Addressing Bias Toward Overweight Patients: a Training Program for First-Year Medical Students

Abstract

Background

Physicians may have biases toward overweight patients which likely influences clinical judgments and can lead to disparities in patient care. An increasing number of adults are considered overweight/obese, so it is important to address these biases in training future physicians.

Methods

Forty-five first-year medical students participated in art museum programs and physician presentations, or were part of the control group. Four validated measures Beliefs About Obese Persons Scale, Attitudes Toward Obese Persons Scale, Fat Phobia Scale, and the Harvard Implicit Association Test (IAT) and researcher-generated questions, measured levels of bias before and after study activities.

Results

All participants demonstrated decreased bias. ANCOVA analysis did not reveal significant differences between the experimental and control groups. However, prior to the study 75% of participants had “preference for thin individuals.” Forty percent of those participating in study activities indicated a positive change by associating more positive traits with obese body shapes, compared to 29% of the control group. Study activities were rated positively.

Discussion/Conclusion

The art museum was an engaging/relaxing place for reflection on body types and biases. Physicians provided important instruction for normalization/de-stigmatization of patient care. Although there were no significant findings, the study has raised questions for continuing this work. What are most effective ways/times to address weight bias within the medical school curriculum? Could this work extend to other marginalized patient groups? The diversity in art and humanities creates a rich resource for discussing viewpoints and experiences. The small number of participants and the timing/lack of focus in museum sessions are noted as limitations.

Background

Recent statistics from the Centers for Disease Control and Prevention (CDC) indicate that 42.4% of American adults are considered obese [1] and the World Health Organization (WHO) estimates that over 1.9 billion adults worldwide are considered overweight1 [2]. The rising rates of obesity have been referred to as “the obesity epidemic,” and accepted as a significant social problem [3]. Obesity is linked to many chronic conditions, e.g., heart disease, diabetes, and many forms of cancer, leading to an unprecedented need for healthcare to patients who are obese [1]. It has been shown that physicians and other healthcare providers have bias towards patients who are overweight [4–6]. The effect of this bias, similar to other biases, e.g., race and gender, influences clinical judgments and leads to disparity in patient care [7–10]. Significantly, Phelan et al. (2014) found that medical students exhibited greater bias toward obese people than toward any other stigmatized minority group [11]. Reducing bias toward overweight patients is essential to assuring effective universal healthcare. Medical school has been identified as a critical time when future physicians appear susceptible to worsening explicit (conscious) bias and diminished empathy towards patients struggling with obesity [12]. It appears that students may not even be aware of their own negative attitudes toward obesity [13], and this lack of awareness can also have a negative impact on the care provided to future patients.

The Liaison Committee on Medical Education (LCME), the accrediting agency for medical schools in the USA, through a series of standards, delineates what must be addressed in all medical school curricula. Two standards are relevant and form the basis for this study: Standard 7.5 states that medical schools must “…ensure that the medical curriculum includes instruction in the diagnosis, prevention.., and treatment of the medical consequences of common societal problems,” and Standard 7.6 refers to Cultural Competence and Health Care Disparities, indicating that schools must provide opportunities “…for medical students to learn to recognize and appropriately address…cultural biases in themselves, in others and in the health care delivery process.” Despite this requirement, there are limited resources available to aid healthcare providers in understanding bias and the needs of patients who are overweight [14, 15] and none could be located that were specifically developed for medical students. The UConn Rudd Center for Food Policy & Obesity developed a video to train healthcare professionals, and recently partnered with the Obesity Action Coalition to create brochures and guides to educate the public, including one to understand weight bias in healthcare settings [16]. The Rudd Center video depicts an uncomfortable interaction between a doctor and an obese patient, provides information about the issue of weight bias, and discusses the complex causes of obesity. This video has been used for training with medical students [14], and was selected for training in the current study. Medical school curricular efforts to reduce weight bias have also focused on perspective-taking [5], education about causes of obesity, e.g., genetics, economic status, living environment [17], positive role-modeling by physicians and peers [18], and greater understanding of a patient’s living situation [19]. The efficacy of the interventions has been variable, and we recognized that further investigation was needed to identify effective strategies to reduce weight bias among medical students.

There has been strong advocacy for the inclusion of humanities education in the medical school curriculum. A recently published monograph (2020) by the Association of American Medical Colleges (AAMC) details steps to integrate the arts into medical education [20], and many medical schools have implemented a variety of humanities programs2. Recently, Harvard Macy announced an art museum–based health profession education fellowship [21]. The efficacy of humanities-based training for medical students has been proven to enhance teamwork [22, 23], to improve clinical observation skills [24–26], and to improve empathy and reduce burnout [27]. One study by Matharu et al. (2014) addressed the issue of weight bias with medical students using dramatic readings written by women who recounted their own experiences with weight prejudice, e.g., bullying and job discrimination [28]. Students who participated in the dramatic readings demonstrated a statistically significant reduction in bias compared to a control group. Since there has been little investigation about using visual arts to address weight bias in medical students, the current study was initiated by a medical student to convey diversity in body types through works of art displayed in an art museum along with providing information from obesity experts, to generate discussion on perspective-taking, empathy and weight bias. Ultimately, the goal was to demonstrate a reduction in weight bias for the students who participated in study activities.

Methods

Subjects

First-year medical students from Rutgers Robert Wood Johnson Medical School (RWJMS) volunteered for this study in September 2018. Students self-selected to be in the experimental group (n = 24) and participate in study activities; the remainder (n = 21) agreed to be part of the control group. All participants completed a demographic survey, including questions about prior art training, and reflection about their own weight and the weight of others in their households, “underweight, overweight, about right.” These questions were relevant to ensuring that the experimental and control groups were comparable, i.e., that the experimental group was not composed primarily of students interested in art or those who were overweight. We were aware that besides BMI, gender identity and race have been shown to be related to fat bias in medical students [11], but decided not to collect this information since our small sample size would not provide enough variation to draw any meaningful conclusions. Demographic information is summarized in Table 1. This study was reviewed and approved as exempt research by the Rutgers University New Brunswick Health Sciences Review Board.

Table 1.

Demographics of first-year medical students participating in the study

Characteristic	Total	Experimental	Control
	(n = 45)	(n = 24)	(n = 21)
	n (%)	n (%)	n (%)
Gender
Male	17 (38)	8 (33)	9 (43)
Female	28 (62)	16 (67)	12 (57)
Age
≤ 24	39 (87)	20 (83)	19 (90)
≥ 25	6 (13)	4 (17)	2 (10)
Art or art history background
High school classes	23 (51)	12 (50)	11 (52)
College classes	22 (49)	14 (58)	8 (38)
Perceptions of own weight
Underweight	1 0(2)	0 0(0)	1 0(5)
About right	38 (84)	21 (88)	17 (81)
Overweight	6 (13)	3 (13)	3 (14)
Perceptions of household weighta
Underweight	0 0(0)	0 0(0)	0 0(0)
About right	38 (84)	21 (88)	15 (71)
Overweight	9 (20)	3 (13)	6 (29)

^aParticipants were asked to describe the weight of those in their childhood household

Research Activities

All participants completed four measurements chosen from among those used in previous studies, the Beliefs About Obese Persons scale (BAOP) [29], Attitudes Towards Obese Persons scale (ATOP) [29], Fat Phobia Scale (FPS) [30], and the Harvard Implicit Association Test (IAT) [31], and answered three researcher-developed questions to rate comfort and knowledge about working with patients who are obese. A short description and scoring of the validated measures are summarized in Table 2. Changes in Harvard IAT scores were also informally analyzed to assess individual student trends from baseline defined in three categories: more positive associations with fat body shapes, more positive associations with thin body shapes, and no change in preference. The three additional questions were all scored on a 5-point Likert scale, ranging from 5 = strongly agree to 1 = strongly disagree. The experimental group also completed a program evaluation.

Table 2.

Measurements utilized in pre- and post-test to assess weight bias

Measurement	Description	Scoring
Beliefs About Obese Persons scale (BAOP)	Participants rate agreement to statements on a 6-point Likert scale (strongly agree…strongly disagree) regarding controllable and uncontrollable factors for obesity. High values represent the belief that obesity is not within the control of an individual (8 question survey).	The scores from each of the 8 questions are summed to determine final score. Score ranges = 0 to 48.
Attitudes Toward Obese Persons scale (ATOP)	Participants rate agreement on a 6-point Likert scale (strongly agree…strongly disagree) to a series of statements about obese people. Higher values indicate more positive attitudes (20-question survey).	The scores from each of the 20 questions are summed to determine final score. Score ranges = 0 to 120.
Fat Phobia scale (FPS)	Participants are presented with two opposite adjectives and select along range of 1 to 5 regarding which adjective best describes their feelings towards obese people (e.g., industrious vs. lazy and strong vs. weak). Higher values indicate negative attitudes and stereotypes about fat people (14-question survey).	Each of the 14 questions receives a score from 1 to 5. The scores are averaged to determine a final score. Score ranges = 1 to 5.
Harvard Implicit Association Test (IAT)	Participants sort positive words, e.g., love, beautiful, and negative words, e.g., hate, nasty with silhouettes of thin and fat body shapes. The IAT is scored on the premise that faster responses indicate more closely associated concepts. A final score, indicated as slight, moderate, or strong preference is a measure of the strength of the associations based on the speed differences between the pairings. Implicit.harvard.edu	Results range from strong preference for fat people (+3) to strong preference for thin people (-3). A score of 0 indicates no preference. .

The intervention included guided viewing of works of art in a museum and interactive presentations by physicians with expertise in nutrition and obesity. The physician presentations provided information on causes of obesity and set the stage for normalization/de-stigmatization of patient care. The experimental group attended three educational sessions at the Rutgers University Zimmerli Museum of Art beginning in Fall 2018, and later in the year participated in two lecture/discussion sessions facilitated by physicians. Timing of study activities is summarized in Table 3. As incentive to complete all parts of the study, students earned a transcript notation indicating completion of a non-credit “Visual Literacy Elective.” During the three visits to the museum, students had time to socialize and enjoy snacks before the museum curator provided an introduction for each session; most time at the museum was spent in guided viewing of selected exhibits. The first classroom session in January 2019 was taught by a Family Medicine physician. As an introduction and to spark discussion, the physician showed the video from the UConn Rudd Center for Food Policy & Obesity [15] (depicting negative reactions of healthcare providers toward obese patients and then demonstrating the positive changes needed to make patients feel more welcome and comfortable). The second classroom session was presented by a physician who had personal struggles with obesity and currently interacts with a wide range of patients with obesity issues. Topics presented included motivational interviewing, challenges when considering weight loss options, and guidelines for choosing bariatric surgery. During both classroom sessions students had a chance to ask questions and interact with the physician speakers.

Table 3.

Study activities and timeline for the experimental and control groups

Group	Pre-intervention	Intervention	Intervention	Post-intervention
	Sept. 2018	Oct. 2017–Jan. 2018	Jan.–Feb. 2019	May 2019
Experimental group	Complete surveys	Museum visits (3)	Clinician talks (2)	Complete surveys and program evaluation
Control group	Complete surveys	No activity	No activity	Complete surveys

Data Analysis

All materials were distributed electronically and kept anonymous by matching student-generated identification codes for the pre- and post-test. Post-test differences between experimental and control groups were analyzed using univariate analysis of covariance (ANCOVA) on SPSS Statistics v25, considering variation in pre-test scores. Open-ended responses to the evaluation questions were independently analyzed qualitatively by the researcher authors.

Results

Measurement of Bias

All the students—those who participated in the study activities as well as those in the control group—demonstrated a decrease in bias over the time of the study. The pre- and post-test scores for all four measurements are listed in Table 4. ANCOVA analysis did not reveal significant differences between the experimental and control groups. On the pretest measure, 75% of all participants had “preference for thin individuals.” Forty percent of those who participated in study activities trended toward more positive associations with fat body shapes, and 30% had no change in preference from baseline score. The control group had just 29% trend towards positive associations with fat body shapes, and 58% had no change from baseline scores. These IAT trends are also reported in Table 4.

Table 4.

Pre- and post-test scores for experimental and control groups

	Control (n = 21) Mean (SD)			Experimental (n = 24) Mean (SD)		ANCOVA analysis between groups
Measurement	Pre	Post		Pre	Post	p value	F-value
BAOP scale	17.38 0 (6.03)	29.91 0 (8.60)		18.83 0 (6.90)	20.46 0 (4.99)	0.370	0.821
ATOP scale	69.33 (12.80)	74.14 (18.21)		68.58 (14.46)	74.83 (11.35)	0.715	0.135
Fat Phobia scale	03.63 0 (0.38)	03.38 0 (0.58)		03.58 0 (0.43)	03.44 0 (0.44)	0.552	0.360
Harvard IAT	-1.30 0(1.38)	-1.00 0(1.48)		-1.46 0(1.18)	-1.21 0(1.25)	0.630	0.236
Individual IAT Student Analysis			Control	Experimental
More positive word associations with fat body shapes			29%	40%

Both experimental and control group participants indicated strong initial levels of knowledge about factors that influence obesity, 4.38 and 4.48; mean of the post-test scores indicated increased knowledge, 4.67 and 4.57. Both groups of participants also indicated high levels of comfort interacting with patients who are overweight or obese, 4.21 and 4.38, with mean post-test scores for both groups indicating greater comfort with those interactions, 4.54 and 4.52. Prior to the study, both the experimental and control groups indicated ambivalence relating to comfort counseling overweight or obese patients about their weight, 3.00 and 3.48. Although participants overall indicated more comfort in counseling on the post survey, the gain in comfort was greater for the experimental group (3.00 to 4.13) compared to the control (3.48 to 3.95). The improvement between groups did not reach statistical significance, but the improvement noted for comfort for counseling neared significance (p = 0.12). This information is summarized in Table 5.

Table 5.

Pre- and post-ratings of knowledge about obesity and comfort in interacting and counseling patients

Questions (Scored on Likert scale 1=strong disagreement/discomfort 5=strong agreement/comfort)	Control (n = 21) Mean (SD)		Experimental (n = 24) Mean (SD)		ANCOVA analysis between groups
	Pre	Post	Pre	Post	p value	F-value
1. I feel knowledgeable about factors that influence obesity (e.g., environment, diet, exercise, genetics).	4.48 (0.51)	4.57 (0.51)	4.38 (0.58)	4.67 (0.48)	0.445	0.594
2. I feel comfortable interacting with patients who are overweight or obese.	4.38 (0.59)	4.52 (0.60)	4.21 (0.72)	4.54 (0.59)	0.666	0.189
3. I feel comfortable counseling overweight or obese patients about their weight.	3.48 (0.87)	3.95 (0.59)	3.00 (1.10)	4.13 (0.61)	0.120*	2.525

^*Nearing significance

Program Evaluation

Student feedback about the program was positive. The majority (91.6%) were satisfied with the training, education and content of the program, 95.8% agreed that the elective was “engaging, interesting, and entertaining,” and 95.8% would “recommend this elective to other first-year medical students.” Many (60.8%) agreed they felt “less biased towards overweight or obese people,” and 70.8% agreed with the statement, “I feel like a more capable medical student after completing the elective.” Several students provided comments expressing that the program focused on an important topic and wanted more opportunities for museum visits and physician lectures, and one student wrote that the elective was “an incredible addition to the curriculum.” Student feedback also suggested changes for future planning, including the selection of additional pieces of art with relevance to the theme of weight bias and the desire for additional program sessions.

Discussion and Limitations

This study was designed to develop an educational program to decrease medical student weight bias, a goal essential to providing universal healthcare to the high-risk population of obese patients. The art museum provided a location to discuss a variety of body types depicted in the works of art and was a unique and engaging space for medical student reflection on empathy building and wellness. The additional instruction by physicians with special interest and teaching experience in nutrition and obesity prevention was important in offering medical-related information. The physician presenters also served as role models in how to empathetically address the needs of patients. Since participation in the study was voluntary, it was important to ensure that all activities were engaging and enjoyable. Reassuringly, student feedback about the program was very positive.

We measured weight bias of first-year medical student participants prior to the intervention. Two of the measurements (BAOP and ATOP) indicated that our students had pre-intervention scores similar to those found in the validation study done with graduate students [29]. Initial results of the Fat Phobia scale indicated a “normal amount of fat phobia” among our participants [30]. The initial average Harvard IAT score of the participants in our study was in the category of “a slight to moderate preference for thin people over fat people” and therefore consistent with 44% of respondents to the IAT website which collected over 1 million IAT scores between April 2004 and December 2015. A positive finding was that most of our study participants indicated decreased bias toward people with obesity over the course of the study, regardless of whether they participated in the study activities. This finding might be explained by the content of their first-year medical school curriculum which included a course, Digestive Systems, Nutrition, and Metabolism, which addressed issues related to diet and obesity and included a required “culinary medicine” session where students discussed healthy and practical eating choices while participating in actual meal preparation with a certified dietitian. Another course, Patient Centered Medicine, discussed negative outcomes related to health disparities and detrimental effects of implicit and explicit bias, and provided opportunities for students to work with a diverse group of standardized patients.

Students who participated in the study activities provided positive comments with the majority reporting the program to be both enjoyable and valuable to their education. The art museum sessions encouraged meaningful contemplation and discussion while also allowing time for students to decompress from routine study, and thus included a “wellness” component. Providing space for self-reflection and exposure to diversity are valuable in reinforcing fundamental equity in healthcare, personal growth, and enhanced student well-being. As described earlier in the background section, incorporating humanities education within the medical school curriculum and through extracurricular activities is being encouraged and has gained momentum [27, 32]. We look forward to continued collaborations between the medical school and the Rutgers University Zimmerli Art Museum.

Although program activities were well-received, outcome measures did not reflect a statistically significant reduction in the bias of our program participants compared to the control group. This is partially explained by the relatively small number of participants in the study, the overall initial high level of participant knowledge about obesity, and the comfort of the students in interacting with patients with obesity, and by factors in the program implementation. Our demographic survey included questions we predicted might influence outcomes, and specifically to assure similarity between the experimental and control groups (art background and self-perception of weight). Although we did not include questions about race, ethnicity, socioeconomic status, or gender identity, we recommend that future studies with larger numbers of participants explore these complex factors in relation to weight bias. When replicating this program, we also recommend that discussion during the art museum sessions focus exclusively on body size and aesthetic preferences of the artist or of the time period, as discussion occasionally shifted to other topics unrelated to the human body. More frequent museum visits are recommended as the once-a-month sessions at the art museum (selected by necessity around the evenings the museum was open and student availability) may not have sufficiently reinforced empathy building and impacted results.

Although we found that all first-year medical student participants showed a reduction in bias during the time of the study, it is contrary to the findings of Phelan et al. [12] who found worsening explicit bias in students during the 4 years of medical school. Sadly, it may be that our student study participants will become more biased as they engage in clinical experiences, particularly during clerkship years, and perhaps observe a greater number of negative interactions between patients who are obese and physicians who display weight bias. It will therefore be important to advocate for ongoing programs to combat weight bias throughout medical school, both in the pre-clerkship and clerkship years. However, a more positive explanation for the reduction in weight bias may be the influence of some societal changes that have happened since 2015 when the Phelan et al. study was published. For example, movements like fat acceptance [33] and body positivity [34] are growing in popularity and being reported in the media. Additionally, a cultural shift has been noted by the increasing number of examples of fashionable plus-sized clothing options, the visibility of curvy models, and the popularity of overweight performers. We will want to continue to monitor to determine if these societal trends lead to enhanced acceptance and fewer negative attitudes towards obese and overweight patients [35].

Conclusion

This study was specifically designed to reduce medical student bias toward patients who are obese, as obesity ranks as one of the most significant factors impacting public health and healthcare in our nation and worldwide. Medical schools are mandated to include training in cultural competency and health care disparities, with opportunities for students to “recognize and appropriately address biases in themselves, in others and in the health care delivery process.” Although our findings did not reach a level of significance, the study provides evidence for promoting programs to integrate bias reduction and the humanities. As mentioned previously, the Association of American Medical Colleges (AAMC) recently published a monograph as part of the AAMC FRAHME (Fundamental Role of the Arts and Humanities in Medical Education) initiative, with the goal to “improve the education, practice, and well-being of physicians through deeper integrative experience with the arts and humanities,” and to enhance “…empathy, communication and teamwork skills” [20]. Empathy/perspective taking is essential for reducing bias, and the immense diversity of people, places, and events depicted through art can be a rich catalyst for promoting reflection and discussion. We believe similar programs could be successfully implemented in both medical and health-related-profession schools to reduce student bias, not only toward weight, but toward people of color, those with disabilities, the LGBTQ+ community, etc. An example of how art is being used to encourage international awareness of social issues [36] are the exhibits organized and promoted by Art Works for Change on topics ranging from race to climate change to advance understanding of societal problems.

Building communities of tolerance, acceptance, and kindness are of the highest priority. We must encourage faculty and students to design and implement programs with a goal to reduce biases throughout medical school, to advocate for humanities-based programs, and to strive to ameliorate the effect of bias on patient care. We must continue to work toward guaranteeing that all patients receive the equal care to which they are entitled.

Declarations

Ethics Approval

This study was reviewed and approved as exempt research by the Rutgers University New Brunswick Health Sciences Review Board.

Conflict of Interest

The authors declare no competing interests.