The relationship of overweight and obesity with subjective health and ...

International Journal of Obesity (2002) 26, 247–252 ß 2002 Nature Publishing Group All rights reserved 0307–0565/02 $25.00 www.nature.com/ijo

PAPER The relationship of overweight and obesity with subjective health and use of health-care services among Spanish women P Guallar-Castilloń1,2, E Lo´pez Garcıá2, L Lozano Palacios1,2, JL Gutie´rrez-Fisac2, JR Banegas Banegas2, PJ Lafuente Urdinguio3 and F Rodriguez Artalejo2* 1 Centro Universitario de Salud Pu´blica, Consejerıá de Sanidad y Universidad Autońoma de Madrid, Madrid, Spain; 2Department of Preventive Medicine and Public Health, Universidad Autońoma de Madrid, Madrid, Spain; and 3Department of Preventive Medicine and Public Health, Universidad del Paı´s Vasco, Bilbao, Spain

OBJECTIVE: To examine the relationship of overweight and obesity with subjective health and use of health-care services among women in Spain. METHODS: Data were drawn from the 1993 Spanish National Health Survey, covering a 13 244-woman sample representative of the non-institutionalised Spanish population aged 16 y and over. Information was collected through home-based interviews. Multiple logistic regression models were used to calculate odds ratios for suboptimal health (fair, poor or very poor) and utilisation of health-care services by women with normal weight (BMI 18.5 – 24.9 kg=m2), overweight (BMI 25.0 – 29.9 kg=m2) and obesity (BMI 30 kg=m2). Analyses were adjusted for age, education level, occupation, civil status, social support, tobacco use, alcohol consumption, physical activity at work and during leisure time, job status and town of residence. RESULTS: Frequency of suboptimal health was higher in women with overweight (OR 1.7; 95% CI 1.5 – 1.9) and obesity (OR 2.1; 95% CI 1.8 – 2.5) than in those with normal weight. Overweight and obese women visited the physician, used hospital emergency services and took medication with greater frequency than did women of normal weight. There was a positive dose – response relationship (P < 0.05) of BMI 18.5 kg=m2 with suboptimal health and utilisation of health-care services. These associations were not wholly explained by BMI-related risk factors and chronic diseases, since their statistical significance remained unchanged and their magnitude was only slightly reduced after adjustment for those factors. The association of overweight and obesity with the use of health-care services did not vary with age, educational level or presence of chronic disease. CONCLUSION: Overweight and obese women have worse subjective health and make greater use of health-care services. This finding is an additional argument for implementing weight-control programmes in Spain. International Journal of Obesity (2002) 26, 247 – 252. DOI: 10.1038=sj=ijo=0801862 Keywords: overweight; subjective health; health-care services; women; Spain

Introduction In Europe, prevalence of overweight and obesity is higher among women than among men.1 Since the frequency of female overweight and obesity rises progressively from northern to southern Europe, Spain has one of the highest

*Correspondence: F Rodriguez Artalejo, Departamento de Medicina Preventiva y Salud Pu´blica, Facultad de Medicina, Universidad Autońoma de Madrid, Avda Arzobispo Morcillo s=n, 28029 Madrid, Spain. E-mail: [email protected] Received 16 February 2000; revised 13 July 2001; accepted 26 July 2001

prevalences of such health problems in women.1 In Spain, as in most other European countries, prevalence of obesity has risen in recent years, with young women being one of the population groups most affected by this rise.2,3 Lastly, female overweight and obesity increase with age and tend to be more marked at the lowest educational levels.4 Among women, overweight and obesity are associated with higher total mortality5 and numerous chronic diseases, such as diabetes, ischaemic heart disease, arterial hypertension, high serum cholesterol, gall bladder disease, rheumatic ailments and certain types of cancer.6 – 8 However, to characterise the overall effect of overweight and obesity on

BMI, subjective health and health services use P Guallar-Castillo´ n et al

248 health, other variables, such as subjective health and use of health-care services, are also relevant. Self-perceived health is an overall measure of health which can be obtained in a reproducible manner, correlates well with other measures of health-related quality of life,9 and is a good predictor of mortality.10 Moreover, identification of factors associated with the use of health-care services is useful for correct allocation of resources in health-care systems marked by rising costs. Most research in this field suggests that overweight and obesity are linked to worse subjective health, more functional limitations and greater use of health-care services.11 – 25 Nevertheless, information on whether this relationship might be attributable to diseases associated with overweight or obesity tends to be sparse,17,21,23,24 and some studies even report the magnitude of the association as varying with age, socio-economic level and presence of chronic disease.17,21,23,24 Therefore, the present study examines the relationship of overweight and obesity with subjective health and use of health-care services among women in Spain. In particular, it seeks to answer two concrete questions: (1) if this relationship indeed exists, is it explained entirely by association with known chronic diseases; and (2) given that obesity is associated with age, lower socio-economic level and the presence of chronic disease, does the relationship under study vary with the presence of those variables?

Subjects and methods Data were drawn from the 1993 Spanish National Health Survey (SNHS),26 which covered a random sample of 13 244 women representative of the non-institutionalised Spanish female population aged 16 y and over, stratified by sex, age and size of town or city of residence. The survey’s response rate was 81.7%. SNHS data were collected through home-based interviews. Information on weight and height was obtained by asking interviewees the following questions: ‘Could you tell me how much you weigh, without shoes and clothes on?’ and ‘Could you tell me how tall you are without shoes on?’ Body mass index (BMI) is calculated by the weight in kilograms divided by the square of the height in metres (kg=m2). Subjects were classified into four BMI categories: underweight (BMI < 18.5), normal weight (BMI 18.5 – 24.9), overweight (BMI 25.0 – 29.9) and obesity (BMI 30).27 The SNHS measured subjective health on the basis of the following question, ‘In the last 12 months, would you say that your state of health has been very good, good, fair, poor or very poor?’ Women were deemed as having suboptimal health when they rated their health as fair, poor or very poor. Information on utilisation of health-care services, and in particular, on hospitalisations, medical visits, hospital emergency services and medical drug use were respectively gathered by asking subjects the following four questions: (1) over International Journal of Obesity

the last 12 months, have you been hospitalised for a minimum of one night? (2) In the last 2 weeks, have you consulted a physician about any complaint, ailment, disorder or disease? (3) In the last 12 months, have you made use of any emergency service as a result of a health problem? (4) In the last 2 weeks, have you taken any type of medication (drops, tablets, injections, suppositories, etc)? It should be noted that medical visits in the SNHS do not include dentistry, and that for the purpose of the study obstetric hospitalisations were excluded. The only permissible answers to these four questions were yes or no. To estimate the prevalence of obesity, subjective health and use of health-care services, survey subjects were weighted by the inverse of the sampling fraction.28 Study associations were summarised with odds ratios obtained from unconditional logistic regression. We used five outcome variables: subjective health, hospitalisations, medical visits, utilisation of hospital emergency services, and medical drug use. Independent variables were categories of BMI. We constructed three models, adjusted respectively for: (1) age; (2) age and obesity-related risk factors, such as educational level, occupation, civil status, social support, tobacco use, alcohol consumption, degree of physical activity during work hours, degree of physical activity during leisure time, and size of town or city of residence; (3) all the previous factors plus the presence of chronic disease. Information on variables of adjustment was also drawn from the SNHS.24 Presence of chronic disease corresponded to self-reported data on medical diagnoses of the following health problems: arterial hypertension, high serum cholesterol, diabetes, heart diseases, asthma and chronic bronchitis. BMI and the variables of adjustment were introduced into the models as ‘dummies’. To study the dose – response relationship of BMI with subjective health and use of health-care services, BMI was modelled as a continuous variable. The modification of the relationship under study by age, educational level and presence of chronic disease was ascertained using two procedures: stratified analysis, and the inclusion of interaction terms between BMI and the above variables in the regression models.29 Analyses were performed using the SAS software package.30

Results According to the SNHS, 28% of Spanish women aged 16 y and over were overweight and 9.5% were obese, and prevalence of overweight and obesity rose substantially with age (Table 1). A total of 32.9% of women reported suboptimal health, 5.8% had been hospitalised in the preceding year, 26.8% had attended a medical visit in the fortnight immediately preceding the interview, 13.1% had used an emergency service during the previous 12 months, and 51.3% had taken some type of medication in the previous 2 weeks. Frequency of suboptimal health and utilisation of healthcare services other than hospital emergency services increased with age (Table 1).


Compared with women of normal weight, women with overweight and obesity showed a higher frequency of suboptimal health (model 1, Table 2), and there was a positive dose – response relationship (P < 0.0001) between BMI 18.5 kg=m2 and frequency of suboptimal health. Moreover, this relationship was not wholly explained by risk factors and chronic diseases associated with overweight and obesity, since its statistical significance remained unchanged and its magnitude was only slightly reduced after adjustment for those factors (models 2 and 3, Table 2). No substantial variation in this relationship was seen with age or presence of chronic disease, yet it did vary with educational level (interaction term, P < 0.05). Thus, a rise in frequency of suboptimal health with overweight and obesity was observed for women with some type of formal (ie primary, secondary or university) education, but not for women without such formal instruction (Figure 1). With respect to health-care services, overweight and obese women consulted the physician, made use of health-care emergency services and took medication more frequently than did women of normal weight (model 1, Table 2). There was a positive dose – response relationship between BMI 18.5 kg=m2 and utilisation of health-care services except hospitalisations (linear trend, P < 0.01 in most cases) (Table 2). Hospitalisation frequency also rose with overweight and obesity, but failed to attain statistical significance, probably because of the low number of hospitalizations in the sample (5.8%). In most instances, these associations remained unchanged after adjustment for the principal BMI-related risk factors (model 2 in Table 2) and chronic diseases (model 3, Table 2). Investigation of possible interactions showed that the above associations were not modified by age, educational level or presence of chronic disease (interaction terms, P > 0.05). Lastly, compared with women of normal weight, underweight women tended to report worse subjective health and were hospitalised with greater frequency (Table 2).

Table 1

249

Discussion Our results show that overweight and obesity are associated with worse subjective health and greater utilisation of health-care services among Spanish women. This means that even levels of BMI which are culturally well accepted for middle-aged women in Spain, such as overweight, have a negative impact on health. The worse subjective health of women with overweight and obesity is not wholly explained by BMI-related risk factors and chronic diseases. This suggests that other mechanisms may be at work, which might include a reduction in mobility and functional limitations, or low selfesteem due to deterioration in body image. The fact that this relationship is observed solely among women having a formal education and, by extension, a higher socioeconomic level, supports the hypothesis that one of the possible mechanisms involved may be linked to social-class expectations of personal body image. With regard to the greater utilisation of health-care services by overweight and obese women, this too cannot be wholly attributed to related risk factors and chronic diseases. While SNHS information does not cover all obesity-related diseases — rheumatic ailments and biliary lithiasis in particular — it is likely that the most important of them have been controlled for in the analysis. It is plausible, moreover, that treatment of obesity might in itself lead to utilisation of certain health-care services. Despite possible differences in the determinants of subjective health and health-care service availability and access between Spain and the Northern European and Anglo-Saxon countries, where more research in this field has been done, our results are largely consistent with those obtained in these countries. Thus, several studies have shown that overweight and obesity are associated with negative aspects of healthrelated quality of life, such as restriction on mobility and functional limitations induced by physical problems,11 – 18 joint or spinal pain,12,14,18 alterations in the mental or social

Body mass index (BMI), suboptimal health and use of health-care services among Spanish women, by age Age Total n

Weight Underweight (BMI < 18.5 kg=m2) Normal weight (BMI 18.5 – 24.9 kg=m2) Overweight (BMI 25.0 – 29.9 kg=m2) 2 Obesity (BMI 30.0 kg=m ) Suboptimal health Hospitalisation Medical visit Hospital emergency Consumption of medical drugs

311 4732 2262 768 2656 451 2082 1017 3979

% (95% CI)

3.85 58.61 28.02 9.52 32.90 5.81 26.83 13.10 51.28

16 – 54 y a

(3.45 – 4.30) (57.53 – 59.69) (27.04 – 29.02) (8.89 – 10.18) (31.88 – 33.94) (5.31 – 6.36) (25.85 – 27.83) (12.37 – 13.88) (50.16 – 52.39)

n

286 3985 1279 365 1434 256 1307 719 2485

% (95% CI)

4.84 67.37 21.63 6.17 24.24 4.53 23.12 12.71 43.94

(4.31 – 5.42) (66.16 – 68.56) (20.58 – 22.70) (5.58 – 6.82) (23.16 – 25.36) (4.01 – 5.11) (21.98 – 24.19) (11.86 – 13.62) (42.65 – 45.26)

55 y and over n

25 747 983 403 1222 195 775 298 1494

% (95% CI)

1.15 34.62 45.53 18.70 56.64 9.23 36.81 14.15 71.00

(0.77 – 1.73) (32.61 – 36.67) (43.44 – 47.68) (17.11 – 20.45) (54.50 – 58.73) (8.03 – 10.55) (34.76 – 38.92) (12.71 – 15.74) (69.12 – 73.04)

a

95% CI, 95% confidence interval.

International Journal of Obesity


250 Table 2 Odds ratios (OR) of suboptimal health and use of health-care services by categories of BMI among Spanish women a

b

c

Model 1 d OR (95% CI)

Model 2 OR (95% CI)

Model 3 OR (95% CI)

Suboptimal health 2 < 18.5 kg=m 18.5 – 24.9 kg=m2 25.0 – 29.9 kg=m2 30.0 kg=m2 e P-value for trend

1.22 (0.92 – 1.61) 1.00 1.67 (1.48 – 1.87){ 2.13 (1.80 – 2.52){ 0.0001

1.23 (0.92 – 1.63) 1.00 1.50 (1.33 – 1.69){ 1.79 (1.50 – 2.13){ 0.0001

1.25 (0.93 – 1.65) 1.00 1.42 (1.26 – 1.61){ 1.58 (1.32 – 1.89){ 0.0001

Hospitalisation < 18.5 kg=m2 18.5 – 24.9 kg=m2 2 25.0 – 29.9 kg=m 2 30.0 kg=m e P-value for trend

1.67 (1.01 – 2.62)* 1.00 1.13 (0.89 – 1.42) 1.40 (1.02 – 1.88)* 0.0461

1.73 (1.04 – 2.73)* 1.00 1.06 (0.84 – 1.34) 1.30 (0.94 – 1.77) 0.2252

1.75 (1.06 – 2.77)* 1.00 1.02 (0.80 – 1.29) 1.20 (0.87 – 1.64) 0.4898

Medical visit 2 < 18.5 kg=m 2 18.5 – 24.9 kg=m 2 25.0 – 29.9 kg=m 30.0 kg=m2 P-value for trende

1.10 (0.83 – 1.45) 1.00 { 1.18 (1.04 – 1.33) 1.53 (1.29 – 1.82){ 0.0001

1.11 (0.83 – 1.46) 1.00 1.13 (1.00 – 1.28) 1.44 (1.21 – 1.72){ 0.0001

1.11 (0.84 – 1.47) 1.00 1.09 (0.96 – 1.24) 1.33 (1.11 – 1.59){ 0.0026

Hospital emergency 2 < 18.5 kg=m 18.5 – 24.9 kg=m2 25.0 – 29.9 kg=m2 30.0 kg=m2 e P-value for trend

0.97 (0.68 – 1.37) 1.00 1.22 (1.03 – 1.43){ 1.41 (1.12 – 1.77){ 0.0033

0.97 (0.67 – 1.37) 1.00 1.24 (1.05 – 1.46)* 1.42 (1.12 – 1.80){ 0.0041

0.97 (0.67 – 1.37) 1.00 1.18 (1.00 – 1.39)* 1.27 (1.00 – 1.61)* 0.0751

Consumption of medical drugs < 18.5 kg=m2 2 18.5 – 24.9 kg=m 2 25.0 – 29.9 kg=m 2 30.0 kg=m e P-value for trend

1.00 (0.78 – 1.26) 1.00 { 1.27 (1.13 – 1.42) { 1.48 (1.25 – 1.76) 0.0001

0.99 (0.78 – 1.26) 1.00 { 1.24 (1.11 – 1.40) { 1.40 (1.17 – 1.67) 0.0001

0.98 (0.77 – 1.26) 1.00 { 1.15 (1.02 – 1.30) 1.17 (0.97 – 1.41) 0.0315

a

Adjusted for age. Adjusted for age (16 – 24, 25 – 44, 45 – 64, 65 y and over), educational level (no formal education, primary, secondary, university), occupation (active, retired, inactive, student, housewife), civil status (married, single), social support (yes, no), tobacco consumption (never-smoker, ex-smoker, current smoker), alcohol consumption (abstainer, 20, > 20 and 40, > 40 and 60, > 60 g=day and over), degree of physical activity during work hours (intense, regular, moderate, inactive), degree of physical activity during leisure time (intense, regular, moderate, inactive), and size of town or city of residence ( < 10 000, 10 001 – 100 000, 100 001 – 400 000, 400 001 – 1 000 000, over 1 000 000 inhabitants). c Adjusted for variables in model 2 and presence of chronic disease. d 95% CI, 95% confidence interval. e Linear trend of BMI 18.5 kg=m2. *P < 0.05. {P < 0.01. {P < 0.001. b

sphere,11,18 – 21 and poor subjective health.11,18,22 However, unlike some of these studies, we have not observed a modification by age of the relationship between BMI and subjective health.17,21 Likewise, with regard to use of health-care services, our results also coincide with those reported by other authors. An increase in the number of medical visits11,14,22 – 25 and consumption of medical drugs11,24,25 has been observed for overweight and obese persons. However, a study conducted in Canada observed that obese persons were hospitalised less frequently.11 Among the study’s limitations, first there is the crosssectional design, which cannot ascertain the direction of International Journal of Obesity

the relationship between variables. Consequently, it cannot be ruled out that poor subjective health and other health problems associated with greater use of health-care services might induce changes in lifestyle, which lead to weight gain. However, the biological plausibility of our findings and their consistency in the literature suggests that the most reasonable explanation is that the results may, at least in part, be due to a mechanism whose direction runs from overweight and obesity to poor health and use of health-care services. Second, compared with women who answered the questions on weight and height, non-responders were older (38 vs 14% aged > 65 y), showed worse subjective health (50 vs 33% with suboptimal health), reported greater use of health-care


251

Figure 1 Odds ratios of suboptimal subjective health by categories of BMI and educational level in Spanish women. Odds ratios are adjusted for the same variables as in model 3 in Table 2, except educational level.

services (32 vs 28% medical visits), registered a higher frequency of chronic disease (38 vs 22%) and had a lower educational level (2 vs 8% with university education). Despite our survey’s high response rate, this probably has underestimated the prevalence of overweight and obesity in our sample, and thus attenuated the impact of these variables on subjective health and utilisation of health-care services among Spanish women. Third, weight and height were self-reported. This tends to underestimate weight and overestimate height, and leads to an overall underestimation of BMI.31 Because, there is no evidence that this error varies with subjective health and use of health-care services, our results are probably conservative.29 Fourth, the SNHS does not inform of the reasons for hospitalisations or medical visits and the specific types of medication taken by overweight and obese women. Last, we have not controlled for the effect of certain factors affecting the utilisation of services, such as their availability and accessibility, due to lack of information in the SNHS. In Spain, there is a publicly funded National Health System that covers all of the population independent of income. However, we cannot exclude barriers to accessing the health services related to socioeconomic status or geographical distance to health services. To limit the influence of those barriers on our results, analyses were adjusted for educational level, occupation, and size of the city or town of residence. It should also be noted that, while the organisation of health services provision varies substantially across countries, our results on the association between BMI and health services use is largely consistent with those found in the international literature. Despite these limitations, this population-based study, conducted on a large sample representative of the Spanish

women, is of practical relevance. First, since subjective health is better, and use of health-care services lower, among subjects of normal weight than in underweight, overweight and obese subjects, this study provides additional evidence that the ideal weight for women lies in the BMI range of 18.5 – 24.9.32 Second, it demonstrates that overweight and obesity are associated with poor subjective health. Because this is a measure of health highly regarded by individuals, who perceive it as correlating with quality of life, our results may stimulate women to change diet and increase physical activity for reduction of such health problems. Finally, since the prevalence of overweight and obesity are both increasing in Spain, it is foreseeable that their negative impact on population health will increase over the coming years. In addition, greater utilisation of health-care services by a growing population with overweight and obesity must result in increased health-care expenditure. Costestimates of obesity drawn up in Europe33 and the USA34 highlight the importance of this aspect. Hence, this study gives a new argument for implementing weight-control programmes; these programmes should particularly address overweight, since this is a prevalent phenomenon that is culturally well accepted among women of adult and advanced age.

Acknowledgements This work was partially funded by a grant from ‘Fundacio´ n de Investigacio´ n y Docencia de Enfermedades Cardiovasculares’ (FIDEC) and a contract with Knoll Laboratories of Spain. The authors were solely and independently responsible for the data management and analysis. International Journal of Obesity


252

References 1 Seidell JC. Obesity in Europe: scaling an epidemic. Int J Obes Relat Metab Disord 1995; 19(Suppl 3): S1 – S4. 2 Gutie´ rrez-Fisac JL. La obesidad: una epidemia en curso. Med Clin (Barc) 1998; 111: 456 – 458. 3 Gutie´ rrez Fisac JL, Banegas Banegas JR, Rodrı´guez Artalejo F, Regidor E. Increasing prevalence of overweight and obesity among Spanish adults, 1987 – 1997. Int J Obes Relat Metab Disord 2000; 24: 1677 – 1682. 4 Gutie´ rrez-Fisac JL, Regidor E, Rodrı´guez C. Prevalencia de la ã. Med Clin (Barc) 1994; 102: 10 – 13. obesidad en Espan 5 Manson JE, Willett WC, Stampfer MJ, Colditz GA, Hunter DJ, Hankinson SE, Hennekens CH, Speizer FE. Body weight and mortality among women. New Engl J Med 1995; 333: 677 – 685. 6 Burton BT, Foster WR, Hirsch J, Van Itallie TB. Health implications of obesity: NIH consensus development conference. Int J Obes Relat Metab Disord 1985; 9: 155 – 169. 7 Manson JE, Colditz GA, Stampfer MJ, Willett WC, Rosner B, Monson RR, Speizer FE, Hennekens CH. A prospective study of obesity and risk of coronary heart disease in women. New Engl J Med 1990; 322: 882 – 889. 8 Pi-Sunyer FX. Comorbidities of overweight and obesity: current evidence and research issues. Med Sci Sports Exerc 1999; 31(Suppl): S602 – S608. 9 Brazier JE, Harper R, Jones NM, O’Cathain A, Thomas KJ, Usherwood T, Westlake L. Validating the SF-36 health survey questionnaire: new outcome measure for primary care. Br Med J 1992; 305: 160 – 164. 10 Ilder EL, Benyamini Y. Self-rated health and mortality: a review of twenty-seven community studies. J Health Soc Behav 1997; 38: 21 – 37. 11 Trakas K, Lawrence K, Shear NH. Utilization of health care resources by obese Canadians. Can Med Assoc J 1999; 160: 1457 – 1462. ¨ ssner S. Obesity and self-perceived health in Sweden. 12 Wolk A, Ro Int J Obes Relat Metab Disord 1996; 20: 369 – 372. 13 Han TS, Tijhuis MA, Lean ME, Seidell JC. Quality of life in relation to overweight and body fat distribution. Am J Public Health 1998; 88: 1814 – 1820. 14 Lean ME, Han TS, Seidell JC. Impairment of health and quality of life using new US federal guidelines for the identification of obesity. Arch Intern Med 1999; 159: 837 – 843. 15 Launer LJ, Harris T, Rumpel C, Madans J. Body mass index, weight change and risk of mobility disability in middle-aged and older women. The epidemiologic follow-up study and NHANES I. JAMA 1994; 271: 1093 – 1098. ¨ vaara M, Knekt P, Reunanen A, Aromaa A, 16 Rissanen A, Helio Maatela J. Risk of disability and mortality due to overweight in a Finnish population. Br Med J 1990; 301: 835 – 837. 17 Seidell JC, Bakx KC, Deurenberg P, Burema J, Hautvast J, Huygen F. The relation between overweight and subjective health according to age, social class, slimming behavior and smoking habits in Dutch adults. Am J Public Health 1986; 76: 1410 – 1415.

International Journal of Obesity

18 Fontaine KR, Cheskin LJ, Barofsky K. Health-related quality of life in obese persons seeking treatment. J Fam Pract 1996; 43: 265 – 270. 19 Sullivan M, Karlsson J, Sjostrom L, Backman L, Bengtsson C, Bouchard C, Dahlgren S, Jonsson E, Larsson B, Lindstedt S. Swedish obese subjects (SOS) — an intervention study of obesity. Baseline evaluation of health and psychosocial functioning in the first 1743 subjects examined. Int J Obes Relat Metab Disord 1993; 17: 503 – 512. 20 Perula de Torres LA, Martıńez de la Iglesia J, Ruiz Moral R, Fonseca del Pozo J. Variables relacionadas con la autopercepcio´ n del estado de salud: un estudio poblacional. Aten Primaria 1995; 16: 323 – 329. 21 Manderbacka K, Lundberg O, Martikainen P. Do risk factors and health behaviours contribute to self-rating of health? Soc Sci Med 1999; 18: 1713 – 1720. 22 Fontaine KR, Faith MS, Allison DB, Cheskin LJ. Body weight and health care among women in the general population. Arch Fam Med 1998; 7: 381 – 384. 23 Quesenberry CP, Caan B, Jacobson A. Obesity, health services use, and health care costs among members of a health maintenance organization. Arch Intern Med 1998; 158: 466 – 472. 24 Sansone RA, Sansone LA, Wiederman MW. The relationship between obesity and medical utilization among women in a primary care setting. Int J Eat Disord 1998; 23: 161 – 167. ¨ ster I, Ferber L. Frequency of obesity in medical 25 Hauner H, Ko records and utilization of out-patient health care by obese subjects in Germany. An analysis of health insurance data. Int J Obes Relat Metab Disord 1996; 20: 820 – 824. 26 Ministerio de Sanidad y Consumo. Encuesta Nacional de Salud de ˜ a 1993. Ministerio de Sanidad y Consumo: Madrid; 1995. Espan 27 World Health Organization Expert Committee. Physical status: the use and interpretation of anthropometry. WHO Technical Report Series no. 854. WHO: Geneva; 1995. 28 Levy PS, Lemeshow S. Sampling for health professionals. Lifetime Learning: Belmont, CA; 1980. 29 Kleinbaum DG, Kupper LL, Morgenstern H. Epidemiologic research. Principles and quantitative methods. Lifetime Learning: Belmont, CA; 1982. 30 SAS=STAT Guide for personal computers, version 6.12. SAS Institute: Cary, NC; 1996. 31 Nieto Garcıá FJ, Bush TL, Keyl PM. Body mass definitions of obesity: sensitivity and specificity using self-reported weight and height. Epidemiology 1990; 1: 146 – 152. 32 Brown WJ, Dobson AJ, Mishra G. What is a healthy weight for middle aged women? Int J Obes Relat Metab Disord 1998; 22: 520 – 528. 33 Seidell JC. The impact of obesity on health status: some implications for health care cost. Int J Obes Relat Metab Disord 1995; 19(Suppl 6): S13 – S16. 34 Wolf AM, Colditz GA. The cost of obesity: the US perspective. PharmacoEconomics 1994; 5(Suppl 1): 34 – 37.