¹Department of Endocrinology, Hospital Dr Negrín, Las Palmas de Gran Canaria, Spain

²Department of Preventive Medicine Hospital Dr Negrin, Las Palmas de Gran Canaria, Spain

Correspondence to: P L de Pablos-Velasco, Department of Endocrinology, Hospital Dr Negrin, Barranco de la Ballena, 350120 Las Palmas de Gran Canaria, Spain. E-mail: depablos@correo.hpino.rcanaria.es

Guarantor: P de Pablos-Velasco.

Contributors: PLdP-V was principal investigator; FJM-M was involved in statistical analysis and discussion; FR-P did the field work.

Objective: We sought to estimate the prevalence of obesity and central obesity, and their association with type 2 diabetes mellitus in the Canarian community of Guía.

Design and setting: Population-based study.

Subjects: A random sample of 691 subjects over 30 y old (stratified by age and sex) was studied.

Data and measures: Age, sex, family history of diabetes and medication use were obtained, height, weight and waist circumference were measured and standard oral glucose tolerance tests were performed

Results: The prevalences of obesity/central obesity were 36.5%/66.5% (women) and 23.6%/32.0% (men). The prevalence of diabetes was 21.0% (women) and 18.4% (men). These rank among the highest in Europe. Bivariate analyses show a strong association of both obesity and central obesity with diabetes mellitus (P<0.001), but in a multivariate model, waist circumference (P<0.001) but not body mass index (P=0.212) was retained as an independent predictor of diabetes.

Conclusion: The prevalences of obesity, central obesity and diabetes in our community are extremely high, and central obesity is a better predictor of diabetes than obesity.

Sponsorship: This study was supported by a grant of the Government of the Canarian Community, Public Health Department, no. BOCA-1996.

European Journal of Clinical Nutrition (2002) 56, 557-560. doi:10.1038/sj.ejcn.1601401

obesity; central obesity; type 2 diabetes mellitus; cardiovascular risk factors

Introduction

In the Canary Islands the prevalence of type 2 diabetes mellitus (DM-2) is among the highest ever reported in Caucasian populations (de Pablos-Velasco et al, 2001). Obesity, especially central obesity, is the major risk factor (Seidell et al, 1997; Modan et al, 1986; Björntorp, 1994; Boyko et al, 2000) for DM-2.

There are wide variations among the known prevalence of obesity in the European countries (Seidell, 1997). Previous estimations of the prevalence of obesity in the Canary Islands indicate that it may be among the highest in Europe (Navarro Rodríguez et al, 2000; Rodríguez-Pérez et al, 1993). The aim of this study was to establish the prevalences of overweight, obesity and central obesity in a Canarian community (Guía) and their associations with DM-2.

Methods

In the council of Guía (Canary Islands) a stratified random sample of the 6355 resident inhabitants aged 30 y or older was taken. Subjects from each sex were chosen randomly from the municipal roster in 10 y age groups, as previously described (de Pablos-Velasco et al, 2001).

One physician and one nurse who were specially trained for this study performed all procedures. Venous blood was obtained after a 12 h fasting period in all subjects; afterwards a standard oral glucose tolerance test (OGTT) was performed according to the WHO recommendations (WHO, 1994), excluding those subjects with a prior diagnosis of DM.

The following information was requested from each subject: name, age, sex and chronic medication use. Personal and familial history of DM were considered positive if the subject stated that they had been diagnosed by their physician. Anthropometric measures were determined according to standardized procedures (WHO, 1995).

Plasma glucose was measured by standard methods (Trinder, 1969) using a Boehringer Mannheim Hitachi 717 autoanalyzer (Tokyo, Japan).

The study design was revised and approved by the Ethical Committee of our Hospital. All subjects gave their informed consent for the study. DM and impaired glucose tolerance (IGT) were defined according to the 1998-WHO diagnostic criteria (Alberti & Zimmet, 1998).

Low weight was defined as a body mass index (BMI)<18.5; overweight as a BMI between 25 and 29.99; obesity as a BMI30¾grade I obesity as having a BMI between 30 and 34.99; grade II obesity as having a BMI between 35 and 39.99; and grade III obesity as having a BMI40 kg/m², as defined by the WHO (1998). 'Central obesity' was defined as an equivalent of 'increased waist circumference', ie having a waist circumference >88 cm in women and >102 cm in men (NIH, 1998).

Numerical variables are summarized as mean±standard deviation. All statistical analyses (contingency table analysis, t-tests and multivariate logistic regression analysis) were performed with SSPS 9.0 (SPSS Inc., Chicago, IL, USA, 1999). The significance level was P<0.05.

Results

Of 900 subjects who were invited, 691 were tested; the mean response rate was 76.8%. The response rate was similar in all age and sex groups (², P=0.29).

The mean BMI was 28.1±4.6 kg/m² and the global prevalence of obesity was 30.8%. The mean waist circumference was 95.3±14.3 cm (women) and 97.6±11.5 (men) and the global prevalence of central obesity was 51.4%. In Table 1 these prevalences are shown by decade and gender.

There were 25 subjects (24 of them women) with normal weight but central obesity, and 130 subjects (118 of them women) with overweight and central obesity. Globally, 23.1% of the population were not obese but had increased waist circumference; 76.1% of them were women.

There were only two subjects with type 1 DM, they were excluded from the analyses, and the rest of the diabetic patients were assumed to have DM-2. The global prevalences of DM-2 and IGT were 19.7 and 16.4%, respectively; 7.7% of the subjects with DM-2 and all of the subjects with IGT were not previously aware of their condition. BMI was 29.97±5.04 kg/m² in the unaware and 29.95±4.73 kg/m² in the aware diabetic subjects (P=0.988, non-paired t-test); waist circumference was 102.8±11.7 cm in the unaware and 103.1±13.1 cm in the aware diabetic subjects (P=0.737, non-paired t-test). The associations between obesity or central obesity and DM-2 or IGT are shown in Table 2; these associations were not significantly different between aware and unaware diabetic subjects.

Of the subjects with a positive family history of DM, 28.5% (91/319) had DM-2, vs 12.2% of those with a negative one (relative risk 2.33; odds ratio 2.88, 95% CI 1.94-4.28, ², P=0.000). Thus, the association of DM-2 with a positive family history seems stronger than with obesity (odds-ratio 2.25; 1.53-3.31) but weaker than with central obesity (odds-ratio 3.01; 1.99-4.56). In order to assess the weight of each of these factors, a multivariate logistic regression analysis was performed, with the presence of DM-2 as the dependent variable, BMI, waist circumference and age as continuous independent variables, and gender and family history of DM as categorical independent variables. The respective odds-ratio, 95% confidence intervals and probabilities for female gender, BMI, family history of DM, age and waist circumference were 1.23 (0.79-1.91; P=0.360), 1.04 (0.98-1.12; P=0.212), 3.08 (1.99-4.76; P=0.001), 1.03 (1.01-1.05; P=0.000) and 1.05 (1.03-1.07; P=0.000), with an overall 82.7% of correct predictions. In order to confirm the exclusion of BMI as an independent predictor of DM-2, a second logistic regression analysis was performed, with BMI and waist circumferences introduced as categorical variables (quartiles of the internal distribution). The P-values for female gender, BMI, waist circumference, family history of DM and age were 0.343, 0.291, 0.017, 0.000 and 0.000, with 80.6% correct predictions. Thus, waist circumference, age and positive family history of DM, but not gender and BMI, were retained as independent predictors of DM-2.

The exclusion of BMI as an independent predictor of DM-2 suggests that in subjects who were centrally obese, the prevalences of DM-2 should be similar in all BMI categories. In fact, these prevalences for normal weight, overweight and obese subjects were 24.0, 23.1 and 31.1% (² not significant, P=0.269). Also, for subjects who were not centrally obese, the respective prevalences were 8.3, 13.0 and 15.8% (² not significant, P=0.337). On the other hand, for each BMI category, the prevalence of DM-2 was significantly higher in centrally obese subjects than in not centrally obese subjects (², P=0.016; P=0.018 and P=0.024 for normal weight, overweight and obese subjects).

Discussion

This is the first population-based study of the prevalence of obesity and central obesity and their relationship with diabetes mellitus in the Canary Islands. The prevalence of obesity in Europe is reportedly highest in the Eastern countries (Seidell, 1997). However, in our community we have found the highest prevalence of obesity in European men and the third highest in women after Lithuania and Russia. Our prevalences of central obesity were also strikingly high.

These results are concordant with the extremely high prevalence of DM-2 in our community (19.7%; de Pablos-Velasco et al, 2001), and could be partly explained because in the last few decades the lifestyle in our community has changed dramatically, from an agriculture-based economy to a tourist services economy, with the subsequent decrease in physical activity and change in dietary habits (Serra-Majem et al, 2000), although the possible contribution of other environmental and genetic factors is unknown.

Within the limitations of a cross-sectional study, our analyses have shown that the impact of central obesity on DM-2 is greater than the impact of obesity as previously described in other populations (Modan et al, 1996; Björntorp, 1994; Boyko et al, 2000). The impact of a positive family history of DM-2 was similar to that of central obesity, while the impact of obesity was lower. In addition, the multivariate logistic regression analyses showed that, when the waist circumference was included in the model, the BMI was no longer a predictor of DM-2, as shown also in Japanese Americans (Seidell et al 1997; Boyko et al, 2000). The association of IGT with central obesity in our population was also highly significant, but the association with obesity was weak. These findings seem to imply that the impact of extra-abdominal fat on glucose metabolism is small.

One remarkable finding of our survey is the fact that the prevalence of central obesity (ie increased waist circumference) clearly exceeds the total prevalence of obesity. This is explained by the fact that many non-obese subjects have increased waist circumference. Almost a quarter of our population (predominantly women) is in this situation, which is associated with a high prevalence of DM-2. We are not aware of previous reports of a similar finding and propose that a new category be included in future obesity classifications, namely 'non-obese subjects with increased waist circumference' or an equivalent descriptor.

The mechanisms that link central obesity and DM-2 are not fully established. The increased production of free fatty acids (Boden, 1999) and other mediators as TNF (Hotamisligil, 1999) and resistin (Steppan et al, 2001) have been implied. Also, it has been shown that decreased adipocyte GLUT-4 expression is associated with obesity and impairs insulin action in muscle and liver (Abel et al, 2001); and that intramyocellular lipid accumulation is associated with insulin resistance (Dobbins et al, 2001). More studies are needed in order to elucidate the relationship of central obesity and DM-2.

Abel ED, Peroni O, Kim JK, Kim Y-B, Boss O, Hadro E, Minnemann T, Shulman GI, Kahn BB. (2001). Adipose-selective targeting of the GLUT4 gene impairs insulin action in muscle and liver. Nature, 409: 729-733. Article MEDLINE

Alberti KG, Zimmet PZ. (1998). Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabetes Med., 15: 539-541.

Björntorp P. (1994). Abdominal obesity and the metabolic syndrome. Ann. Med., 24: 465-468.

Boden G. (1999). Role of fatty acids in the pathogenesis of insulin resistance and NIDDM. Diabetes, 46: 3-10.

Boyko EJ, Fujimoto WY, Leonetti DL, Newell-Morris L. (2000). Visceral adiposity and risk of type 2 diabetes: a prospective study among Japanese Americans. Diabetes Care, 23: 465-471. MEDLINE

de Pablos-Velasco PL, Martínez-Martín FJ, Rodríguez-Pérez F, Anía BJ, Losada A, Betancor P. (2001). Prevalence and determinants of diabetes mellitus and glucose intolerance in a Canarian Caucasian population. Comparison of the 1997-ADA and the 1985-WHO criteria. The Guía Study. Diabetes Med., 18: 235-241.

Dobbins RL, Szczepaniak LS, Bentley B, Esser V, Myhill J, McGarry JD. (2001). Prolonged inhibition of muscle carnitine palmitoyltransferase-1 promotes intramyocellular lipid accumulation and insulin resistance in rats. Diabetes, 50: 123-130. MEDLINE

Hotamisligil GS. (1999). The role of TNF and TNF receptors in obesity and insulin resistance. J. Intern. Med., 245: 621-625. Article MEDLINE

Modan M, Karasik A, Halkin H, Fuchs Z, Lusky A, Shitrit A, Modan B. (1986). Effect of past and concurrent body mass index on prevalence of glucose intolerance and type 2 (non-insulin-dependent) diabetes and on insulin response. The Israel study of glucose intolerance, obesity and hypertension. Diabetologia, 29: 82-89. MEDLINE

Navarro Rodríguez MC, Laínez Sevillano P, Ribas Barba L, Serra Majem L. (2000). [Anthropometric values and cardiovascular risk factors in Canary Islands.]. Arch. Latinoam. Nutr., 50: ((Suppl 1)) 34-42, (in Spanish). MEDLINE

NIH. (1998). Clinical guidelines on the identification evaluation and treatment of overweight and obesity in adults. The evidence report. Obes. Res., 6: ((Suppl 2)) 51S-209S.

Rodríguez Pérez JC, Calonge Ramirez S, Bichara Antanios G. (1993). [Prevalence of ischemic cardiopathy risk factors on the island of Lanzarote.]. Med. Clin. (Barc.), 101: 45-50, (in Spanish). MEDLINE

Seidell JC. (1997)). Time trends in obesity: an epidemiological perspective. Horm. Metab. Res., 29: 155-158. MEDLINE

Seidell JC, Han TS, Feskens EJM, Lean MEJ. (1997)). Narrow hips and broad waist circumferences independently contribute to increased risk of non-insulin dependent diabetes mellitus. J. Intern. Med., 242: 401-406. MEDLINE

Serra-Majem L, Santana-Armas JF, Salmona E. (2000). Dietary habits and nutritional status in Spain. Wld Rev. Nutr. Diet, 87: 127-159.

Steppan CM, Bailey ST, Bhat S, Brown EJ, Banerjee RR, Wright CM, Patel HR, Ahima R, Lazar M. (2001). The hormone resistin links obesity to diabetes. Nature, 409: 307-312. Article MEDLINE

Trinder P. (1969). Determination of glucose in blood using glucose oxidase with an alternative oxygen acceptor. Ann. Clin. Biochem., 6: 24-27.

WHO. (1994). Prevention of Diabetes Mellitus Report of a WHO Study Group. Second report of the WHO expert committee on diabetes mellitus. Technical Report Series 844 Geneva: WHO.

WHO. (1995). Physical Status: the Use and Interpretation of Anthropometry Report of a WHO Expert Committee. Technical Report Series 854 Geneva: WHO.

WHO. (1998). Obesity: Preventing and Managing the Global Epidemic Report of a WHO Consultation on obesity. Programmme of Nutrition, Family and Reproductive Health Geneva: WHO.

Table 1 Prevalences of low weight, normal weight and obesity by decade and gender, in the population (over 30 y old) of Guía

Table 2 Associations between the presence of type 2 diabetes mellitus (upper panel) or impaired glucose tolerance (lower panel) and the presence of obesity or central obesity, globally and by gender, in the population (over 30 y old) of Guía