|Home | About | Journals | Submit | Contact Us | Français|
Clinician counseling is a catalyst for lifestyle modification in obesity. Unfortunately, clinicians do not appropriately counsel all obese patients about lifestyle modification. The extent of disparities in clinician counseling is not well understood. Obese participants (BMI ≥30 kg/m2, N = 2097) in the Dallas Heart Study (DHS), a probability-based sample of Dallas County residents ages 18–65, were surveyed regarding health-care utilization and lifestyle counseling over the year prior to DHS enrollment. Health-care utilization and counseling were compared between obese participants across three categories based on the presence of 0, 1, or 2+ of the following cardiovascular (CV) risk factors: hypertension, hypercholesterolemia, or diabetes. Logistic regression modeling was used to determine likelihood of counseling in those with 0 vs. 1+ CV risk factors, stratified by race, adjusting for age, sex, insurance status, and education. Among obese subjects who sought medical care, those with 0 CV risk factors, compared to those with 1 or 2+ CV risk factors, were less likely to report counseling about losing weight (41% vs. 67% vs. 87%, P trend <0.001), dietary changes (44% vs. 71% vs. 85%, P trend <0.001), and physical activity (46% vs. 71% vs. 86%, P trend <0.001). Blacks and Hispanics without CV risk factors had a lower odds of receiving counseling than whites without risk factors on weight loss (adjusted odds ratio (OR), 95% confidence interval (CI) for nonwhites 0.19, [0.13–0.28], whites 0.48, [0.26–0.87]); dietary changes (nonwhites 0.19, [0.13–0.27], whites 0.37, [0.21–0.64]); and physical activity (nonwhites 0.22, [0.16–0.32], whites 0.32, [0.18–0.57]). Lifestyle counseling rates by clinicians are suboptimal among obese patients without CV risk factors, especially blacks and Hispanics. Systematic education about and application of lifestyle interventions could capitalize on opportunities for primary CV risk prevention.
Given the high prevalence of obesity in the United States (1,2) and the strong association between obesity and an increased lifetime risk of cardiovascular events (3–7), interventions aimed at slowing or reversing the obesity epidemic are vitally important to the public’s health. One such intervention is health professional counseling of patients regarding lifestyle modification, including weight loss, dietary changes, and increasing physical activity. Among obese individuals, clinician counseling has been associated with motivation to lose weight, attempted weight loss, and a better understanding of the adverse health effects of obesity (8–13). Current guidelines from the National Institutes of Health, the American Heart Association, the American College of Cardiology Foundation, and the American Diabetes Association consistently advocate physician counseling on lifestyle modification for the treatment of obesity (14–17). In addition, the US Preventive Services Task Force guidelines recommend that clinicians screen for obesity and intensively counsel all obese adults about weight loss using dietary changes and/or physical activity (18).
Unfortunately, data from population-based cohorts suggest that health-care professionals do not appropriately counsel all obese patients about lifestyle modification (8,11,19–21). Time constraints during visits, inadequate training on counseling, or lack of reimbursement are several barriers to weight loss counseling and potential explanations for the lack of consistency in health professionals’ counseling of obese patients about lifestyle changes (22,23). However, the extent of disparities in health professional counseling for those with obesity, particularly in the absence of cardiovascular (CV) risk factors, is not well understood. Moreover, despite the disproportionate impact of obesity on ethnic minority populations (24), racial disparities in health professional counseling about lifestyle modification for obese individuals without obesity-related CV risk factors have not been fully characterized. Prior findings have been inconsistent with studies suggesting that obese blacks or nonwhites have a similar (8,11,20) or higher (25–27) likelihood of reported counseling on lifestyle changes as compared to whites. However, many of these studies have not accounted for the higher prevalence of obesity-related CV risk factors that may influence counseling on lifestyle modification. Understanding racial disparities in counseling may identify ways to combat the long-standing, disproportionately higher prevalence of obesity among blacks and Hispanics.
To better understand the pattern of health professional counseling regarding obesity in a multi-ethnic urban cohort at high lifetime risk for obesity-related complications, we studied obese participants in the Dallas Heart Study (DHS). We characterized the pattern of health professional counseling on lifestyle modification among obese individuals with and without CV risk factors and explored the influence of CV risk factors on health professional counseling when accounting for other potential confounders of counseling such as an individual’s age, race, sex, education, and insurance status.
The DHS is a multiethnic, probability-based population sample of Dallas County adults ages 18–65 (N = 6,101) designed to study cardiovascular disease. Blacks were intentionally over-sampled to comprise 50% of the study cohort. Details of the DHS study design and cohort have been previously reported (28). Visit 1 for DHS involved a home visit for collection of demographic and survey data and measurement of anthropometrics. Visit 2 (N = 3,398) involved collection of fasting blood and urine samples. The University of Texas Southwestern Medical Center at Dallas institutional review board approved this study, and all subjects provided written informed consent.
Demographic information, including race/ethnicity, household income, highest level of achieved education, health insurance status, and medical histories, were determined by self-report. BMI was calculated based on measured height and weight at study entry. Obesity was defined according to National Heart, Lung, and Blood Institute criteria (14) as a BMI greater than or equal to 30 kg/m2. Hypertension was defined as one of the following: systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg, or the use of anti-hypertensive medication. Hypercholesterolemia was defined either by self-report, use of lipid-lowering medication or by a fasting low-density lipoprotein ≥160 mg/dl. Diabetes mellitus was defined either by self-report, use of anti-hyperglycemic medication or by fasting serum glucose ≥126 mg/dl. Questions abstracted from the 1999 Behavioral Risk Factor Surveillance System were incorporated into the DHS survey instrument to assess beliefs about general health perceptions and health-care access (29). Physician counseling on lifestyle modification was surveyed in the questionnaire using the following yes/no questions: (i) ‘Has a health professional talked with you about changing dietary or eating habits?’; (ii) ‘Has a health professional talked with you about physical activity or exercise?’; (iii) ‘Has a health professional talked with you about losing weight?’.
Obese individuals in the DHS, defined as those having a BMI ≥30 kg/m2, were stratified based on the presence or absence of prevalent hypertension, hypercholesterolemia, or diabetes, and were divided into those with 1 and 2+ of these risk factors (with CV risk factors) and those with none (without CV risk factors). Of the 2,097 obese subjects from the DHS, we excluded subjects missing prevalence data on hypertension (N = 43), diabetes (N = 1), or hypercholesterolemia (N = 32), those with ethnicity other than black, white, or Hispanic (N = 16), those who had not seen a health-care professional in the past 12 months (N = 572), leaving a final sample size of 1,433 subjects for analysis.
Baseline characteristics for obese DHS participants with 0, 1, and 2+ CV risk factors were compared using the Jonckheere–Terpstra test for categorical variables and for continuous variables as appropriate. Two-sided P values ≤0.05 were considered statistically significant. Proportions of participants answering specific questions about general health perceptions, perceived lifetime risk of disease, and utilization of health-care resources were compared using the Jonckheere–Terpstra or χ2 test. Comparison of reported physician counseling on weight loss, dietary changes, and physical activity between obese subjects with vs. without any of the previously mentioned CV risk factors was done using multivariable logistic regression modeling. These models were adjusted for race, age, sex, achieved education, and insurance status. A statistical interaction between race and risk factor status was also evaluated and race was stratified into white and nonwhite, which comprised blacks and Hispanics, in these analyses. Odd ratios for the likelihood of health professional counseling were compared between racial/ethnic groups using the Lagakos method (30). All analyses were performed using Statistical Analysis Systems software, version 9.2 (SAS Institute, Cary, NC).
In the entire cohort of obese subjects seen by a health professional at least once in the last 12 months, 60% had at least one CV risk factor, as shown in Table 1. Compared to obese subjects with 1 or 2+ CV risk factors, obese subjects with 0 CV risk factors were younger (mean age 36 vs. 45 vs. 50 years, P trend <0.001), more likely to be Hispanic (22% vs. 11% vs. 11%, P trend <0.001), and had a modestly smaller mean BMI (36.2 vs. 36.8 vs. 37.2 kg/m2, P trend = 0.02), waist circumference (110.5 vs. 112.9 cm vs. 117.0, P trend <0.001), and waist-to-hip ratio (0.90 vs. 0.92 vs. 0.95, P < 0.001). Of obese participants with one CV risk factor, 64% had hypertension, 9% had diabetes, and 28% had hyperlipidemia. Of participants with 2+ risk factors, 96% had hypertension, 78% had hypercholesterolemia, and 51% had diabetes. Regarding socioeconomic status, there were no significant differences in the proportion of participants with 0 vs. 1 or 2+ CV risk factors in each achieved education strata. However, family income at or above $50,000 was significantly less likely among those with 0 CV risk factors (P trend = 0.02). As shown in Table 2, 20% of obese participants with 0 CV risk factors reported lack of health insurance as a barrier to health care, while 18% of those with one CV risk factor and 14% of those with 2+ CV risk factors reporting this barrier (P trend = 0.03). There were no differences across the three groups in the number of participants who reported high cost as a barrier to accessing health care. In addition, trust of health-care professionals did not differ significantly across the three groups.
Reported health professional counseling about lifestyle modification among obese participants is shown in Table 2. In the overall study population, 62% of participants reported counseling about weight loss, 64% reported counseling on dietary changes, and 65% reported counseling on physical activity. Obese subjects without CV risk factors were significantly less likely than those with one CV risk factor or those with 2+ risk factors to report counseling about weight loss (41% vs. 67% vs. 87%, P trend <0.001), dietary changes (44% vs. 71% vs. 85%, P trend <0.001), and physical activity (46% vs. 71% vs. 86%, P trend < 0.001).
Figure 1 illustrates the adjusted odds ratios for reported physician counseling on weight loss, dietary changes, and physical activity in obese individuals with 0 CV risk factors vs. 1+ CV risk factors. There was a statistically significant interaction between race and risk factor status for counseling on weight loss (P interaction for race × risk factor status <0.05); therefore, results from the multivariable logistic regression models were stratified by race. Whites with obesity and no CV risk factors had as much as a 68% lower likelihood of receiving counseling on lifestyle modification as compared to obese whites with CV risk factors (adjusted OR, (95% confidence interval (CI) for obese whites without CV risk factors: counseling on weight loss 0.48, [0.26, 0.87]; counseling on dietary changes 0.37, [0.21, 0.64]; counseling on physical activity 0.32, [0.18, 0.57]). In contrast, obese nonwhites without risk factors were up to 81% less likely to receive counseling on weight loss, dietary changes, or physical activity as compared to nonwhites with CV risk factors (adjusted OR, [95% CI] for obese nonwhites without CV risk factors: counseling on weight loss 0.19, [0.13, 0.28]; counseling on dietary changes 0.19, [0.13, 0.27]; counseling on physical activity 0.22, [0.16, 0.32]).
When comparing specific racial/ethnic groups (Table 3), obese blacks without risk factors were up to two times less likely than obese whites without CV risk factors to receive physician counseling in each area of lifestyle modification (adjusted OR, [95% CI] for obese blacks without CV risk factors: counseling on weight loss 0.21, [0.14, 0.32], P < 0.05 when compared to whites; counseling on dietary changes 0.21, [0.14, 0.31]; counseling on physical activity 0.24, [0.16, 0.35]). Across racial/ethnic groups, Hispanics were the least likely to report physician counseling. Hispanics without CV risk factors had more than a 50% lower likelihood than blacks without risk factors of receiving physician counseling on lifestyle changes (adjusted OR, [95% CI] for obese Hispanics without CV risk factors: counseling on weight loss 0.14, [0.05, 0.40]; counseling on dietary changes 0.11, [0.04, 0.30]; counseling on physical activity 0.15, [0.06, 0.41]). Hispanics without CV risk factors had a two- or three-fold lower likelihood of reporting counseling than whites without risk factors. These findings persisted in sensitivity analyses with BMI added to the adjusted models stratified by race (data not shown). As shown in Figure 2, obese individuals without CV risk factors were also more likely to perceive a low lifetime risk of CV events and related diseases, with 60% perceiving a low lifetime risk of myocardial infarction, 59% perceiving a low lifetime risk of diabetes, and 50% perceiving a low lifetime risk of hypertension as compared to those with risk factors (P < 0.05 for all groups).
In summary, these data from a large, contemporary population sample illustrate that therapeutic lifestyle counseling remains suboptimally applied by health-care providers, with disparities in application most notable among (i) those without prevalent hypertension, hyperlipidemia, or diabetes, and (ii) ethnic minorities. These disparities in counseling about lifestyle modification demonstrate that health-care providers commonly fail to capitalize on an important opportunity for primary prevention through lifestyle intervention to impact the development of obesity-related morbidity, especially in black and Hispanic populations at high risk of developing cardiovascular disease (31). The progressive increase in reported counseling on weight loss, dietary changes, and physical activity with an increasing number of prevalent CV risk factors seen in the DHS population supports the concept that comorbidities, as opposed to the presence of obesity, serve as the impetus for counseling on lifestyle modification. Biases that direct health professional counseling towards those with existing hypertension, hyperlipidemia or diabetes may contribute to a lack of awareness among obese individuals without CV risk factors that obesity itself is a risk factor for cardiovascular morbidity, potentially making them more likely to perceive themselves with a low lifetime risk of cardiovascular disease.
Our study suggests that counseling on lifestyle modification is substantially driven by the presence of cardiac risk factors, rather than by obesity itself, despite guideline recommendations that support lifestyle counseling in all obese individuals (14,18). Data from the 1996 Behavioral Risk Factor Surveillance System (BRFSS) in over 12,000 obese subjects demonstrate that only 42% of obese individuals who visited a physician in the last 12 months reported counseling about weight loss. In this study, subjects with diabetes were more likely to report counseling than those without, suggesting bias towards preferentially counseling obese individuals with existing risk factors (8). Further analyses of the 1996 BRFSS data demonstrated that while the presence of a weight-related comorbidity predicted whether an individual reported counseling, only 47% of those with a BMI ≥30 kg/m2 and with hypertension, diabetes, or hypercholesterolemia reported counseling to lose weight (19). However, these data were collected prior to the publication of guidelines for counseling in obesity by the National Institutes of Health in 1998 or by the US Preventive Services Task Force in 2003 (14,18).
Unfortunately, establishing guidelines does not appear to have meaningfully improved physician counseling rates. Despite the dramatic increase in obesity prevalence over the past decades, studies of more contemporary cohorts demonstrate that less than half of obese individuals, even among those with BMI at or above 40 kg/m2, report receiving specific advice on weight loss, with prevalent diabetes and hypertension serving as independent predictors of reported counseling (32). In addition, comparison of trends in counseling from samples of the US population in the BRFSS have shown a concerning decline in reported weight loss counseling from 1994 to 2000 (11), with only 44% of obese adults from 11 states reporting counseling in 2005 (27). In the DHS, we found that 62% of obese individuals seen by a physician in the past year reported physician counseling to lose weight. The higher percentage of individuals reporting counseling on weight loss in the DHS population compared to random samples of the US population likely reflects a higher prevalence of diabetes, hypertension or hyperlipidemia in the DHS population compared to the previously sampled populations (32), regional variation in medical practice, and differences in data collection between the studies. These findings suggest that counseling to lose weight remains underutilized in the treatment of obesity, and interventions should be developed to encourage physicians to counsel all obese individuals about lifestyle modification based on established guidelines, regardless of the absence of comorbidities.
Prior studies have not recognized race/ethnicity as an important determinant of whether obese patients report counseling for lifestyle changes, with several studies showing that race is not an independent predictor of lifestyle counseling (8,11,19–21,32). Others have reported higher rates of counseling on lifestyle modification among ethnic minority populations (9,25–27,33), although individuals with obesity were not the primary focus of all of these studies (25–27,33). Most have not taken into account that blacks and Hispanics may have more prevalent factors that lead to counseling (8,9,19–21,32). Among women in the 2000 BRFSS survey, Mack et al. identified that 44% of black women with obesity reported counseling to lose weight as compared to 40% of obese white women and 43% of obese Hispanic women, suggesting that obese black and Hispanic women were actually more likely to be counseled than were white women (9). However, black and Hispanic women were also more likely to have diabetes in this cohort, which likely drove the higher rates of weight loss counseling. Stratification of the DHS cohort by race/ethnicity allows us to ascertain that, while those with CV risk factors are counseled similarly across racial/ethnic groups, black and Hispanic men and women without risk factors are less likely to report counseling. Given that blacks and Hispanics are more likely to be obese (2), and, therefore, more likely to suffer the cardiovascular sequelae of this disease (31), it is imperative that lifestyle modification be promoted in these most vulnerable populations even before factors develop. This distinction also provides valuable insight into groups that might benefit from focused interventions to better promote lifestyle modification.
More frequent counseling on lifestyle modification, particularly among blacks and Hispanics with obesity, may lead these high risk subgroups to a better understanding of the health effects of obesity. Data from overweight and obese individuals from United States community health centers suggest that while blacks and Hispanics were less likely to perceive excess weight as damaging to health as compared to whites, blacks and Hispanics who were counseled about being overweight or obese by their provider were four-fold and over ten-fold, respectively, more likely to perceive that excess weight impacted health (12). Another analysis of the 2000 BRFSS data revealed that advice to lose weight is an independent predictor of attempted weight loss among obese individuals (10). Simple acknowledgement of weight status and lifestyle counseling appear to be associated with a more realistic perception of the effects of obesity and motivation for weight loss (13,19,27); therefore, improvements in counseling may prove beneficial for an obese population without existing CV risk factors.
A key strength of our study is the large, diverse population-based study cohort of the DHS, which fosters the extrapolation of these findings to other contemporary multi-ethnic urban populations. In addition, we used measured height and weight, allowing a more accurate representation of the obese population in DHS, in contrast to a number of older studies that relied on survey data of self-reported height and weight to estimate BMI (8,9,11). More recent studies of health professional counseling on lifestyle changes, however, are beginning to use measured height and weight to calculate BMI (21,32). The limitations of this study must also be considered. These data are cross-sectional and, therefore, we cannot infer causality in our findings. Reported counseling cannot be independently and objectively verified from medical records; therefore, we are unable to know if DHS participants actually received lifestyle counseling. Health-care professionals’ characteristics, such as physician specialty, are not available for the participants who reported counseling, and these unmeasured factors may explain some variability in counseling. In addition, limited English proficiency may serve as a barrier and a potential confounder for lifestyle counseling among Hispanic participants; however, prior data from the National Health Interview Survey suggest that English proficient Hispanics have higher rates of obesity than Hispanics with limited English (34). Finally, reported counseling data are subject to recall bias, and may not precisely reflect actual rates of counseling for all subjects. However, those who cannot recall counseling may be similar to those who were never counseled at all, as counseling on lifestyle modification is unlikely to change health beliefs and behaviors for those who do not remember its occurrence.
In conclusion, health professional counseling for lifestyle modification is underutilized in all obese individuals without cardiovascular risk factors, particularly among racial/ethnic minority populations. Education efforts to improve health professional counseling on lifestyle modification must focus on both clinicians and patients, encouraging health-care providers to counsel all obese patients about lifestyle modification as current guidelines recommend. Patients must be better educated about obesity as a cardiovascular risk factor and its impact on lifetime CV risk. Health professional counseling better targeted at obese individuals without other CV risk factors may help reduce persistent disparities in obesity prevalence and related CV comorbidities among blacks and Hispanics.
Funding support for Dr. Powell was provided by the Ruth Kirschstein National Research Service Award from the National Institutes of Health (#2-T32-HL007360-31). Funding support for Dr. Das was provided by an American College of Cardiology Foundation/GE Healthcare Career Development Award. Funding support for the Dallas Heart Study was provided by the Donald W. Reynolds Foundation (Las Vegas, NV) and the United States Public Health Service General Clinical Research Center grant #MO1-RR00633 from the National Institutes of Health/National Center Research Resources-Clinical Research.
The authors declared no conflict of interest.