Most complications of gestational diabetes are now attributed to accelerated fetal growth before 28 to 30 weeks of gestation. Prompt, effective treatment is required to minimize complications, but there is controversy concerning the optimal intervention. Most women are initially offered dietary treatment; insulin therapy is usually reserved for use in those with severe cases and in those who do not respond to a reasonable trial of dietary control. This approach may delay effective therapy for mothers at high risk. McFarland and colleagues studied the optimal duration of diet therapy and sought to identify factors that predict which women need insulin therapy for gestational diabetes.

The authors studied 269 women attending a university obstetric clinic from 1994 to 1995. Women were screened for gestational diabetes between 24 and 28 weeks of gestation unless they had risk factors (e.g., history of gestational diabetes, stillbirth or congenital malformation, and those with a body mass index greater than 27), in which case they were screened at the first clinic visit. A three-hour glucose tolerance test was performed if the plasma glucose level was greater than 130 mg per dL (7.2 mmol per L) on screening. Diagnosis of gestational diabetes was based on the criteria of the National Diabetes Data Group.

The women studied had no history of diabetes and had fasting glucose levels below 140 mg per dL (7.8 mmol per L). These women were prescribed diets for four weeks, but if the mean weekly glucose level exceeded 200 mg per dL (11.1 mmol per L), they were immediately changed to insulin therapy. After the four-week dietary intervention, women whose blood glucose was in poor control (over 105 mg per dL [5.8 mmol per L]) were prescribed insulin therapy.

Diets were constructed to contain 50 to 55 percent carbohydrates, 20 to 25 percent protein and 20 to 25 percent fat. Each woman received intensive dietary instruction and was counseled every week. The patients were instructed to keep food diaries, measure their blood glucose levels seven times daily and check for morning ketonuria.

The authors found that maternal weight and body mass index were related to the degree of hyperglycemia. Over 85 percent of the women reported compliance with their dietary regimens, and glucose was monitored an average of 4.7 times daily during the four-week intervention. No insulin was prescribed during the four-week dietary intervention, but 43 percent of the women eventually required insulin for glycemic control. Statistical analysis showed that the cutoff point that best predicted need for insulin was a fasting value on a glucose tolerance test of around 95 mg per dL (5.3 mmol per L). Most women with lower levels achieved good glycemic control within two weeks using diet alone.

The authors conclude that women with gestational diabetes should attempt glycemic control by diet alone for at least two weeks. In those with fasting glucose levels above the cutoff point of 95 mg per dL, insulin is likely to be required and one week of dietary intervention may be a more appropriate trial period. Women who have additional risk factors (e.g., obesity) and fasting levels above the cutoff point should be considered for immediate insulin therapy.