Patient Education and Counseling 44 (2001) 227±233 General practitioners' and patients' models of obesity: whose problem is it? Jane Ogden*, Induja Bandara, Howard Cohen, David Farmer, Joan Hardie, Harry Minas, Jane Moore, Shabbeer Qureshi, Fiona Walter, Mary-Anne Whitehead Department of General Practice, Guys Kings and St Thomas' Medical Schools, Kings College London, 5 Lambeth Walk, London SE 11 6SP, UK Received 15 June 2000; received in revised form 14 October 2000; accepted 26 November 2000 Abstract Primary care literature emphasises the importance of agreement and shared models in the consultation. This study compared general practitioners' (GPs') and patients' models of obesity. Questionnaires concerning beliefs about the causes, consequences and solutions to obesity were completed by 89 general practitioners (GPs) and 599 patients from practices across UK. In terms of causes, the results showed that the patients were more likely to attribute obesity to a gland/hormone problem, slow metabolism and stress than the GPs, whereas the GPs were more likely to blame eating too much. In terms of consequences, the patients rated dif®culty getting to work more highly whereas the GPs regarded diabetes as more important. For the solutions to obesity, the two groups reported similar beliefs for a range of methods, but whereas the patients rated the GP and a counsellor as more helpful, the GPs rated the obese person themselves more highly. It is argued that patients show a self serving model of obesity by blaming internal uncontrollable factors for causing obesity yet expecting external factors to solve it. In contrast, GPs show a victim blaming model by attributing both cause and the solution to internal controllable factors. Such differing models have implications regarding the form of intervention likely to be used in primary care and indicate that whereas patients would prefer a more professional based approach, GPs would prefer a more patient-led one. Further, the results suggest that even if an intervention could be negotiated, success rates would be low as either the patient or the GP would be acting in contradiction to their beliefs about the nature of obesity. # 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Agreement; Models; Obesity; Doctor/patient communication 1. Introduction Research in primary care has increasingly focused on the multidimensional nature of the primary care intervention. This approach takes into account the * Corresponding author. Tel.: ‡44-20-7735-8881/ext. 417; fax: ‡44-20-7793-7232. E-mail address: jane.ogden@kcl.ac.uk (J. Ogden). complexity of the doctor±patient relationship and the communication which occurs between them. In particular, it has emphasised agreement between these two parties. For example, Pendleton et al. [1] argued that the central tasks of a consultation involved agreement with the patient about the nature of the problem, the action to be taken and subsequent management. Likewise, Neighbour [2] provided a detailed practical approach to the consultation central to which was an 0738-3991/01/$ ± see front matter # 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0 7 3 8 - 3 9 9 1 ( 0 0 ) 0 0 1 9 2 - 0 228 J. Ogden et al. / Patient Education and Counseling 44 (2001) 227±233 emphasis on understanding the patient's language and perspective and the use of matching as a consultation skill. Star®eld et al. [3] and Bass et al. [4] also examined the degree of agreement between doctors and patients within consultations and suggested that concordance may predict patient outcomes. Further, Krupat et al. [5] examined congruence between the practice orientation of physicians and their patients and concluded that patients were more satis®ed when their physicians showed a similar orientation to themselves. This literature has highlighted the importance of agreement in terms of the patients' and doctors' models of the consultation. It has also addressed agreement concerning the presenting problem. How people make sense of their problems has been addressed by psychological research over recent years within the context of explanatory models and illness representations. Attribution theory formed the basis of this work, e.g. [6,7] and illustrated that individuals are motivated to make sense of their world by looking for causes. Developing this perspective, Brickman et al. [8] argued that people also make attributions for the solutions to their problem and suggested that attributions for causality and solutions may be different. For example, their work with alcoholics indicated that whereas an individual may make internal attributions for the cause of the problem they could make external attributions for its solutions. Leventhal et al. [10] also examined how individuals make sense of their illness and suggested that symptom perception results in the development of illness cognitions which consistently re¯ect the identity, causes, consequences, time line and potential cure of any given problem [9,10]. In line with this perspective research has also explored how people make sense of problems including the adjustment to chronic illness [11], the experience of cancer treatment [12] and the experience of being HIV positive [13]. Therefore, if agreement is to occur between practitioner and patient then not only should these parties agree as to the nature of the consultation but also about the illness being considered. Tuckett et al. [14] addressed this issue and following an analysis of over 1000 consultations concluded that the consultation should be conceptualised as a `meeting between experts' highlighting the importance of the patient's and doctor's potentially different views of the problem. Research has also compared both GPs and patients models of health per se [15] and their models of depression [16]. The former study indicated that whereas GPs adhere to a more medical model of health in general, patients show more complex models drawing upon both contemporary and historical perspectives [15]. The latter study showed that whereas doctors hold a medical model of depression and emphasise the importance of somatic symptoms patients focus more on mood related symptoms [16]. Disagreement between doctors and patients has implications for successful communication. It also has implications for the success of any intervention. The present study addressed doctors' and patients' models of obesity which is of particular interest as obesity raises issues concerning blame and responsibility [17,18] and is associated with negative stereotypes often derived from assumptions about behaviour [19]. Furthermore, success rates for primary care interventions for obesity are exceptionally poor [20,21] and disagreement between practitioner and patient may provide a possible explanation for this phenomenon. In particular, in line with previous research [8±10] the present study explored GPs' and patients' models of obesity in terms of its causes, consequences and possible solutions. 2. Methods 2.1. Participants Questionnaires were distributed to 900 consecutive patients visiting nine practices based in Essex (n ˆ 1), Surrey (n ˆ 3), Cambridge (n ˆ 1), London (n ˆ 1), Bristol (n ˆ 1), Liverpool (n ˆ 1) and Lincoln (n ˆ 1). General practitioners (n ˆ 135) in and around these practices were also given questionnaires. The response rate for the patients was 66.6% (n ˆ 599) and for the GPs was 66.0% (n ˆ 89). 2.2. Design The study used a cross-sectional survey design with comparisons made between GPs and patients. 2.3. Measures Participants were asked to complete a questionnaire consisting of the following sections. J. Ogden et al. / Patient Education and Counseling 44 (2001) 227±233 229 2.3.1. Pro®le characteristics Participants were asked to record their age (years), sex, present weight, height (to compute body mass index (BMI)), ethnic group (white/Asian/black/other) and subjective social class (working/lower middle/ middle/other). compare their beliefs about obesity. Following tests for homogeneity of variance, analysis used ANOVA for parametric data. 2.3.2. Beliefs about obesity Participants rated a series of a statements relating to the causes, consequences and solutions to obesity on ®ve point Likert scales ranging from `not at all' (1) to `totally' (5). These are shown in Fig. 1. 3.1. Pro®le characteristics 3. Results The GPs' and patients' pro®le characteristics are shown in Table 1. 3.2. Beliefs about obesity 2.4. Statistics The results were analysed ®rstly to describe GPs' and patients' pro®le characteristics and then to 3.2.1. Causes of obesity The participants' beliefs about the causes of obesity are shown in Table 2. For medical causes, the results Fig. 1. Measuring beliefs about obesity 230 J. Ogden et al. / Patient Education and Counseling 44 (2001) 227±233 Table 1 GP and patient pro®le characteristics Variables GPs (n ˆ 89) Patients (n ˆ 599) Age (years) 41.46  7.83 43.19  16.07 Sex Male Female 49 (60.5%) 32 (39.5%) BMIa 24.9  3.51 Ethnic group White Asian Black Other 65 15 2 1 (78.3%) (18.1%) (2.4%) (1.2%) 500 46 26 12 (85.6%) (7.9%) (4.5%) (2.1%) Class Working Lower middle Middle Upper 1 6 62 11 (1.3%) (7.5%) (77.5%) (13.8%) 208 125 214 12 (37.2%) (22.4%) (38.3%) (2.1%) a a 195 (34.3%) 394 (65.7%) 26.19  5.34 Mean/S.D. showed that the GPs and patients were comparable in their rating of genetics as a cause of obesity, but that the patients rated a gland/hormone problem, and slow metabolism as more important causes than the GPs. For psychological causes, the results showed that the patients rated anxiety/stress as a more important cause than the GPs but that the two groups were comparable for beliefs about low self esteem and depression. For behavioural and social causes, the GPs showed a higher rating for eating too much and patients showed a higher rating for low income as causes of obesity. The two groups were similar in their ratings of other behavioural and social causes. 3.2.2. Consequences of obesity The participants ratings of the consequences of obesity are shown in Table 3. For medical consequences, the results showed that the GPs reported a greater rating of diabetes as a consequence of obesity than the patients but the two groups were comparable for their ratings of painful joints, heart disease and high blood pressure. For psychological consequences, the results showed no differences between the GPs and patients for any of the ratings and for social consequences the results indicated that the GPs and patients were comparable in their ratings of dif®culty making friends and getting medical/surgical treatments. However, the patients rated dif®culty getting work as a more likely consequence than the GPs. 3.2.3. Solutions to obesity Participants ratings for the solutions to obesity are shown in Table 4. For methods used, the results Table 2 GPs and patients beliefs about the causes of obesitya Variables GPs (n ˆ 89) Patients (n ˆ 599) F-statistic P-value Medical Genetics Gland/hormone problemb Slow metabolismb 3.04  0.97 2.34 ‡ 1.09 2.36  1.06 2.83  0.94 3.21  1.05 3.07  1.07 3.11 48.59 32.43 0.078 0.00001 0.0001 Psychological Low self esteem Anxiety/stressb Depression 2.92  0.99 2.71  1.03 2.97  1.03 2.93  1.15 3.05  1.02 3.04  1.07 0.007 7.99 0.30 0.93 0.004 0.58 Behavioural Eating too muchb Not enough exercise Eating the wrong foods 4.2  0.6 3.83  0.74 3.83  0.71 3.8  1.0 3.74  0.92 3.84  0.98 9.4 0.62 0.02 0.002 0.42 0.88 Social Unemployment Low incomeb 2.65  1.02 2.67  0.96 2.53  1.21 2.36  1.2 0.74 5.03 0.38 0.025 a b Means and S.D. Signi®cant main effect of group. 231 J. Ogden et al. / Patient Education and Counseling 44 (2001) 227±233 Table 3 GPs and patients beliefs about consequences of obesitya Variables GPs (n ˆ 89) Patients (n ˆ 599) F-statistic P-value Medical Diabetesb Painful joints Heart disease High blood pressure 3.91 3.57 3.73 3.74 3.22 3.34 3.69 3.75 20.08 3.67 0.11 0.007 0.00001 0.055 0.7 0.9 Psychological Depression/anxiety Not feeling attractive Not feeling good about yourself 3.28  0.98 3.66  0.85 3.44  0.8 3.48  1.00 3.64  1.17 3.51  1.07 2.64 0.02 0.36 Social Difficulty making friends Difficulty getting workb Difficulty getting medical/surgical treatments 2.8  0.94 2.71  0.91 2.89  0.98 2.83  1.16 3.16  1.16 2.67  1.19 0.04 11.68 2.47 a b     0.79 0.87 0.82 0.85     1.11 1.07 1.11 1.02 0.1 0.8 0.55 0.83 0.0006 0.115 Means and S.D. Signi®cant main effect of group. Table 4 GPs and patients beliefs about solutions for obesitya Variables GPs (n ˆ 89) Patients (n ˆ 599) F-statistic P-value Methods used Slimming pills Exercise Calorie controlled diet Healthy eating Surgery 2.00 4.13 4.16 4.3 2.22      1.26 0.82 0.9 0.82 1.07 1.87 4.15 4.12 4.43 2.15      1.04 0.93 0.96 0.84 1.1 0.95 0.05 0.15 1.82 0.3 0.328 0.8 0.7 0.17 0.58 Professionals GPb Counsellorb Obese person themselvesb Dietician Practice nurse Friend/family 2.96 2.89 4.08 3.67 3.33 3.66       1.07 1.25 1.02 1.00 1.02 1.02 3.33 3.28 3.78 3.95 3.14 3.66       1.14 1.13 1.17 0.98 1.08 1.12 7.6 8.36 4.98 5.64 2.37 0.000 0.005 0.004 0.025 0.057 0.12 0.99 a b Means and S.D. Signi®cant main effect of group. showed no differences between the GPs and the patients for their ratings of sliming pills, exercise, calorie controlled diets, healthy eating and surgery. For the role of other people, the results showed that the two groups were similar in their ratings of the practice nurse, a dietician and the obese person's friends and family. However, whereas the GPs rated the obese person as more helpful, the patients rated the GP and a counsellor as more helpful. 4. Discussion The present study aimed to explore GPs' and patients' models of obesity. In terms of beliefs about the causes of obesity the results showed that the patients were more likely to attribute obesity to a gland/hormone problem, slow metabolism, stress and low income. For consequences, patients showed a higher rating for dif®culty getting work and for 232 J. Ogden et al. / Patient Education and Counseling 44 (2001) 227±233 solutions the patients rated the GP and a counsellor as more helpful. In contrast, the GPs were more likely to blame obesity on eating too much, rated diabetes as a more important consequence and rated the obese persons themselves as a more helpful solution. In line with previous research [9,10], these results provide insights into the ways in which patients and doctors make sense of illness and suggest that illness models can be characterised by dimensions of cause, consequence and solution. In addition, the results indicate that individuals may not always make consistent attributions for these different dimensions. Speci®cally, the patients in the present study made causal explanations which according to attribution theory [6,7] could be considered to be both internal to the individual and uncontrollable. In contrast, the same patients showed external attributions for the solutions to obesity. This provides support for the work of Brickman et al. [8] and indicates that individuals make different attributions for causes and solutions. In particular, it suggests that the patients showed a self serving and ego protective model of obesity by blaming internal factors for the etiology of obesity and external factors for its solution. In direct contrast to the patients, the doctors reported internal and controllable explanations for both causes and solutions. Such a model of obesity re¯ects a victim blaming approach as the individual is deemed to be responsible for both the cause and solution to their problem. This ®nds re¯ection in papers highlighting the emphasis on blame and control in the obese [17] and suggests that whereas this perspective may not be adhered to by patients it is re¯ected in the beliefs of doctors. In summary, these results, therefore, indicate considerable disagreement between doctors and patients and suggest that not only do doctors and patients have different models of the consultation [5], different models of health per se [15] and different models of depression [16], they also have different models of obesity. In particular, whereas the patients reported a self serving model of obesity, the GPs' analysis re¯ected a more victim blaming model of the problem. management in primary care. In particular, the results from this study suggest that whereas patients would prefer a professional based intervention, the GPs would prefer a patient based approach. This is particularly pertinent at a time when new drug therapies are being launched and marketed widely [22] and it could be predicted from the present study that patients would be more in favour of such solutions than GPs. This mismatch concerning how the doctor and patient make sense of obesity could create con¯ict over `best practice' in the consulting room. Secondly, the results also have implications for the effectiveness of any primary care based intervention and could provide an explanation of why the success of interventions for obesity are so poor [20,21]. Previous research has highlighted the importance of agreement, shared models and shared understanding in the doctor±patient relationship [1,2]. Further, it has suggested that disagreement may relate to poorer patient outcomes [3,4]. Therefore, as patients and GPs appear to disagree about the nature of obesity, if a management strategy were established it would seem unlikely that it would be a success. For, if they disagree about obesity, either the patient would be trying to comply to a strategy which did not concur with his/her model of obesity, or the GP would be recommending a strategy which they did not believe in. Accordingly, an improvement in communication and a potential increase in effectiveness in any subsequent intervention could be achieved by changing either the GPs' or the patients' models of obesity. Alternatively, GPs should be encouraged to tolerate the differences with their patients and agree to disagree. Acknowledgements The study was completed as part of the research methods module for the MSc. in general practice at Kings College London. References 5. Implications for practice Such a mismatch between GPs and patients primarily has implications for the form of obesity [1] Pendleton D, Scho®eld T, Tate P, Havelock P. The consultation: an approach to learning and teaching. Oxford: Oxford Medical Publications, 1984. [2] Neighbour R. The inner consultation. Reading: Petroc Press, 1987. J. Ogden et al. / Patient Education and Counseling 44 (2001) 227±233 [3] Star®eld B, Hess K, Gross R, Birk PS, D'Lugoff BC. The in¯uence of patient±practitioner agreement on outcome of care. Am J Pub Health 1981;71:127±31. [4] Bass MJ, Buck C, Turner C, Dickie G, Pratt G, Robinson HC. The physicians actions and the outcome of illness in family practice. J Fam Pract 1986;23:43±7. [5] Krupat E, Rosenkranz SL, Yeager CM, Barnard K, Putman SM, Inui TS. The practice orientations of physicians and patients: the effect of doctor patient congruence on satisfaction. Patient Edu Couns 2000;39:49±59. [6] Heider F. Social perception and phenomenal causality. Psychol Rev 1944;51:358±74. [7] Kelley HH. Attribution in social interaction. In: Jones EE, et al., editors. Attribution: perceiving the causes of behaviour. Morristown (NJ): General Learning Press, 1972. [8] Brickman P, Rabinowitz VC, Karuza J, Coates D, Cohn E, Kidder L. Models of helping and coping. Am Psychol 1982;37:368±84. [9] Leventhal H, Nerenz D. The assessment of illness cognition. In: Karoly P, editor. Measurement strategies in health psychology. New York: Wiley, 1985. p. 517±54. [10] Leventhal H, Benyamini Y, Brownlee S, Diefenbach M, Leventhal EA, Patrick Miller L, Robitaille C. Illness representations: theoretical foundations. In: Petrie KJ, Weinman JA, editors. Perceptions of health and illness. Amsterdam: Harwood, 1997. p. 1±18. [11] Radley A. The role of metaphor in adjustment to chronic illness. Worlds of illness: biographical and cultural perspectives on health and disease. London: Routledge, 1995 [chapter 6]. 233 [12] Schou KC, Hewison J. Health psychology and discourse: personal accounts as social texts in grounded theory. J Health Psychol 1998;3(3):297±312. [13] Crossley M. Sick role or empowerment: the ambiguities of life with an HIV positive diagnosis. SHI 1998;20:507±31. [14] Tuckett D, Boulton M, Olson C, Williams A. Meetings between experts. London: Tavistock Publications, 1985. [15] Ogden J, et al. What is health?: where GPs and patients worlds collide. Pat Educ Couns 2001, in press. [16] Ogden J, et al. `You're depressed, no I'm not': GPs and patients different models of depression. Br J Geb Pract 1999;49:123±4. [17] Brownell KD. Personal responsibility and control over our health: when expectation exceeds reality. Health Psychol 1991;10:303±10. [18] Ogden J. Restraint theory and its implications for obesity treatment. Clin Psychol Psychotherapy 1994;1:191±201. [19] Ogden J, Awal M. Obesity stereotypes: the role of body size and eating behaviour. Submitted for publication. [20] Wadden TA. Treatment of obesity by moderate and severe calorie restriction: results of clinical research trials. Ann Int Med 1993;119:688±93. [21] NHS Centre for Reviews and Dissemination, University of York. Systematic review of interventions in the treatment and prevention of obesity, 1997. [22] Sjostrom L, Rissanen A, Andersen T, Boldrin M, Golay A, Koppeschaar HPF, Krempf M. Ramdomised placebo controlled trial of orlistat for weight loss and prevention of weight regain in obese patients. The Lancet 1998;352:167±72.