Insulin resistance (IR) is a characteristic feature of non-insulin-dep
endent diabetes mellitus (NIDDM) as well as obesity, and a majority of
NIDDM patients are obese. To assess the effect of obesity independent
of NIDDM on IR, we studied the relationship between IR and obesity in
65 normal and 58 NIDDM subjects; we used body mass index (BMI) as a m
easure of obesity and glucose infusion rate (GINF) during a euglycemic
hyperinsulinemic (120 mU . m(-2). min(-1)) glucose clamp as a measure
of IR. In lean normal subjects, GINF was 57.7 +/- 2.2 mu mol . kg(-1)
. min(-1) (10.4 +/- 0.4 mg . kg(-1). min(-1)) and the lean NIDDM subje
cts were markedly insulin-resistant, with a GINF of 34.4 +/- 2.8 mu mo
l . kg(-1). min(-1) (6.2 +/- 0.5 mg . kg(-1). min(-1)). Obese normal s
ubjects were also insulin-resistant compared with lean normal subjects
, with a GINF of 36.1 +/- 2.2 mu mol . kg(-1). min(-1) (6.5 +/- 0.4 mg
. kg(-1). min(-1)), and obesity caused an increase in IR in NIDDM, wi
th a GINF of 21.1 +/- 1.4 mu mol . kg(-1). min(-1) (3.8 +/- 0.25 mg .
kg(-1). min(-1)) in the obese NIDDM subjects. Therefore, similar to 61
% of the IR in obese NIDDM subjects is due to NIDDM, with 39% due to o
besity, demonstrating a greater impact of NIDDM than of obesity in cau
sing IR. The correlation between GINF and BMI was much better in norma
l subjects (r = -0.75) than in NIDDM subjects (r = -0.50) as was the r
elationship between fasting insulin level and BMI (r = -0.59 in normal
subjects, r = -0.48 in NIDDM subjects). As expected, the fasting insu
lin level was also strongly correlated to GINF in normal subjects (r =
-0.61); however, this relationship was weaker in NIDDM subjects (r =
-0.46). In conclusion, 1) obesity has a major impact to cause insulin
resistance in nondiabetic subjects, but the effect of obesity on iii i
n NIDDM is less; 2) NIDDM per se is the major contributor to IR in NID
DM; and 3) the fasting insulin level is a better surrogate marker of I
R in nondiabetic subjects than in NIDDM patients.