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


Multiple insulin injection therapy delayed microvascular complications in NIDDM

ACP J Club. 1996 May-June;124:68. doi:10.7326/ACPJC-1996-124-3-068

Source Citation

Ohkubo Y, Kishikawa H, Araki E, et al. Intensive insulin therapy prevents the progression of diabetic microvascular complications in Japanese patients with non-insulin-dependent diabetes mellitus: a randomized prospective 6-year study. Diabetes Res Clin Pract. 1995 May;28:103-17.



To compare multiple insulin injection therapy (MIT) with conventional therapy (CT) for prevention of microvascular complications in Japanese patients with non-insulin-dependent diabetes mellitus (NIDDM).


6-year randomized controlled trial.


Outpatient clinic in Kumamoto, Japan.


110 patients who had NIDDM and received 1 to 2 daily injections of intermediate-acting insulin. Exclusion criteria were age > 70 years, neuropathy, nephropathy, or severe retinopathy. Patients were divided into a primary prevention cohort (n = 55) (no retinopathy or microalbuminuria) and a secondary intervention cohort (n = 55) (simple retinopathy or microalbuminuria). 102 patients (93%) (mean age 49 y, 52% women) completed the study.


Patients were allocated to CT (n = 55) or to MIT (n = 55). CT consisted of 1 to 2 daily insulin injections to maintain a fasting blood glucose level < 7.8 mmol/L without symptoms of hyperglycemia or hypoglycemia. MIT consisted of ≥ 3 daily insulin injections to maintain a fasting blood glucose level < 7.8 mmol/L, postprandial level < 11.1 mmol/L, hemoglobin A1c level < 7.0%, and mean amplitude of glycemic excursions < 5.6 mmol/L.

Main Outcome Measures

Development and progression of retinopathy, nephropathy, and neuropathy.

Main Results

Near normoglycemia was achieved within 3 months and was maintained throughout the study by patients receiving MIT. MIT led to fewer patients developing worsening diabetic retinopathy than did CT (13% vs 38%, P = 0.007). {This absolute risk reduction (ARR) of 25% means that 4 patients would need to receive (NNT) MIT (compared with CT) to prevent 1 additional patient from developing retinopathy, 95% CI 2 to 13; the relative risk reduction (RRR) was 65%, CI 26% to 84%.}* MIT also led to fewer patients developing diabetic nephropathy (10% vs 30%, P = 0.005) {ARR 20%; NNT 5, CI 3 to 19; RRR 68%, CI 22% to 87%}*. The differences in retinopathy and nephropathy were similar in both cohorts. Nerve conduction velocity improved over 6 years in patients receiving MIT (P < 0.05 for MIT compared with CT). The groups did not differ for episodes of hypoglycemia.


Intensive glycemic control with multiple insulin injections delayed the onset and progression of retinopathy, nephropathy, and neuropathy in Japanese patients with NIDDM.

Source of funding: Ministry of Health and Welfare, Japan.

For article reprint: Dr. Y. Ohkubo, Department of Metabolic Medicine, Kumamoto University School of Medicine, 1-1-1 Honjo, Kumamoto 860, Japan. FAX 81-96-366-8397.

*Numbers calculated from data in article.


The Diabetes Control and Complications Trial (DCCT) (1) showed conclusively that intensive diabetes management dramatically reduced the development and progression of diabetic retinopathy, nephropathy, and neuropathy in insulin-dependent diabetes mellitus (IDDM). A major question that arose out of the DCCT study is whether these findings are applicable to patients with NIDDM. Using a similar study design, Ohkubo and colleagues evaluated the effects of intensive insulin therapy in the form of multiple insulin injections on microvascular complications in 110 Japanese patients with NIDDM. The separation in glycemia and hemoglobin A1c level was almost identical to that achieved in the DCCT, and the benefits for microvascular disease were very closely mirrored. The authors thus conclude that intensive therapy is equally effective in preventing the microvascular complications of diabetes in NIDDM and IDDM.

The conclusion of the authors may well be correct; however, several aspects of this study require closer scrutiny. Although the patients were described as having NIDDM, they do not share the phenotypic characteristics commonly seen in European and North American patients with NIDDM. Specifically, the body mass index was in the low normal range, and the insulin dose was similarly low, approximately 50% of that observed in the DCCT and much lower than the usual insulin requirements of patients with NIDDM. In addition, the patients in this study do not appear to share the common characteristics of central obesity, insulin resistance, hypertension, elevated triglyceride levels, and low high-density lipoprotein levels frequently seen in Caucasian, African American, Hispanic, and aboriginal persons with NIDDM.

Despite these limitations, the study by Ohkubo and colleagues is consistent with other observations that showed that hyperglycemia, irrespective of its cause (IDDM, NIDDM, pancreatectomy, or hemochromatosis), has similar effects on the microvascular complications of diabetes. The clinical implications are clear, namely, that the management of diabetes—both IDDM and NIDDM—requires a serious and concerted effort to obtain the best possible glycemic control to improve long-term outcome.

Bernard Zinman, MD
University of TorontoToronto, Ontario, Canada


1. The Diabetes Control and Complications Trial Research Group. N Engl J Med. 1993;329:977-86.