Paper | Published:

Revised QUICKI provides a strong surrogate estimate of insulin sensitivity when compared with the minimal model

International Journal of Obesity volume 28, pages 222227 (2004) | Download Citation



OBJECTIVE: To compare insulin sensitivity (Si) from a frequently sampled intravenous glucose tolerance test (FSIGT) and subsequent minimal model analyses with surrogate measures of insulin sensitivity and resistance and to compare features of the metabolic syndrome between Caucasians and Indian Asians living in the UK.

SUBJECTS: In all, 27 healthy male volunteers (14 UK Caucasians and 13 UK Indian Asians), with a mean age of 51.2±1.5 y, BMI of 25.8±0.6 kg/m2 and Si of 2.85±0.37.

MEASUREMENTS: Si was determined from an FSIGT with subsequent minimal model analysis. The concentrations of insulin, glucose and nonesterified fatty acids (NEFA) were analysed in fasting plasma and used to calculate surrogate measure of insulin sensitivity (quantitative insulin sensitivity check index (QUICKI), revised QUICKI) and resistance (homeostasis for insulin resistance (HOMA IR), fasting insulin resistance index (FIRI), Bennetts index, fasting insulin, insulin-to-glucose ratio). Plasma concentrations of triacylglycerol (TAG), total cholesterol, high density cholesterol, (HDL-C) and low density cholesterol, (LDL-C) were also measured in the fasted state. Anthropometric measurements were conducted to determine body-fat distribution.

RESULTS: Correlation analysis identified the strongest relationship between Si and the revised QUICKI (r=0.67; P=0.000). Significant associations were also observed between Si and QUICKI (r=0.51; P=0.007), HOMA IR (r=−0.50; P=0.009), FIRI and fasting insulin. The Indian Asian group had lower HDL-C (P=0.001), a higher waist–hip ratio (P=0.01) and were significantly less insulin sensitive (Si) than the Caucasian group (P=0.02).

CONCLUSION: The revised QUICKI demonstrated a statistically strong relationship with the minimal model. However, it was unable to differentiate between insulin-sensitive and -resistant groups in this study. Future larger studies in population groups with varying degrees of insulin sensitivity are recommended to investigate the general applicability of the revised QUICKI surrogate technique.


Insulin resistance is an independent risk factor in the aetiology of cardiovascular disease (CVD) and noninsulin-dependent diabetes mellitus (NIDDM). 1 It is also reported to be associated with obesity and dyslipidaemia2 contributing to the cluster of abnormalities known as the metabolic syndrome.3 Features of the metabolic syndrome are commonly reported among Indian Asians living in the UK4,5,6,7 and are thought to contribute to the increased mortality rate observed among this ethnic group compared to the native Caucasian population.8,9 Reliable measurement of insulin resistance is therefore important to enable further investigation into the links between insulin resistance and other features of the metabolic syndrome, as well as to evaluate effective preventative strategies for the metabolic syndrome.

The most valid of the currently available procedures for measuring insulin sensitivity in vivo involves dynamic intervention with glucose and insulin to mimic normal body homeostasis. Although the glucose-clamp technique10 is considered the gold standard, the more recently developed minimal model provides a well-validated alternative.11,12 Both procedures are time consuming and invasive rendering them unsuitable for use when determining the insulin-sensitivity status of large study populations. Recent research has focused on the development of simpler, less invasive surrogate measures of insulin resistance and sensitivity. Most of these are based on measurement of fasting concentrations of insulin and glucose and can be easily applied in large-scale studies or where it is not practical to conduct more comprehensive tests.

The homeostasis model for insulin resistance (HOMA IR)13 is the most commonly employed surrogate measure and provides a reliable alternative to the more robust glucose clamp14 and minimal model techniques.15 Other surrogate methods, including fasting insulin,16 Bennetts index17 and the insulin-to-glucose ratio,18 also demonstrate good concordance with values obtained using the more invasive techniques.19,20,21 The recently developed quantitative insulin-sensitivity check index (QUICKI),22 which takes the product of both the reciprocal and the logarithm of fasted insulin and glucose concentrations, gives values that correlate well with the glucose-clamp and minimal model techniques and may provide a better surrogate measure of insulin resistance compared to HOMA IR.22 The revised version of QUICKI (revised QUICKI)23 that incorporates a value for fasting nonesterified fatty acids (NEFA) into the QUICKI formula is reported to have a stronger relationship with the glucose clamp than QUICKI alone and therefore may provide an adequate substitute for more sophisticated measures of insulin sensitivity. However, as yet the revised QUICKI has not been widely employed in studies to date and its further validation in different populations is required.

The aim of this study was to compare a range of surrogate measures of insulin sensitivity (QUICKI and revised QUICKI) and resistance (HOMA IR, fasting insulin resistance index (FIRI), Bennetts index, fasting insulin, insulin-to-glucose ratio) with insulin sensitivity (Si) derived from the minimal model technique, in a group of male volunteers of Caucasian and Indian Asian ethnic origin. This study also aimed to compare a range of metabolic parameters including insulin sensitivity between the two ethnic groups.

Subjects and methods


In all, 27 healthy male volunteers (13 Indian Asians, Sikhs were targeted as the Indian Asian group, and 14 Caucasians), aged 35–70 y were recruited from the Reading and Slough areas of the UK, through a combination of e-mail contact, newspaper advertisements, distributed flyers and word of mouth. Healthy individuals were recruited on the basis of a screening blood sample (glucose <8 mmol/l, triacylglycerol (TAG) 0.5–4 mmol/l, total cholesterol <8 mmol/l) and completion of a medical and lifestyle questionnaire. The exclusion criteria included diagnosed CVD, diabetes, liver disease, smoking, hypertension, a BMI >35 kg/m2 and those who were on a weight-reducing or other diet. Those who took part in strenuous exercise exceeding 20 min, more than three times per week were not permitted to take part in the study. Additionally, those taking hypolipidaemic therapy or other medication known to affect lipid or glucose metabolism and those consuming fatty acid or other supplements were also excluded. Indian Asian volunteers who have been resident in the UK for a minimum of 2 y were required. Volunteer characteristics are presented in Table 1. Ethical approval for this study was given from the University of Reading, West Berkshire, East Berkshire and Hounslow Research Ethics Committees, and all volunteers gave their written informed consent for participation.

Table 1: Baseline characteristics for the study group according to ethnicity

Insulin action measurements

Insulin sensitivity was assessed from a frequently sampled intravenous glucose-tolerance test (FSIGT)24 with minimal model computer analyses25 subsequent to a 12-h overnight fast. Volunteers were requested to refrain from exercise and alcohol on the day prior to their insulin-sensitivity test and to consume a low-fat meal the evening before their visit. On the morning of the FSIGT, a cannula was inserted into the antecubital vein of both forearms under local anaesthetic. A fasting blood sample was taken to assess insulin, glucose, TAG, NEFA and cholesterol concentrations. At the start of the test, a bolus of 50% glucose solution (Phoenix Pharma Ltd, Gloucester, UK) (0.3 g/kg body weight (BW)) was infused into one of the cannula over a 1 min period. After 20 min, a dose of insulin (Novo Nordisk Pharmaceuticals Ltd, West Sussex, UK) (0.03 U/kg BW) was infused into the same cannula. Meanwhile, blood was sampled through the other cannula at frequent intervals over a 3-h period (−5, 0, 2, 4, 8, 19, 22, 30, 40, 50, 70, 100 and 180 min after the start of the glucose injection).26 Glucose and insulin values were entered into the MINMOD computer program (version 3.0, Richard N Bergman) to determine Si using mathematical modelling methods.11,25 Insulin sensitivity from the minimal model was taken as a gold standard and compared with a selection of surrogate measures of insulin resistance (HOMA IR, FIRI, Bennetts index, fasting insulin, insulin-to-glucose ratio) and insulin sensitivity (QUICKI and Revised QUICKI), which were calculated using fasting insulin, glucose and NEFA concentrations calculated from the mean of fasted blood samples collected at −5 and 0 min of the FSIGT. The calculation of these surrogate measures is summarised in Table 2.

Table 2: Surrogate measures of insulin action

Anthropometric measurements

Anthropometric measurements were conducted by a single trained investigator on the morning of the insulin-sensitivity test. Height was measured to the nearest 1 cm using an upright stadiometer and weight to the nearest 0.5 kg on standard medical scales. Waist circumferences were determined mid-way between the lowest rib margin and the iliac crest to the nearest 1 mm in a standing position. Hip circumferences were determined as the largest circumference around the hips to the nearest 1 mm in the standing position. Skinfold thickness was measured using Harpenden's calipers at the triceps, biceps, subscapular and suprailiac sites as previously described.27

Biochemical analysis

Blood samples were collected into 5 ml potassium EDTA tubes and 1 ml fluoride oxalate tubes. Blood samples were centrifuged at 3000 rpm for 10 min and the plasma obtained was aliquoted into 3 ml flat bottomed plastic tubes (Thermo Lifesciences, Basingstoke, UK) and stored at –20°C for later determination of plasma TAG, NEFA, insulin, high-density lipoprotein cholesterol (HDL-C) and total cholesterol concentrations from the EDTA tubes and glucose concentration from the fluoride oxalate tubes. Low density lipoprotein cholesterol (LDL-C) was determined by the Friedewald formula.28

Plasma TAG, glucose and total and HDL-C concentrations (using test kits supplied by Instrumentation Laboratories Ltd, Warrington, Ches., UK) and NEFA (using the Wako NEFA C test kit; Alpha Laboratories Ltd, Eastleigh, Hants., UK) were determined using an automated analyzer (Instrumentation Laboratories UK Ltd). Insulin was measured using a specific commercial ELISA kit (DAKO Ltd, High Wycombe, Bucks., UK). The mean intra- and inter-CVs for the total cholesterol, glucose, TAG, NEFA and insulin were 2.1, 1.0, 1.4, 1.8 and 4.0%, respectively, and 4.0, 3.7, 3.1, 1.8 and 5.5%, respectively.

Statistical analysis

All statistical analyses were performed using SPSS (version 11.0;. SPSS, Chicago, IL, USA) and a P value of <0.05 was considered to be statistically significant. Prior to statistical analysis, all data were examined for normality using the Shapiro–Wilks test and log transformed where necessary. Comparison between the measurement of Si derived from the minimal model and the surrogate estimates of insulin action was conducted using Pearson's correlation analyses. Significant differences between the Caucasian and Indian Asian groups were determined by independent t-tests. The results are presented as means±s.e.m.


Minimal model vs surrogate measures

The results comparing surrogate measures of insulin sensitivity (QUICKI and revised QUICKI) and resistance (HOMA IR, FIRI, Bennetts index, fasting insulin, insulin-to-glucose ratio) with the minimal model estimate of Si are presented in Table 3. Significant positive Pearson's correlations were observed between Si and revised QUICKI and QUICKI and significant negative Pearson's correlations between Si and HOMA IR, FIRI and fasting insulin measurements. The strongest and most significant relationship with Si was the positive association observed with insulin sensitivity measured by revised QUICKI (r=0.67; P=0.000), which differs from the other surrogate methods investigated as it incorporates fasting NEFA concentrations into an equation in combination with fasting insulin and glucose concentrations. A scatterplot of the relationship between revised QUICKI and Si is illustrated in Figure 1a. The most commonly used surrogate measure for insulin resistance, HOMA IR, showed a significant negative relationship with Si (r=−0.50; P=0.009), as illustrated in Figure 1b. Weaker nonsignificant negative relationships were observed between Si and the insulin-to-glucose ratio and Bennetts index.

Table 3: Pearson's correlation analysis comparing insulin sensitivity (Si) derived from the minimal model with surrogate measures of insulin action
Figure 1
Figure 1

Scatterplots of (a) Si vs revised QUICKI, (b) Si vs HOMA IR.

Ethnic differences in metabolic parameters

The blood lipid profile, insulin action measurements and anthropometric measurements for the total study group based on ethnicity, are presented in Table 1. The Indian Asian group had a significantly lower HDL-C concentration compared to the Caucasians (P=0.001). Additionally, this ethnic group had higher TAG and NEFA concentrations, although differences between the groups for these measurements did not reach significance. The Indian Asian group had a higher waist–hip ratio than the Caucasians (P=0.01), although similar waist circumferences were observed in the two groups. The Indian Asians had a greater amount of fat over the subscapular region (P=0.03) as determined by higher subscapular skinfold thickness compared to the Caucasian group.

When insulin sensitivity was assessed using the minimal model the Indian Asian group was identified as significantly less insulin sensitive than the Caucasians. However, differences in values did not reach significance for any of the surrogate measurements of insulin sensitivity (revised QUICKI, QUICKI) or insulin resistance (HOMA IR, FIRI, Bennetts index, fasting insulin, insulin-to-glucose ratio) applied in this study.


The purpose of the present study was to compare a range of surrogate markers of insulin sensitivity (QUICKI, revised QUICKI) and resistance (HOMA IR, FIRI, Bennetts index, fasting insulin, insulin-to-glucose ratio) with Si determined from the minimal model. This study also compared features of the metabolic syndrome between Indian Asian and Caucasian study participants. The minimal model11 was chosen as the gold standard due to its previously reported significant relationship with the glucose-clamp technique12 and its greater ease of use in volunteer studies. The results showed significant positive relationships between surrogate measures of insulin sensitivity (QUICKI, revised QUICKI) and significant negative relationships between measures of insulin resistance (HOMA IR, FIRI, fasting insulin) and Si. The strongest relationship emerged between Si and revised QUICKI, the most recently developed surrogate technique, which includes fasted NEFA concentrations in an equation with fasted insulin and glucose measures. The Indian Asian group exhibited some of the classical features of the metabolic syndrome including lower HDL-C, higher WHR and higher truncal fat. However, the minimal model was the only technique that discriminated between the two population groups with respect to insulin sensitivity.

The HOMA IR is the most commonly applied surrogate marker of insulin resistance, although few studies in the literature have investigated its relationship with Si. However, it has been concluded that HOMA IR provides the best surrogate measure in situations where direct measures cannot be implemented.29 The present study demonstrated a moderate relationship between HOMA IR and Si similar to previous results.15 Previously, fasting insulin has been reported to be as accurate as HOMA IR, Bennetts index and the insulin-to-glucose ratio, for prediction of insulin resistance in normoglycaemic subjects.19,20,21 In the present study, correlations between Si and the insulin-to-glucose ratio and Bennetts index were weaker than other surrogate markers and did not reach significance. However, fasting insulin did show a significant moderate relationship with Si and may be useful in situations where fasting glucose concentrations are unavailable.

The present study also evaluates two recently developed surrogate measures of insulin sensitivity, QUICKI and revised QUICKI. The QUICKI is based on the logarithm and the reciprocal of the product of fasting insulin and glucose concentrations and demonstrates a stronger relationship with the glucose clamp than HOMA IR.22 The QUICKI technique has been reported to have a significant relationship with the minimal model (r=0.52, P=0.000) over a wide range of insulin sensitivities (obese, nonobese and type II diabetic patients).22 A similar relationship was observed in the present study investigating normoglycaemic individuals (r=0.52, P=0.007). In an attempt to further improve the power of QUICKI to determine insulin sensitivity, fasting NEFA concentrations were incorporated into the equation with insulin and glucose.23 NEFA is a postabsorptive metabolic marker of insulin action that remains elevated in insulin resistance due to the failure of hormone sensitive lipase to respond to antilipolytic actions of insulin or possibly to impaired fatty acid esterification.30 Its measurement therefore reflects the effects of insulin beyond those of stimulation of glucose transport and reflects the disturbance in fasting NEFA commonly observed in insulin-resistant subjects.31 Inclusion of NEFA in the equation was shown to strengthen the relationship between QUICKI and the glucose clamp with the regression coefficient increasing from r=0.27 to 0.51.23 This is similar to the findings in the present study, where the strength of the relationship between QUICKI and Si increased from r=0.51 to r=0.67 on inclusion of NEFA. Therefore, inclusion of fasted NEFA concentrations as an additional marker of insulin resistance appears to improve the predictive power of fasting insulin and glucose alone.

Indian Asians in the UK have a high prevalence of the characteristics of the metabolic syndrome (insulin resistance,4 central obesity 5,6,7 and dyslipidaemia32,33), compared to Caucasians. Consistent with previous reports, the Indian Asians in the present study were significantly less insulin sensitive as determined by the minimal model, had significantly lower HDL-C concentrations and a significantly higher mean waist–hip ratio compared to the Caucasian group, despite having comparable plasma glucose concentrations. Absolute values for insulin sensitivity in the present study were comparable with a previous study that reported greater insulin sensitivity, from the minimal model, among Caucasians compared to Indian Asians (Si=3.01 and 1.92 × 10−4.min−1.μ−1, respectively).4 A previous larger study in our research group observed that Indian Asians (n=55) were significantly less insulin sensitive than Caucasians (n=55) when revised QUICKI was applied as a surrogate measure of insulin sensitivity.34 In the present study, it was observed that Indian Asians were less insulin sensitive using the revised QUICKI, than Caucasians, although these values did not reach statistical significance, which probably reflected the small numbers in each study group. Furthermore, the minimal model technique was the only method that discriminated, statistically, between the two normoglycaemic ethnic groups, illustrating greater sensitivity and value of this technique when relatively small numbers of volunteers are studied.

In conclusion, the present study established that the revised QUICKI surrogate index of insulin sensitivity had the strongest relationship with the gold standard minimal model in comparison to a range of other techniques for assessing insulin sensitivity (QUICKI) and resistance (HOMA IR, FIRI, Bennetts index, fasting insulin, insulin-to-glucose ratio). This method may represent a reliable surrogate measure of insulin sensitivity in large-scale studies where fasting insulin, glucose and NEFA concentrations are available. However, further research in larger study groups and in population groups with various ranges of insulin sensitivity, including obese and diabetic individuals, is necessary to determine whether this technique is of value. Finally, this study highlighted the sensitivity of the minimal model to discriminate between insulin-resistant and -sensitive population groups in small-scale studies.


  1. 1.

    , . The assessment of insulin resistance in man. Diabet Med 2002; 19: 527–534.

  2. 2.

    . Do non insulin dependant diabetes mellitus and cardiovascular disease share common antecedents? Annal Intern Med 1996; 124: 110–116.

  3. 3.

    . Role of insulin resistance in human disease. Diabetes 1988; 37: 1595–1607.

  4. 4.

    , , , , , . Relation of plasma lipids to insulin resistance, nonesterified fatty acid levels and body fat in men from three different ethnic groups: relevance to variation in risk of diabetes and coronary heart disease. Metabolism 2000; 49: 245–252.

  5. 5.

    , , , , . Body composition, visceral fat, leptin and insulin resistance in Asian Indian men. J Clin Endocrinol Metab 1998; 84: 137–144.

  6. 6.

    , , , , , , , . C-reactive protein, insulin resistance, central obesity and coronary heart disease risk in Indian Asians from the United Kingdom compared with European Whites. Circulation 2001; 104: 145–150.

  7. 7.

    , , , . Increased prevalence of smaller and denser LDL particles in Asian Indians. Arterioscler Thromb Vasc Biol 1999; 19: 2749–2755.

  8. 8.

    British Heart Foundation. Coronary heart disease statistics. British Heart Foundation: London; 2002.

  9. 9.

    , , . Coronary heart disease in South Asians overseas: A review. J Clin Epidemiol 1989; 42: 597–609.

  10. 10.

    , , . Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol 1979; 237: 214–223.

  11. 11.

    . Toward physiological understanding of glucose tolerance. Minimal-model approach. Diabetes 1989; 38: 1512–1527.

  12. 12.

    , , , , , , , , , . A comparison between the minimal model and the glucose clamp in the assessment of insulin sensitivity across the spectrum of glucose tolerance. Diabetes 1994; 43: 1114–1121.

  13. 13.

    , , , , . Homeostasis model assessment: Insulin resistance and β-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985; 28: 412–419.

  14. 14.

    . Homeostasis model assessment closely mirrors the glucose clamp technique in the assessment of insulin sensitivity. Diabetes Care 2000; 23: 57–63.

  15. 15.

    , , , , , , , . Assessment of insulin sensitivity from a single sample. Diabetes Care 2000; 23: 1434–1435.

  16. 16.

    , , , , , , , . Relation of carotid wall thickness to diabetes mellitus, fasting glucose and insulin, body size and physical activity. Atherosclerosis Risk in Communities (ARIC) Study Investigators. Stroke 1994; 25: 66–73.

  17. 17.

    , , , , , , , . Exploration of simple insulin sensitivity measures derived from frequently sampled intravenous glucose tolerance (FSIGT) tests. Am J Epidemiol 1995; 142: 724–732.

  18. 18.

    . Insulin resistance in obese and nonobese man. J Clin Endocrinol Metab 1991; 73: 691–695.

  19. 19.

    , , , , , , . Diagnosing insulin resistance in the general population. Diabetes Care 2001; 24: 460–464.

  20. 20.

    , , , , . Clinical significance of the insulin resistance index as assessed by the homeostasis model assessment. Endocr J 2001; 48: 81–86.

  21. 21.

    . How good a marker is insulin level for insulin resistance. Am J Epidemiol 1993; 137: 959–964.

  22. 22.

    , , , , , , . Quantitative insulin sensitivity check index: a simple accurate method for assessing insulin sensitivity in humans. J Clin Endocrinol Metab 2000; 85: 2402–2410.

  23. 23.

    , , , , . Incorporation of the fasting plasma FFA concentration into QUICKI improves its association with insulin sensitivity in nonobese individuals. J Clin Endocrinol Metab 2001; 86: 4776–4781.

  24. 24.

    , , . Assessment of insulin sensitivity in vivo. Endocr Rev 1985; 6: 45–86.

  25. 25.

    , . MINMOD: a computer program to calculate insulin sensitivity and pancreatic responsitivity from the frequently sampled intravenous glucose tolerance test. Comput Meth Prog Biomed 1986; 23: 113–122.

  26. 26.

    , , , , , , , , , , . Increased insulin resistance and insulin secretion in non diabetic African-Americans and Hispanics compared with non-Hispanic Whites: the insulin resistance atherosclerosis study. Diabetes 1996; 45: 742–748.

  27. 27.

    , . Body fat assessed from total body density and its estimation from skinfold thickness measurements on 481 mean and women aged 16 to 72 years. Br J Nutr 1974; 32: 77–97.

  28. 28.

    , . Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972; 18: 499–502.

  29. 29.

    , , . The homeostasis model in the San Antonio heart study. Diabetes Care 1997; 20: 1087–1092.

  30. 30.

    . Non esterified fatty acid metabolism and postprandial lipaemia. Atherosclerosis 1998; 141: S41–S46.

  31. 31.

    , , . Are increased plasma non esterified fatty acid concentrations a risk marker for coronary heart disease and other chronic diseases? Clin Sci 1996; 90: 243–253.

  32. 32.

    , , , , , , , , , , , , . Coronary risk factors on people from the Indian subcontinent living in west London and their siblings in India. Lancet 1995; 345: 405–409.

  33. 33.

    , , . Relation of central obesity and insulin resistance with high diabetes prevalence and cardiovascular risk in South Asians. Lancet 1991; 337: 382–386.

  34. 34.

    , , , , , . Lack of association between central adiposity and lipaemia in UK Sikh men. Int J Obesity 27: 1373–1382.

  35. 35.

    , , , , . Reduced adipocyte insulin sensitivity in Caucasian and Asian subjects with coronary heart disease. Diabet Med 1998; 15: 1003–1009.

Download references


We thank the Food Standards Agency (FSA) who provided funding for this research. We also thank Dr John Wright who cannulated and infused all the volunteers with glucose and insulin on the insulin sensitivity study days and Kangmei Ren who helped with the minimal model analysis. Finally, we thank the volunteers who gave up their time to participate in this study.

Author information


  1. School of Food Biosciences, The University of Reading, UK

    • L M Brady
    • , S S Lovegrove
    • , C M Williams
    •  & J A Lovegrove
  2. School of Health Related Professions, The University of Alabama at Birmingham, USA

    • B A Gower


  1. Search for L M Brady in:

  2. Search for B A Gower in:

  3. Search for S S Lovegrove in:

  4. Search for C M Williams in:

  5. Search for J A Lovegrove in:

Corresponding author

Correspondence to J A Lovegrove.

About this article

Publication history