To elucidate the glycemic response and antibody formation in gestational diabetic women treated with insulin injected by a needle or a jet. The American Diabetes Association's position statement on jet injectors raised the concern that “insulin could be denatured as a result of forceful injection through a tiny port, which could lead to an increase in antibody formation” (Diabetes Care11:600, 1988). However, the pharmacokinetics of jet-injected insulin suggest that it might be useful in controlling postprandial glucose levels.


We randomized 20 women with gestational diabetes mellitus (<34 wk gestation) who required insulin to receive either jet-injected or needle-injected human NPH and regular insulin. Variables of interest were evaluated at the start of therapy, weekly until delivery, and 6-wk postpartum that included: 1) insulin antibodies in the mother and her infant, 2) HbA1c, 3) insulin dose, 4) fasting and postprandial glucose levels, and 5) subject acceptance and preference.


Of the 10 women in the needle group, 6 developed significant insulin antibodies compared with 1 of 10 in the jet group (P < 0.001). HbA1c and insulin doses were the same in both groups. During the test meal, glucose levels in the jet group were significantly lower (P < 0.01), yet none of the women in the jet group experienced blood glucose 70 mg/dl (3.89 mM) at 3–4 h after the meal, compared with 5 in the needle group (P < 0.001). Jet injection was associated with less variability (P < 0.001) in postprandial glucose values but slightly greater variability (P < 0.05) in fasting glucose. Jet-injected insulin was more readily accepted by subjects than needle injections.


Jet injection is associated with a diminished antibody response and postprandial variability compared with needle-injected insulin. Thus, this warrants consideration as a therapeutic option for women with gestational diabetes mellitus and may also be applicable to nonpregnant, insulin-requiring diabetic patients.

This content is only available via PDF.