Are Obese Patients Assisted in Losing Weight?
Published Online: April 17, 2014
Armina Sepehri, MPH; Vicente F. Gil-Guillén, MD, PhD; Antonio Palazón-Bru, MPH; Domingo Orozco-Beltrán, MD, PhD; Concepción Carratalá-Munuera, PhD; Ernesto Cortés Castell, PhD; and Mercedes Rizo-Baeza, PhD
Obesity is a very common disorder worldwide.1 In addition, obesity is associated with such healthcare problems as diabetes mellitus, hypertension, dyslipidaemia, and cardiovascular diseases (CVD).2 Clinical practice guidelines recommend that healthcare professionals intervene to reduce the prevalence of this problem. These professionals should help those patients who need to lose weight, by means of personalized counselling about a healthy lifestyle (food and physical exercise).3-6
The Valencian Community is situated in the Mediterranean area of eastern Spain, and has a population of 4,518,126 inhabitants (figures from January 2004).7 Primary healthcare is given at health centers, and is universal and free for patients. The patients who attend these health centers are mainly women of older age with cardiovascular risk factors (CVRF), and they are frequent visitors.8 In this community, the Valencia study analyzed the impact of obesity in the population. From 1991 to 2005, the prevalence of obesity rose from 7.3% to 12.4%, and was the most important problem for men aged 34 years and older and women aged 50 years and older. Also of note was that these patients had other CVRFs.9 In Spain, the health costs associated with obesity account for 7% of total healthcare costs. Over 35% of those costs correspond to obesity-related diseases such as CVD, diabetes mellitus, and dyslipidemia. The remaining 65% of costs are due to different types of cancer, kidney and liver disorders, sleep apnea, and even urinary incontinence—all related to obesity.10 Notable is the yield of bariatric surgery for both health and cost-benefit reasons, both in Spain and in other countries.11,12
A program of preventive activities was started in this Community at the end of 2003, aimed at the whole population over age 40 years. Each person was invited by mail to participate, and then contacted by phone to arrange an appointment at their health center. There, they underwent a preventive examination by medical and nursing personnel, and were given a report with the result of the examination together with the opportune recommendations; a copy of this report was also left at the health center. This program followed the recommendations of the Programme for Prevention and Health Promotion (PPHP) of the Spanish Society of Family and Community Medicine.3,13-15 Briefly, this program contains cardiovascular screening (hypertension, diabetes mellitus, dyslipidemia, smoking, obesity, etc), gynecological screening (cytology, mammography, etc), and a vaccination campaign (flu, tetanus, and pneumococcus).
Phillips et al in 2001 defined clinical inertia as failure by the physician to start or intensify treatment when this was indicated.16 A few years later, Andrade et al defined the concept of therapeutic inertia.17 Reflection about the definition of these concepts suggests that inertia not only influences the therapeutic process, but may also affect other parts of the clinical care process, such as personalized counseling about healthy lifestyle habits for those patients who need it.
As part of the preventive activities program, this study analyzed inertia associated with advising obese patients about a healthy lifestyle in order to lose weight, together with the possible associated factors. Others have also analyzed this behavior,18-26 assessing the advice and its association with a history of cardiovascular risk factors. We therefore wondered whether the healthcare professionals were paying more attention to already diagnosed cardiovascular risk factors rather than considering the current status of the patients. Accordingly, we calculated the cardiovascular risk of these patients using the REGICOR score,27 which is a calibration of one of the scales from the Framingham study designed for the Spanish population,28 and determined the association of this risk with the behavior of the healthcare professionals when aiding an obese patient to lose weight. The need for measures to improve the counselling of obese patients about weight loss can be seen from the results.
Design and Study Subjects
This cross-sectional study involved a sample of obese persons over age 40 years who participated in the preventive activities program of the Valencian Community during its first 6 months, and who wished to collaborate. Patients were considered to be obese if their body mass index (BMI) was at least 30 kg/m2. Any patient who was not obese, according to this definition, was excluded from this study.
Variables and Measurements
All the cardiovascular information recorded at the health examinations was studied. The main outcome measure was obesity inertia (OI). A patient was considered to have experienced OI if that patient’s healthcare professional failed to provide personalized advice about both diet and exercise together as a means to lose weight. The healthcare professional also recorded the following variables: gender; personal history of hypertension, dyslipidaemia, diabetes mellitus, smoking, acute myocardial infarction, and stroke; BMI (in kg/m2); age (in years); blood pressure (BP) (systolic [SBP] and diastolic [DBP] in mm Hg); total cholesterol and high-density lipoprotein (HDL) cholesterol (in mmol/L).
In order to calculate BMI, the weight and height were measured with a calibrated scale and stadiometer, removing all objects that could affect the weight, including shoes. BP was measured following current recommendations with well-calibrated semiautomatic aneroid devices (mercury) in adequate conditions. The lipid profile was measured first thing in the morning after a minimum 8-hour fast with calibrated equipment. The personal history of disease, gender, and age was obtained during the patient interview and corroborated from the clinical records.
After gathering all the data, the following groups of variables were made: (1) BMI groups according to the World Health Organization (WHO) classification: Class I obesity (BMI ≥30 kg/m2 and <35 kg/m2) and Class II and III obesity (BMI ≥35 kg/m2)29; and (2) personal history of CVD, or having had an acute myocardial infarction or stroke.
After collecting and grouping the variables, the REGICOR cardiovascular risk was defined (Registre Gironí del Cor) in those patients for whom it was applicable. These patients were then classified in risk groups27: high (≥20%) and low (<20%). This scale is an adaptation of the Wilson scale for the Spanish population and it estimates the risk of having a coronary event in the next 10 years in patients aged 30 to 74 years who have not had any prior CVD.28 The predictive variables on this scale are gender, age, total cholesterol, HDL cholesterol, BP (SBP and DBP), diabetes mellitus, and smoking.
There were no missing data, as the healthcare professionals took particular care to complete the whole preventive activities program in all the participants.
The overall sample size was 8687 patients with obesity. Of these, 7700 fulfilled the criteria necessary to be evaluated with the REGICOR (primary cardiovascular prevention and age <75 years). Thus, using a significance level of 5% and a maximum expected proportion (P = q = .50), the expected error in the estimation of OI was 1.05% in the overall sample and 1.12% in the patients whose REGICOR was calculated.
Absolute and relative frequencies were used to describe the qualitative variables, whereas means and standard deviations were used for the quantitative variables. Multivariate logistic regression models were calculated to estimate the adjusted odds ratios (ORs) in order to analyze the relation between OI and the study variables. For the overall sample, the ORs were adjusted for gender; personal history of hypertension, dyslipidemia, diabetes mellitus, smoking, and CVD; BMI; and age groups. For the REGICOR sample, the ORs were adjusted in 2 ways: (1) REGICOR risk group and BMI as a quantitative variable; and (2) REGICOR risk group and BMI group. The likelihood of OI in the multivariate models was used to create figures to help interpret the results. The likelihood ratio test was carried out for the goodness-of-fit of the models. All analyses were performed at a 5% significance level and associated confidence intervals (CIs) were estimated for each relevant parameter. All of the analyses were performed using SPSS 19 (IBM, Armonk, New York).
The first statistical analysis done (overall sample) was similar to that done by others,18‑26 mainly prioritizing the personal history of CVRF. The second analysis (REGICOR sample) was an innovative examination of the association between inertia and cardiovascular risk. Thus, we assessed the BMI both qualitatively and quantitatively. Though this increases the complexity of the paper, both forms provide clinically relevant information.
This study was approved by an institutional review board of the Valencian Community, permitting data analysis and complying with current legislation on medical ethics. This institution had no role in data collection, analysis, or interpretation; nor did it have the right to approve or disapprove publication of the finished manuscript. Furthermore, the data were anonymized and encrypted, satisfying the data protection law.
Table 1 summarizes the information concerning the overall sample (n = 8687). Most of those who participated in the study were women; there was a high prevalence of CVRF (over 7% had CVD); and the immense majority of patients had a BMI associated with Class I obesity.29
The magnitude of OI was 16.6% (95% CI, 15.8-17.4). Factors associated with OI were being male (OR = 1.19; 95% CI, 1.06-1.35); no personal history of hypertension (OR = 0.85; 95% CI, 0.74-0.97), or dyslipidemia (OR = 0.86; 95% CI, 0.73-1.01), or diabetes mellitus (OR = 0.80; 95% CI, 0.64-1.00), or CVD (OR = 0.79; 95% CI, 0.62-1.01); and having a BMI representing Class I obesity (OR = 0.83; 95% CI, 0.72-0.96).
Table 2 shows the results in the REGICOR sample (n = 7700). A small proportion of persons had a high risk according to the REGICOR (0.7%). Their characteristics were very similar to those of the overall sample (gender, age, personal history of diseases, and smoking), with a mean blood pressure representing prehypertension (134/81 mm Hg),30 and mean total and HDL cholesterol levels much higher than normal (5.5 mmol/L and 1.6 mmol/L, respectively).31 The proportion of OI was 16.9% (95% CI, 16.0-17.7). The associated factors, after adjusting for BMI as a quantitative variable, were a high REGICOR (OR = 2.27’ 95% CI, 1.30-3.99) and a low BMI (OR = 0.98; 95% CI, 0.96-1.00), whereas in the model adjusted for BMI group the associated factors were a high REGICOR (OR = 2.27; 95% CI, 1.30-3.98) and a borderline BMI representing Class I obesity (OR = 0.82; 95% CI, 0.71- 0.95). In the model with the BMI as a quantitative variable a Cartesian chart (the Figure) was designed showing the following elements: BMI on the X axis, likelihood of OI on the Y axis, and REGICOR risk groups by symbols (crosses and circles). This chart shows that persons with a high risk have a greater likelihood of experiencing OI, and the greater the BMI the lower the likelihood of experiencing OI.
The first analysis of the results of this study, considering the whole sample with obesity, shows that OI occurs in about 1 in 6 patients. A search of the literature showed studies evaluating advice about losing weight in obese patients. However, these studies diff er concerning both the type of population and the mode of evaluation (eg, some only evaluate physical exercise or diet whereas others evaluate both). The rate of OI found in these studies ranged from 35 to 63%.18-27 In our study, though, the rate of OI was much lower, possibly because our study took place during a preventive campaign.
PDF is available on the last page.