Original Article | Published:

Relationships between glycaemic abnormalities, obesity and insulin resistance in nondiabetic Polynesians of New Caledonia

International Journal of Obesity volume 31, pages 109113 (2007) | Download Citation



Polynesians in New Caledonia have an increased risk for developing diabetes, compared to Melanesians or Europeans. They are also more prone to obesity. The aim of this study was to analyse differences in the pre-diabetic state that may explain the varying susceptibility to diabetes between these three ethnic groups, focusing on the balance between insulin resistance and capacity of pancreatic cells to secrete insulin.

Design and subjects:

The CALDIA Study is a population-based cross-sectional survey of diabetes prevalence conducted in New Caledonia. All participants who did not have diabetes, according to the results of a 0–2 h oral glucose tolerance test (n=392), were selected for analysis.


Compared to Europeans, Polynesians and Melanesians had significantly higher body mass indices (BMI) and waist-to-hip ratios (WHRs). Polynesians had higher fasting plasma glucose values than Europeans or Melanesians (6.03 mmol/l, vs 5.78 and 5.46, respectively; P<0.0001). Fasting plasma insulin level and the estimate of insulin resistance by homeostasis model assessment were not significantly different between the three ethnic groups. Homeostasis model assessment estimate of β-cell secretory capacity was lower in Polynesians compared to the two other ethnic groups (83.1 mU/mmol, vs 119.3 and 125.2, respectively; P<0.02).


Despite a high prevalence of central obesity, as judged by high BMI and WHR, in Polynesians of New Caledonia, their high risk of diabetes may be more strongly related to a defect in insulin secretion capacity than to insulin resistance.


Type 2 diabetes is considered to be the result of a two-step process.1 In the first ‘prediabetic’ step characterized by insulin resistance, compensatory hyperinsulinaemia helps maintain normal glucose levels. In the second step, there is a decline in β-cell secretory capacity together with a progressive elevation of glucose levels that define diabetes.2 This two-step mechanism has been shown to apply to many populations around the world (e.g. Pima Indians, Micronesians, Mexican Americans, South Asians).3 However, the balance between insulin resistance and deficient insulin secretion in the pathogenesis of type 2 diabetes appears to vary between ethnic groups.4 Regarding Polynesians, a population with a high prevalence of obesity and diabetes, one study in New Zealand reported that, after adjustment for body mass index (BMI), Polynesians were not more insulin-resistant than Europeans, who are much less susceptible to diabetes.5

Previous publications from the CALDIA Study, a large diabetes screening survey conducted in the multiethnic population of New Caledonia, have shown that Polynesians indeed had the highest degree of central obesity, and the highest prevalence rate of diabetes compared to the other two major ethnic groups living in the archipelago (15.3 vs 8.4% in Melanesians or Europeans).6, 7 In order to further examine the abnormalities that could explain their higher risk of diabetes, we selected from the CALDIA database nondiabetic subjects with the purpose of comparing the degree of insulin resistance and of β-cell secretory capacity (using the homeostasis model assessment, or HOMA) between ethnic groups.

Population and methods

The CALDIA Study was conducted from 1992 to 1994 to determine the prevalence of diabetes in New Caledonia. The rationale, design and methods of the study were previously described in detail.6 To summarize, the target population for the CALDIA Study were subjects aged 30–59 years, resident in New Caledonia for more than 10 years. Subjects were recruited all over the territory (North province, Noumea and Loyalty Islands). In the North province, 12 small towns and 101 villages out of 199 were randomly selected. In Noumea, six suburbs were chosen because they included all the ethnic groups. In the Loyalty Islands, all 85 villages participated. As a whole, 9390 subjects (representing a response rate of 78%) were visited at home for screening, where they all had a capillary blood glucose (CBG) measurement with a reflectance meter (One Touch®, LifeScan, Johnson & Johnson, Milpitas, CA, USA). All previously known diabetic subjects and subjects having a CBG value 6.1 mmol/l when fasting, or CBG 7.8 mmol/l when nonfasting (n=643), were invited to come to the health centre for a more detailed examination. The response rate for this examination was 91.5% (588 subjects). At the same time, a selection of 517 subjects with CBG <6.1 mmol/l, matched by ethnic group, gender, age and location, also underwent the examination.

At the health centre, participants answered a standardized questionnaire and underwent anthropometric measurements. BMI (weight (kg)/height (m)2) was used to report general obesity. According to the World Health Organization (WHO) criteria,8 subjects with BMI 25–29.9 Kg/m2 were considered overweight and subjects with BMI 30 Kg/m2 were classified as obese. Waist and hip circumferences were measured in the standing position to the nearest centimetre, the first at the umbilicus and the second at the iliac crest. The waist-to-hip ratio (WHR) was calculated as an index of upper-body adiposity.

A 2-hour oral glucose tolerance test (OGTT) with a 75-g glucose load was performed according to the WHO recommendations.9 Blood samples were locally centrifuged, frozen and sent to Noumea central laboratory. Fasting plasma glucose (FPG) and 2-h plasma glucose (2h-PG) levels were assessed by the glucose oxidase method in Noumea. Fasting plasma insulin (FPI) was measured by radioimmunoassay (RIAgnost, Behring, Los Angeles, CA, USA) at the Henri Mondor Hospital, Créteil, France.

The study protocol and procedures were approved by all local medical commissions involved, and all participants gave informed consent for their participation.

Among the 1105 subjects having a complete examination, we excluded diabetic subjects, known or newly diagnosed, that is those with FPG 7 mmol/l or 2h-PG 11.1 mmol/l at the OGTT.9 One unclassified subject was also excluded. Of the remaining subjects, only those of European, Melanesian or Polynesian origin were retained. Subjects were classified as normoglycaemic (FPG <6.1 mmol/l and 2h-PG <7.8 mmol/l), having impaired fasting glucose (IFG) (6.1 mmol/l FPG <7.0 mmol/l and 2h-PG <7.8 mmol/l), or impaired glucose tolerance (IGT) (7.8 mmol/l 2h-PG <11.1 mmol/l, irrespective of FPG values).9 To assess insulin resistance and β-cell function, the HOMA10 was applied to fasting plasma glucose and insulin values as follows: insulin resistance (HOMA-IR)=FPI (mU/l) × FPG (mmol/l)/22.5; β-cell function (HOMA-BC)=[20 × FPI (mU/l)]/[FPG (mmol/l)-3.5].

Age, sex, anthropometric variables and variables characterising glucose-insulin regulation were compared between the three ethnic groups using χ2 test for categorical variables and analysis of variance for continuous variables. Glucose homeostasis variables were also compared after multiple adjustment for age, sex and BMI, by analysis of covariance. When comparisons were statistically significant, two-by-two comparisons were carried out, using the Bonferroni correction. Continuous variables with a log-normal distribution were log-transformed before testing, and back-transformed into natural values for presentation. All statistical analyses were performed with SAS statistical package software version 9.1 (SAS Institute, Cary, NC, USA).


After exclusion of subjects with missing values for the variables of interest, analysis was performed on 392 subjects (57 Europeans, 287 Melanesians, 48 Polynesians). Their general characteristics are shown in Table 1. Age and sex distribution were not significantly different between the groups. Mean BMI and WHR were highest in Polynesians and lowest in Europeans. Overweight and obesity were significantly more prevalent (P<0.0001) in Polynesians or Melanesians than in Europeans.

Table 1: General characteristics of the participants, by ethnic group

Nonadjusted and age, sex and BMI-adjusted comparisons of variables characterising glucose-insulin regulation are shown in Table 2. Fasting plasma glucose values were significantly different between the three groups, with the highest mean value observed in Polynesians, who also had the highest prevalence rate of glycaemic anomalies (IFG or IGT: 43.7 vs 35.1% in Europeans and 22.6% in Melanesians, P<0.01). Non-adjusted 2-h plasma glucose was not significantly different between the three groups, but the differences became significant when adjusted for age, sex and BMI. Despite the marked differences in morphotype between the groups, the geometric means of fasting insulin and HOMA-IR were not significantly different. By contrast, there was a significant difference (P<0.02) for the marker of insulin secretion capacity, HOMA-BC, with the lowest value observed in Polynesians. When the analysis was performed after adjustment for BMI, the differences became significant for FPI (P<0.004) and HOMA-IR (P<0.002). The two-by-two comparisons with Bonferroni adjustment (Table 3) indicated that for these markers of insulin resistance, the differences were between Europeans and the two other ethnic groups, with Europeans having significantly higher indices of insulin resistance. By contrast, insulin secretory capacity was significantly lower in Polynesians compared to the two other ethnic groups.

Table 2: Markers of glucose-insulin regulation by ethnic group
Table 3: Two-by-two comparisons between the ethnic groups for markers of glucose-insulin regulation

To better understand this finding, the three ethnic groups were divided according to BMI (normal, overweight or obese). At any given body mass, Polynesians had the highest fasting glucose values together with the lowest fasting insulin values, denoting a clear lack of insulin, both absolute and relative to glucose values (Figure 1, panels a and b). This was confirmed by the fact that the curve for HOMA-BC in Polynesians was consistently and largely lower, at any given body mass, compared to the two other groups (Figure 1, panel c).

Figure 1
Figure 1

Levels of fasting plasma glucose (FPG, a), fasting plasma insulin (FPI, b) and HOMA estimate of β-cell function (HOMA-BC, c) by degree of overweight in nondiabetic Europeans, Melanesians and Polynesians in New Caledonia.


Polynesians are known to be a population at high risk for type 2 diabetes. This was confirmed in the CALDIA Study, where diabetes prevalence rate was twice that observed in Europeans or Melanesians.6 Moreover, in our analysis restricted to nondiabetic subjects, Polynesians were markedly overweight with upper-body obesity, and they already exhibited abnormalities of glucose homeostasis (higher FPG values and lower prevalence of normoglycaemic subjects), which are signs of susceptibility to diabetes.11 Although obesity is commonly associated with insulin resistance and hyperinsulinaemia, and the classical view is that these anomalies are markers of the prediabetic state,12, 13, 14 the only accompanying anomaly we found in Polynesians was a relative defect in insulin secretion. The fact that our radioimmunoassay was not specific for insulin does not challenge the finding, as, if there are differences of proinsulin proportion between the study groups, it is probably higher in Polynesians given their ‘worse’ glucose homeostasis. After adjustment for BMI, insulin sensitivity in both Polynesians and Melanesians was actually greater than in Europeans. Although we used an indirect method (HOMA) for estimation of insulin secretion and insulin resistance, the HOMA indices have been validated against the glucose clamp, for various degrees of glucose tolerance and plasma insulin levels,15 and have been largely used in epidemiological studies.5, 16

Our results are consistent with those of Simmons et al.5, who found that Polynesians of New-Zealand were not ‘intrinsically insulin resistant’ despite a high prevalence rate of diabetes and a greater central and overall obesity compared to Europeans of the same area. In their study, although Polynesians had significantly higher markers of insulin resistance in univariate comparisons, there were no differences after adjustment for BMI (in our study, this was the case even before adjustment for BMI).

Altogether, these two studies suggest that in Polynesians, the progression from normal glucose tolerance to type 2 diabetes may not include an insulin resistance state characterized by compensatory hyperinsulinaemia. It is already known that the pathogenesis of type 2 diabetes is a combination of insulin resistance and reduced insulin secretion,2 and the predominant mechanism is possibly different in various ethnic groups.4 Our study points out to pancreatic deficiency being the main anomaly taking part in the Polynesians' high susceptibility to diabetes mellitus. However, the absence of a frank defect of insulin sensitivity is unexpected, given their general morphotype, as excess adiposity is normally accompanied by insulin resistance17 and decrease in insulin sensitivity is even considered a regulatory response to weight gain, which it serves to limit.18 Of course, we cannot exclude that HOMA-IR is not a correct estimate of insulin resistance in this ethnic group, although it has been validated in subjects with normal to low insulin concentrations at least in one study.15 Besides, fasting plasma insulin levels and insulin sensitivity are reciprocally related,19 making it unlikely that Polynesians would actually be insulin resistant although they have normal insulin levels. If we consider that Polynesians are truly not insulin resistant, one explanation may be that BMI is not a good estimate of adiposity in this particular ethnic group. Indeed, the universal BMI standards for defining ‘overweight’ and ‘obesity’ in adults are based on the risk of adiposity-related diseases in Caucasians,20 and this may not apply to all populations. For instance, in a meta-analysis about ethnic differences in the associations between BMI and per cent body fat, Polynesians had the highest mean BMI, whereas their degree of body fatness was below that of some ethnic groups such as Indonesians, Thais or Ethiopians.21 Another study showed that, at higher BMI levels, Polynesians had significantly more lean mass than Europeans.22 Another possible explanation is that, as Polynesians in New Caledonia are generally employed in heavy, energy-consuming work, this may significantly enhance their insulin sensitivity.4, 23 Also, it cannot be excluded that Polynesians have an insulin-deficient form of diabetes, different from type 2.2, 24, 25

Whatever the explanation, if it is confirmed that insulin resistance in the prediabetic state is not the prominent anomaly in this ethnic group (and maybe others), a new approach to the prevention of type 2 diabetes should be designed.


  1. 1.

    , , , , , . A two-step model for development of non-insulin-dependent diabetes. Am J Med 1991; 90: 229–235.

  2. 2.

    Expert Committee. Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 1997; 20: 1183–1197.

  3. 3.

    , , . The global epidemiology of non-insulin-dependent diabetes mellitus and the metabolic syndrome. J Diabetes Complications 1997; 11: 60–68.

  4. 4.

    , . The impact of ethnicity on type 2 diabetes. J Diabetes Complications 2003; 17: 39–58.

  5. 5.

    , , . Polynesians: prone to obesity and Type 2 diabetes mellitus but not hyperinsulinaemia. Diabetes Med 2001; 18: 193–198.

  6. 6.

    , , . Prevalence of diabetes mellitus in New Caledonia: ethnic and urban–rural differences. CALDIA Study Group. CALedonia DIAbetes Mellitus Study. Am J Epidemiol 1996; 143: 1018–1024.

  7. 7.

    , , , . Nutritional status in adults in the pluri-ethnic population of New Caledonia. The CALDIA Study Group. Int J Obes Relat Metab Disord 1997; 21: 61–66.

  8. 8.

    World Health Organization. Report of a WHO consultation on obesity: preventing and managing the global epidemic. Geneva: WHO, 1998.

  9. 9.

    World Health Organization. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus. WHO: Geneva, 1999.

  10. 10.

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

  11. 11.

    , , , , , et al. Impaired fasting glucose or impaired glucose tolerance - What best predicts future diabetes in Mauritius? Diabetes Care 1999; 22: 399–402.

  12. 12.

    , , , , , et al. Insulin resistance and insulin secretory dysfunction as precursors of non-insulin-dependent diabetes mellitus. Prospective studies of Pima Indians. N Engl J Med 1993; 329: 1988–1992.

  13. 13.

    . Incidence of NIDDM and the natural history of IGT in Pacific and Indian Ocean populations. Diabetes Res Clin Pract 1996; 34 (Suppl): S45–S50.

  14. 14.

    , , , , . Incidence of type II diabetes in Mexican Americans predicted by fasting insulin and glucose levels, obesity, and body-fat distribution. Diabetes 1990; 39: 283–288.

  15. 15.

    , , , , , et al. Homeostasis model assessment closely mirrors the glucose clamp technique in the assessment of insulin sensitivity: studies in subjects with various degrees of glucose tolerance and insulin sensitivity. Diabetes Care 2000; 23: 57–63.

  16. 16.

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

  17. 17.

    , , . Relationship between insulin sensitivity and plasma leptin concentration in lean and obese men. Diabetes 1996; 45: 988–991.

  18. 18.

    , . Pathophysiology of obesity. Lancet 1992; 340: 404–408.

  19. 19.

    , . Insulin: in search of a syndrome. Diabetes Med 2002; 19: 724–729.

  20. 20.

    , , , , , et al. Age, body mass index and Type 2 diabetes – associations modified by ethnicity. Diabetologia 2003; 46: 1063–1070.

  21. 21.

    , , . Body mass index and percent body fat: a meta analysis among different ethnic groups. Int J Obes Relat Metab Disord 1998; 22: 1164–1171.

  22. 22.

    , , , . Body size and composition in Polynesians. Int J Obes Relat Metab Disord 1999; 23: 1178–1183.

  23. 23.

    , , , , , et al. Effect of long-term strength training on glucose metabolism. Implications for individual impact of high lean mass and high fat mass on relationship between BMI and insulin sensitivity. J Endocrinol Invest 2002; 25: 520–525.

  24. 24.

    , . Genetic, metabolic and clinical characteristics of maturity onset diabetes of the young. Eur J Endocrinol 1998; 138: 233–239.

  25. 25.

    . Antibodies to glutamic acid decarboxylase in the prediction of insulin dependency. Diabetes Res Clin Pract 1996; 34 (Suppl): S125–S131.

Download references


We thank each member of the CALDIA Study Group who permitted the design and/or conduction of the CALDIA Study. Coordination, INSERM - in France: L Papoz (principal coordinator), D Simon, J Cubeau, A Lacroux, A Forhan and A Ponton; in New Caledonia: S Barny (local coordinator), D Juranville, H Lmahdi and R Manuohalalo. Health and Social Services in New Caledonia - Medical and Research Commission: F Ledoux and C Merger; Territory: M Germain; North Province: P Calen and P Genty; Loyalty Islands: J Coscoquela and P Buffet; South Province: JG Lambert and P Baqué; and Biochemistry (Territorial Hospital): S Solar and Y Barguil.

We thank V Wirquin of the Henri Mondor Hospital Laboratory (Créteil, France) for the measures of fasting plasma insulin.

This research was financed by Grant 492015 from the INSERM (Institut National de la Santé et de la Recherche Médicale), Grant 607006 from the French Ministry of Cooperation, with additional support from the Medical Commission and the Provinces of New Caledonia, the Rotary and Lions Clubs of Noumea, and Behring Diagnostic Laboratory.

Author information


  1. INSERM, UR024 Epiprev, IRD, Montpellier, France

    • R Defay
    • , A Lacroux
    •  & A Fontbonne
  2. INSERM E0361, Montpellier, France

    • I Jaussent


  1. Search for R Defay in:

  2. Search for I Jaussent in:

  3. Search for A Lacroux in:

  4. Search for A Fontbonne in:

Corresponding author

Correspondence to A Fontbonne.

About this article

Publication history







Further reading