| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrine Care |
Departments of Medicine (Division of Endocrinology and Metabolism) (G.E.D., P.W.S.), Biostatistics (A.D.H.), and Biochemistry and Molecular Biology (P.W.S.), University of Florida, Gainesville, Florida 32610
Address all correspondence and requests for reprints to: Glen E. Duncan, Ph.D., Box 100226 JHMHSC, University of Florida, Gainesville, Florida 32610-0226. E-mail: gduncan{at}ufl.edu
Abstract
A novel index of insulin sensitivity, the quick insulin sensitivity check index, termed QUICKI (1/[log (insulin) + log (glucose)]), was recently developed. We examined whether QUICKI accurately reflects changes in insulin sensitivity after exercise training, a perturbation known to improve insulin sensitivity. Sedentary, nondiabetic adults underwent a frequently sampled iv glucose tolerance test before and after 6 months of training. Insulin sensitivity was estimated from the glucose tolerance test using Bergman’s minimal model (insulin sensitivity–minimal model), and QUICKI was calculated from basal insulin and glucose. Exercise increased (P = 0.003) insulin sensitivity–minimal model but did not change (P = 0.12) QUICKI. Before and after training, the rank-correlation between QUICKI and insulin sensitivity–minimal model was significant (r = 0.79, P = 0.0005; r = 0.56, P = 0.03, respectively). However, the rank-correlation between fasting insulin alone with insulin sensitivity–minimal model was as good (before training r = -0.77, P = 0.0009; after training r = -0.55, P = 0.03) as that between QUICKI and insulin sensitivity–minimal model. Fasting glucose was not related to insulin sensitivity–minimal model at either time. When difference scores (i.e. after pretraining values) were examined, neither QUICKI nor fasting insulin correlated with insulin sensitivity–minimal model (QUICKI vs. insulin sensitivity–minimal model r = 0.24, P = 0.39; fasting insulin vs. insulin sensitivity–minimal model r = -0.40, P = 0.14). We conclude that fasting insulin is equivalent to fasting insulin plus glucose (i.e. QUICKI) at estimating basal insulin sensitivity in nondiabetic adults. However, QUICKI does not accurately reflect exercise-induced changes in insulin sensitivity within individual subjects.
INSULIN RESISTANCE (IR) is defined as an inappropriately high level of insulin required to maintain metabolic homeostasis (1), and is characterized by decreased insulin sensitivity (SI) or responsiveness to the metabolic actions of insulin. The ability to quantify SI in large groups of individuals is important from a public health standpoint because of the well-established role that IR plays in the pathophysiology of type 2 diabetes mellitus (2). Furthermore, IR is a common feature of many metabolic disturbances associated with coronary heart disease, including central obesity, hypertension, and dyslipidemias (3).
The hyperinsulinemic-euglycemic clamp technique is considered the standard for quantifying SI. However, minimal model (MM) analysis of the frequently sampled iv glucose tolerance test (FSIVGTT) appears to provide an equivalent assessment of overall SI in normal and insulin-resistant nondiabetic subjects (4). Unfortunately, these methods are not applicable to studies employing large numbers of subjects, or for routine medical screening, because of their greater relative expense, time commitment, and invasive nature.
In contrast to these dynamic interventions, steady-state assessments of SI offer an inexpensive and rapid alternative to quantifying SI in a large number of individuals. These methods include a measure of fasting insulin concentration, the homeostasis model assessment, or HOMA, defined as insulin x glucose/22.5 (5), and a recently developed method, the quick insulin sensitivity check index, termed QUICKI, defined as 1/(log [insulin]) + log [glucose] (6).
Katz et al. (6) demonstrated that the overall correlation between the insulin-glucose clamp and the QUICKI estimate of SI was substantially better than the relationship between either the clamp and HOMA or fasting insulin, or between the clamp and the MM estimate of SI (SI-MM) in a group of obese, nonobese, and diabetic subjects combined. These findings suggest that QUICKI is as good as or better at estimating SI than are more invasive (e.g. SI-MM) and noninvasive (e.g. fasting insulin, HOMA) methods. Furthermore, QUICKI offers the advantage of being fast, inexpensive, and applicable to large populations of both diabetic and nondiabetic individuals.
A potential application of QUICKI would be in determining changes in SI that occur owing to specific metabolic interventions. We investigated whether QUICKI could accurately reflect changes in SI after exercise training, which is known to improve SI (7, 8, 9). Specifically, we tested the hypothesis that QUICKI significantly correlates with SI-MM both before and after a 6-month exercise training intervention in a group of sedentary, nondiabetic adults. We also tested the hypothesis that the change in QUICKI significantly correlates with the change in SI-MM within individual subjects after an exercise intervention in nondiabetic adults.
Materials and Methods
Subjects
Subjects for the present study were enrolled in a 6-month exercise intervention designed to examine the dose response of aerobic training on SI and markers of fat metabolism in sedentary adults. Participants were recruited primarily through newspaper advertisement. Interested individuals were provided with a thorough description of the study, and were asked several questions related to their current and past health status, and their current and past physical activity habits. We excluded individuals who had known disease or who were physically active (defined as engaging in structured physical activity on more than two occasions per week, each lasting 30 min or more, over the previous 12 months, and regular brisk walking for exercise or leisure for 20 min or more per occasion). Subjects who passed the preliminary telephone screen were invited to attend an information session at which they completed demographic, health history, and two physical activity questionnaires (the Seven-Day Physical Activity Recall [10 ] and Aerobics Center Longitudinal Study Physical Activity Questionnaire [11 ]) to further establish that they were sedentary. Individuals interested in continuing their participation were scheduled for a series of medical evaluations that included standard tests of cardiac, endocrine, hematological, and metabolic function.
This study was conducted on the General Clinical Research Center, Shands Hospital, at the University of Florida. Before testing, all subjects provided written informed consent, according to the standards established by the Health Science Center Institutional Review Board.
Measures
Anthropometric. Body mass index ([BMI] = kg/m2) was calculated as an index of adiposity. For all BMI measurements, height was measured using a wall-mounted stadiometer and weight on a balance beam scale, both with shoes removed. The waist-to-hip (W:H) ratio was calculated as an index of body fat distribution. The waist circumference was made at the narrowest part of the torso between the ziphoid process and the umbilicus, and the hip circumference at the point of maximal circumference of the buttocks above the gluteal fold. Both measures were made with a spring-retractable steel tape measure.
Insulin sensitivity. An estimate of SI was calculated using the minimal model analysis of the FSIVGTT (SI-MM) as described by Bergman et al. (12). For the baseline assessment, subjects were instructed to keep physical activity to a minimum on the day preceding the test. For the 6-month assessment, subjects were studied within 24 h of their last training bout. For both tests, subjects arrived at the General Clinical Research Center in the morning following a 10-h fast. A Teflon catheter was placed in the antecubital space of each arm for blood sampling and for glucose infusion. Baseline blood samples were obtained following a 10-min rest (after placement of the catheters) for fasting insulin and glucose concentrations. A dextrose-saline solution was infused (0.5 g dextrose/kg body weight) over 3 min, and 14 blood samples were obtained over the subsequent 3 h at the following time points: 3, 6, 9, 12, 15, 20, 25, 30, 40, 50, 65, 80, 120, and 180 min. Approximately 2 ml whole blood was collected into tubes containing sodium fluoride and potassium oxalate, gently mixed, and placed on ice. These samples were measured for glucose in duplicate on an automated analyzer (YSI, Inc., Yellow Springs, Ohio), and the average value was used for subsequent data analysis. Approximately 10 ml whole blood was collected into serum separator tubes and allowed to clot, and the serum was separated and kept frozen at -70 C for subsequent analysis of insulin, using a double-antibody, competitive RIA technique (13).
Insulin sensitivity was also estimated from the physiological steady-state values of insulin and glucose collected before the glucose infusion, using the QUICKI method (6).
Aerobic capacity. Both before and after exercise training
(see below), graded treadmill exercise (Bruce protocol) was performed
to volitional fatigue to measure maximal oxygen consumption
(VO2max) and maximal heart rate. Subjects were
required to meet two of four standard criteria for having achieved
VO2max (plateau in VO2,
heart rate 
age predicted maximum heart rate, respiratory
exchange ratio 1.10, rating of perceived exertion 19).
Pulmonary gas exchange was measured continuously using a metabolic cart
(TrueMax 2400, ParvoMedics, Inc., Sandy, UT). Pulmonary
ventilation was measured via pneumotach, which was calibrated daily,
and fractions of O2 and CO2
via analyzers calibrated with gases of known concentration before each
test. Maximal heart rate was measured via continuous, standard 12-lead
electrocardiography. Resting heart rate was the average of three seated
measurements performed on two different days.
Exercise training. Following baseline assessments, subjects were randomized to one of three groups that differed with respect to the frequency and/or intensity of aerobic training. Because we were interested in assessing the overall utility of QUICKI to reflect changes in SI owing to exercise training, and because many different types of training interventions can be used, we did not analyze changes in SI separately by group. Thus, the data presented reflect pooled subjects from all three groups. In brief, the training program consisted of walking exercise prescribed at a frequency of either 3–4 (low frequency) or 5–7 (high frequency) d/wk, an intensity of either 45–55% (moderate intensity) or 65–75% (high intensity) of individual heart rate reserve (heart rate reserve = maximal heart rate - resting heart rate), and a duration of 30 min. The three training groups were defined as the following: high intensity, high frequency (1); high intensity, low frequency (2); and moderate intensity, high frequency (3). Subjects wore a heart rate monitor (Polar Electro, Woodbury, NY) during each training session to gauge exercise intensity. Subject adherence and compliance with the exercise prescription was ensured through continuous communication with the principal investigator (telephone and e-mail) and by review of subject-completed training logs. All subjects achieved a level of at least 85% of the prescribed exercise.
Statistical analysis. Standard descriptive statistics were used to summarize demographic and clinical variables. We used the Spearman rank-correlation to examine the correlation between SI-MM, QUICKI, and fasting insulin, and between pre- and post-training values for SI-MM, QUICKI, and fasting insulin. The choice of the rank-correlation coefficient over the standard Pearson correlation was dictated by the necessity to examine all possible monotone relationships between the variables (i.e. we did not want to restrict our examination to purely linear relationships). Changes in SI-MM, QUICKI, and fasting insulin were tested using the Wilcoxon signed-rank test. All tests were two-sided and carried out at an alpha level of 0.05.
Results
Subjects consisted of 10 female and 5 male sedentary adults
(52.9 ± 5.4 yr). As expected, exercise training resulted in a
significant (P = 0.003) increase in
SI-MM from 2.57 ± 2.65
min-1·µU·ml before training to 4.43
± 3.41 min-1·µU·ml after training.
However, there was no change in QUICKI (0.3509 ± 0.0297
vs. 0.3633 ± 0.0382, P = 0.121),
fasting insulin (9.87 ± 4.72 vs. 8.13 ± 5.13
uU/ml, P = 0.147), VO2max
(24.81 ± 5.42 vs. 25.07 ± 4.96 ml
O2·kg-1·min-1,
P = 0.720), BMI (29.19 ± 4.95 vs.
29.26 ± 4.97 kg/m2, P =
0.820), or W:H ratio (0.83 ± 0.10 vs. 0.83 ±
0.10, P = 0.368) with exercise training. In contrast,
there was a significant change in fasting glucose (84.5 ± 6.8
vs. 88.6 ± 9.8 mg/dl, P = 0.021) over
the course of the 6-month exercise training program. Pre- and
post-training values are summarized in Table 1
.
|
).
However, the rank-correlation coefficient between fasting insulin alone
with SI-MM was as good as that between QUICKI and
SI-MM both before (r = -0.77,
P = 0.0009) and after (r = -0.55,
P = 0.03) exercise training (Fig. 2). Fasting glucose was not related to
SI-MM at either time point.
|
|
), and between
SI-MM and fasting insulin were then examined.
Neither the difference scores for QUICKI nor fasting insulin correlated
with the difference scores for SI-MM (QUICKI
vs. SI-MM r = 0.24,
P = 0.39 and fasting insulin vs.
SI-MM r = -0.40, P =
0.14).
|
We used two different methods to examine changes in SI with exercise training, a perturbation known to improve SI, in a group of sedentary adults. One approach involved a dynamic intervention (SI-MM) that required obtaining multiple blood samples over a 3-h time period after an iv glucose injection. The other method (QUICKI) involved a mathematical expression of fasting insulin and glucose concentrations. Obviously, QUICKI is superior in terms of time to completion, expense, and independence from specialized equipment and/or techniques. More importantly, however, QUICKI has the potential to determine changes in SI in large groups of individuals who have undergone a specific metabolic perturbation (e.g. large-scale exercise training or pharmacological studies) in whom it would be impractical to perform glucose-insulin clamps or minimal model FSIVGTTs to estimate SI.
When analyzing the baseline data, we found that QUICKI correlated well
(r = 0.79) with SI-MM (Fig. 1A
). This
relationship was stronger than the study recently reported by Katz
et al. (6) in which the overall correlation
between these two indexes was r = 0.52. This discrepancy is likely
owing to differences in subject characteristics. Katz et al.
(6) enrolled nonobese (31 ± 2 yr, 24.2 ± 0.5
kg/m2), obese (41 ± 3 yr, 38.6 ± 1.7
kg/m2), and diabetic (46 ± 2 yr, 35.1
± 2.9 kg/m2) subjects, whereas our subjects were
older (52.9 ± 5.4 yr) and less obese (29.19 ± 4.95
kg/m2), and all had normal glucose tolerance
(i.e. fasting glucose < 109 mg/dl). Individual group
correlations between QUICKI and SI-MM in the Katz
study were r = 0.36 for the nonobese, r = 0.75 for the obese,
and r = 0.67 for the diabetic subjects (6). Thus, the
marginal relationship between QUICKI and SI-MM in
their nonobese, insulin-sensitive subjects attenuated the overall
correlation between these two variables for all subjects combined.
Despite these subgroup differences, the authors concluded that QUICKI
could accurately assess insulin sensitivity in vivo over a
wide range in a diverse population (e.g. nonobese, obese,
and diabetic subjects) (6).
The strong association between QUICKI and SI-MM
found in our study notwithstanding, the correlation between fasting
insulin and SI-MM (r = -0.77) at baseline
(Fig. 2A) was as good as that between QUICKI and
SI-MM (Fig. 1A). Fasting glucose was not related
to SI-MM at baseline. Together, these findings
demonstrate that fasting insulin alone is as good at estimating basal
SI as is combining fasting insulin and glucose
(i.e. QUICKI), at least in nondiabetic adults.
The correlation between QUICKI and SI-MM was
still significant after the 6-month exercise training intervention
(Fig. 1B), however, was less robust (r = 0.56) than that found
between these variables at baseline (Fig. 1A). Similar to the situation
at baseline, the correlation between fasting insulin and
SI-MM after training (r = -0.55) (Fig. 2B)
was as good as that between QUICKI and SI-MM
(Fig. 1B). Fasting glucose was not related to
SI-MM after exercise training. These findings
again demonstrate that fasting insulin alone is as good at estimating
SI, this time determined after exercise training,
as is combining fasting insulin and glucose (i.e. QUICKI) in
nondiabetic adults.
In our final analysis, we calculated the difference between pre- and
post-training values for SI-MM, QUICKI and
fasting insulin and then examined the correlations among the difference
scores for these variables. By doing so, we found no relationship
between either the change in SI-MM vs.
the change in QUICKI (r = 0.24) (Fig. 3), or the change in
SI-MM vs. the change in fasting
insulin (r = -0.40). These results demonstrate that neither
changes in QUICKI nor fasting insulin are related to changes in
SI-MM after exercise training.
Fasting insulin per se provides a reasonable and simple estimate of basal SI and appears to predict the development of diseases associated with insulin resistance (14, 15). Despite this apparent utility, the use of fasting insulin alone suffers from a number of interpretive problems (16). The primary concern is that measurement of the basal insulin concentration reflects SI in the basal, postabsorptive state (16). Because the majority of glucose uptake following an overnight fast occurs in insulin-independent tissues (e.g. brain and splanchnic tissues, such as liver and gut) (17), the fasting insulin concentration alone does not reflect insulin action in insulin-dependent tissues (e.g. muscle) (16). Defects in insulin-stimulated glucose transport and/or phosphorylation activity represent the primary mechanisms of insulin resistance in obesity and genetic models (i.e. offspring of patients with type 2 diabetes) of diabetes (18, 19). Therefore, the method by which whole-body SI is determined should reflect these processes. Because the QUICKI derived estimate of SI depends largely on the fasting insulin concentration, this method is unable to provide an adequate assessment of changes in whole-body SI with exercise training. Dynamic interventions, such as the glucose-insulin clamp or MM analysis of the FSIVGTT, are thus the methods of choice under these circumstances.
In conclusion, our results demonstrate that fasting insulin alone is as good at estimating basal SI as is combining fasting insulin and glucose (i.e. QUICKI) in nondiabetic adults. However, neither QUICKI nor fasting insulin alone accurately reflects exercise-induced changes in SI within individual subjects when changes in SI with training are determined during MM analysis of the FSIVGTT. We therefore advise against the use of QUICKI as an estimate of SI when a change in this parameter owing to a specific metabolic perturbation, such as exercise training, is the primary outcome measure of the study.
Acknowledgments
We acknowledge the GCRC at the University of Colorado Health Sciences Center and, in particular, Teddi Wiest-Kent for performing the insulin assays.
Footnotes
This work was supported by an American Heart Association/ Florida-Puerto Rico Affiliate Postdoctoral Fellowship (9920174V), an American College of Sports Medicine Foundation/Polar Research grant, and General Clinical Research Center Grant RR00082.
Abbreviations: BMI, Body mass index; FSIVGTT, frequently sampled iv glucose tolerance test; HOMA, homeostasis model assessment; IR, insulin resistance; MM, minimal model; SI, insulin sensitivity; SI-MM, MM estimate of SI; VO2max, maximal oxygen consumption; W:H, waist to hip ratio.
Received March 2, 2001.
Accepted May 15, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  H. Chen, G. Sullivan, and M. J. Quon Assessing the Predictive Accuracy of QUICKI as a Surrogate Index for Insulin Sensitivity Using a Calibration Model Diabetes, July 1, 2005; 54(7): 1914 - 1925. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Tripathy, P. Almgren, T. Tuomi, and L. Groop Contribution of Insulin-Stimulated Glucose Uptake and Basal Hepatic Insulin Sensitivity to Surrogate Measures of Insulin Sensitivity Diabetes Care, September 1, 2004; 27(9): 2204 - 2210. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Z. Lewanczuk, B. W. Paty, and E. L. Toth Comparison of the [13C]Glucose Breath Test to the Hyperinsulinemic-Euglycemic Clamp When Determining Insulin Resistance Diabetes Care, February 1, 2004; 27(2): 441 - 447. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Sheeder, S. H. Travers, and C. Stevens-Simon Is This Patient Insulin Resistant? How Much Does It Matter? Clinical Pediatrics, November 1, 2003; 42(9): 835 - 839. [PDF]
|
|
|
|
 |
|
 R. Rabasa-Lhoret, J.-P. Bastard, V. Jan, P.-H. Ducluzeau, F. Andreelli, F. Guebre, J. Bruzeau, C. Louche-Pellissier, C. MaItrepierre, J. Peyrat, et al. Modified Quantitative Insulin Sensitivity Check Index Is Better Correlated to Hyperinsulinemic Glucose Clamp than Other Fasting-Based Index of Insulin Sensitivity in Different Insulin-Resistant States J. Clin. Endocrinol. Metab., October 1, 2003; 88(10): 4917 - 4923. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Yokoyama, M. Emoto, S. Fujiwara, K. Motoyama, T. Morioka, M. Komatsu, H. Tahara, T. Shoji, Y. Okuno, and Y. Nishizawa Quantitative Insulin Sensitivity Check Index and the Reciprocal Index of Homeostasis Model Assessment in Normal Range Weight and Moderately Obese Type 2 Diabetic Patients Diabetes Care, August 1, 2003; 26(8): 2426 - 2432. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. N. Bergman, D. J. Zaccaro, R. M. Watanabe, S. M. Haffner, M. F. Saad, J. M. Norris, L. E. Wagenknecht, J. E. Hokanson, J. I. Rotter, and S. S. Rich Minimal Model-Based Insulin Sensitivity Has Greater Heritability and a Different Genetic Basis Than Homeostasis Model Assessment or Fasting Insulin Diabetes, August 1, 2003; 52(8): 2168 - 2174. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Chen, G. Sullivan, L. Q. Yue, A. Katz, and M. J. Quon QUICKI is a useful index of insulin sensitivity in subjects with hypertension Am J Physiol Endocrinol Metab, April 1, 2003; 284(4): E804 - E812. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. I. Uwaifo, E. M. Fallon, J. Chin, J. Elberg, S. J. Parikh, and J. A. Yanovski Indices of Insulin Action, Disposal, and Secretion Derived From Fasting Samples and Clamps in Normal Glucose-Tolerant Black and White Children Diabetes Care, November 1, 2002; 25(11): 2081 - 2087. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Katsuki, Y. Sumida, E. C. Gabazza, S. Murashima, H. Urakawa, K. Morioka, N. Kitagawa, T. Tanaka, R. Araki-Sasaki, Y. Hori, et al. QUICKI Is Useful for Following Improvements in Insulin Sensitivity after Therapy in Patients with Type 2 Diabetes Mellitus J. Clin. Endocrinol. Metab., June 1, 2002; 87(6): 2906 - 2908. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Quon QUICKI Is a Useful and Accurate Index of Insulin Sensitivity J. Clin. Endocrinol. Metab., February 1, 2002; 87(2): 949 - 950. [Full Text]
|
|
|
|
|
|
G. E. Duncan, A. D. Hutson, and P. W. Stacpoole QUICKI Is Not a Useful and Accurate Index of Insulin Sensitivity following Exercise Training J. Clin. Endocrinol. Metab., February 1, 2002; 87(2): 950 - 951. [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
