitus. Diabetes 1967; 15: 48792.
  • 6
    Beck-Nielsen H, Hother-Hielsen O, Pederson O. Mechanism of action of sulphonylureas with special reference to the extrapancreatic effect: an overview. Diabet Med 1988; 5: 61320.
  • 7
    UKPDS Study Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 1998; 352: 83753.
  • 8
    Safran MA & Vinicor F. The war against diabetes. How will we know if we are winning? Diabetes Care 1999; 22: 50816.