Frequency and Associated Predictors of Screening for Eating Disorders and Obesity in Female Students in College Health Centers

Abstract

Objective:

Screening for obesity and eating disorders (EDs) offers a pathway to treatment. The current study surveyed U.S. college healthcare providers regarding screening for obesity and EDs.

Participants:

Providers (n=1,159) from a nationally-representative sample of 384 college health centers

Methods:

Providers completed surveys regarding obesity and ED screening practices and beliefs. Provider-level and organizational-level factors were examined as predictors of screening frequency.

Results:

Providers reported screening larger percentages of female students for obesity (70%) than EDs (30%) and were more likely to report a center-wide policy to screen for obesity (59.1%) than EDs (33.9%). Providers reporting a routine screening policy predicted screening frequency in both obesity and EDs. Most providers believed screening for obesity (75.6%) and EDs (82.5%) was a good idea.

Conclusion:

Obesity is screened for more often than EDs in college health centers. Understanding screening barriers will be beneficial in connecting students with obesity and/or EDs to care.

Introduction

Obesity and eating disorders (EDs) are both weight-related conditions that negatively impact mental and physical health.^1–3 Research suggests that the conditions are often related. For example, obesity is a risk factor for developing an ED⁴ and certain EDs, such as binge eating disorder, are associated with increased energy consumption that can contribute to excess weight.⁵ Emerging adulthood (18–25 years old) is an important time period for both conditions, with high incidence rates observed for obesity and EDs.^{6, 7} This pattern of high obesity and ED rates translates to college campuses, where approximately 40% of emerging adults are enrolled.⁸ Studies of college campuses show rates of overweight/obesity have increased from 33.5% in 2010 to 38.3% in 2022⁹ and positive screens for EDs have increased from 14.9% in 2013 to 28% in 2020–2021.¹⁰

One step to reduce student experiences with obesity and EDs is to screen for these conditions and refer to evidence-based treatment. Screening for obesity (i.e., measuring height and weight and calculating body mass index according to obesity cut-offs) is recommended universally by the United States Preventive Services Task Force (USPSTF)¹¹ and the American Academy of Pediatrics recommends screening for EDs at annual visits and/or at examinations prior to participation in sports.¹² The goal of screening is to identify individuals who currently have or are at risk for developing obesity and/or an ED. Screening is a critical first step in raising patient awareness, increasing treatment access, and preventing progression of disease. At an organizational level, screening can also be used to identify the likely rates of obesity and/or EDs in the patient population, which could then influence organizational decision-making and initiatives (e.g., hiring or training staff to treat EDs, offering a weight management intervention). A number of studies have sought to quantify the prevalence of EDs on college campuses using various screening methods. A review by Fitzsimmons-Craft and colleagues outlined prevalence rates, potential screening tools, recommendations for screening and future directions.¹³ Existing screening data (i.e., number of positive screens) have primarily come out of research efforts to understand ED prevalence rates on campuses. As noted by Fitzsimmons-Craft et al., only one prior study conducted by the National Eating Disorders Association in 2013 examined screening practices for EDs on college campuses. Results from that survey indicated that fewer than half (45%) of the 115 participating colleges offered ED screenings on campus. However, these findings did not provide specific information about routine screening practices for obesity and EDs within college health centers, which may be the optimal place to identify students in need of treatment and connect them to care. Moreover, no studies of obesity screening in college health settings have been conducted. Understanding current college healthcare provider practices and individual- and organization-level influences of practice will identify care caps and inform future efforts to improve screening practices and clinical care for students.

To address this research gap, the 2022 National College Healthcare Provider Survey examined the clinical practices of providers in college health centers focusing on screening for seven priority health conditions in female college students, including obesity and eating disorders.¹⁴ This comprehensive survey collected data from college healthcare providers at 384 higher education institutions in 49 states and Washington, D.C. Providers’ self-reported rates of routine screening varied widely by screening focus (HPV vaccination status, intimate partner violence, tobacco use, alcohol use, depression and anxiety, obesity and eating disorders).¹⁴ Results showed that college healthcare providers reported screening 70% of female students for obesity and 30% of female students for eating disorders. The type of healthcare provider (e.g., nurse practitioner, physician) was not related to these self-reported screening rates. The current study sought to identify individual- and organization-level factors that acted as facilitator or inhibitors of screening for obesity and EDs.

There are a number of factors that may influence screening practices. The Theory of Planned Behavior (TPB) posits that an individuals’ behavioral beliefs, social normative beliefs, and control beliefs / perceived behavioral control are the primary determinants of their behavioral intentions, and that behavioral intentions, in turn, are the primary determinants of the actual behavior.¹⁵ It should be noted that the TPB construct control beliefs/perceived behavioral control is similar to and sometimes referred to as self-efficacy and/or confidence in one’s ability to perform a behavior. The TPB has been widely used to identify factors that influence individuals’ behavioral intentions and engagement in health behaviors and health screenings for a wide variety of conditions, such as breast cancer¹⁶, cervical cancer.¹⁷ It has also been used to understand healthcare providers’ practice behaviors, including screening practices for sexual and intimate partner violence.¹⁸ Studies have found that providers’ positive attitudes towards and perceived behavioral control of screening increase their screening intentions and behaviors.^{17, 18}

The TPB has been applied to a lesser degree to screening for obesity and eating disorders. One existing study of general practitioners found that more positive attitudes towards referral for ED treatment were more likely to be related to greater intentions to refer a patient with ED symptoms.¹⁹ Thus, the TPB may be a valuable lens to consider variations in college healthcare provider ED and obesity screening practices given the success of predicting screening for other conditions.

By employing an ecological or multi-system-level lens, one can appreciate that organizational factors and the context in which an organization exists can also impact individual provider practices.²⁰ Thus, organization-level factors must be included in efforts to understand influences of current practice behaviors and incorporated into the design of interventions to change practice and/or implement evidence-based practices.^{18, 20} The Consolidated Framework for Implementation Research (CFIR) identifies determinants in the outer setting (e.g., state, campus) and inner setting (e.g., clinic) that may influence clinical practices including screening for obesity and EDs.²⁰ Related to the outer setting, obesity rates have been shown to vary across higher education institutions by geographical region and school type (i.e., public vs. private)²¹ and resources (e.g., ED specialist or center) may be more available in and around colleges that are located in urban compared to rural locations.²² If college health centers are serving a student population with greater obesity rates, they may be more likely to screen for obesity. Moreover, if greater treatment resources for EDs exist and practitioners have more options for referrals, they may be more likely to screen for EDs. Examining school-level demographics may offer a window into outer setting factors influencing screening in college health centers.

In the inner setting, one CFIR determinant that may influence screening is access to knowledge and information about an evidence-based practice, or access to training on how to implement that practice. Trainings, particularly those that include information as well as role-modeling, skills training, practice and feedback, are effective ways to increase providers’ self-efficacy, confidence and/or perceived behavioral control and are effective ways to promote evidence-based practice in a healthcare organization.²³ Provider trainings have been used effectively to increase providers’ knowledge, skills, and comfort in identifying and appropriately addressing obesity and EDs in practice.^{24, 25} Recent participation and completion of training focused on screening for obesity and EDs may promote provider screening of obesity and EDs in college health centers.

A second CFIR determinant of screening within the inner setting could be the workflow support and resources an organization has to promote use of a practice or innovation.^{20, 26} Related to screening of obesity and EDs, this could include having a center-wide policy for screening at certain time intervals or appointment types within the patient population. Research on screening for sexual violence found that having a policy to screen and/or a prompt to screen in the electronic health record was associated with higher rates of screening for sexual violence in college health centers.¹⁸ Similarly, having a protocol for violence screening was associated with higher screening rates in primary care.²⁷ The effects of policies and prompts to screen have not been examined as a possible factor influencing screening rates for obesity and EDs in college health centers.

The current study aimed to examine and describe college healthcare providers’ screening practices for obesity and EDs and identify provider-level (i.e., attitudes, perceived behavioral control) and organizational-level (i.e., region, school type, urbanicity, policies, and trainings) factors associated with providers’ self-reported screening rates for these conditions. We hypothesized that more positive attitudes and greater perceived behavioral control at the provider level and presence of organizational characteristics that would promote implementation of screening would be associated with greater rates of self-reported screening for obesity and EDs.

Methods

Study design, Sample, and Procedures

Data for this study came from the National College Healthcare Provider Survey (NCHPS) funded by the Agency for Healthcare Research and Quality (AHRQ; R01HS027154). College healthcare providers (n=2,994) from 471 accredited, general four-year colleges and universities were contacted via the U.S. mail at their college health center addresses and invited to complete either paper-and-pencil or online surveys. College providers included nurse practitioners (NPs), physicians (MD/DOs), and physician assistants (PAs). Altogether, 1,159 providers from 384 colleges/universities completed surveys; this represented a response rate of close to 40%. Participants were offered $20 gift cards as incentives for their participation. The procedures for sampling, recruitment and data collection followed best practices²⁸ and are described in detail in Sutherland et al.¹⁴ Notably, the parent study focused on intimate partner and sexual violence in female college students, therefore most questions asked specifically about screenings with female students. The Institutional Review Board at University of Rhode Island reviewed and approved all human subjects procedures.

Measures

Screening rates.

Providers’ self-reported rates of routine screening for obesity and EDs were assessed using two questions that asked, “Of the female college students who you saw at the college health center during the spring 2022 semester, approximately what percentage (%) did you screen for or ask about [obesity OR eating disorders]?” Response options were provided in 10% increments ranging from 0% to 100%. This question structure has been used previously to survey providers about screening rates.¹⁸

Provider-level variables.

Providers’ perceived behavioral control in their ability to screen was assessed with the following questions: “During the next semester, I am confident that I could… . (a) regularly screen male and female students for obesity, (b) discuss obesity status with students who meet criteria, (c) provide guidance around weight loss for students with obesity, (d) regularly screen for eating disorders, and (e) refer students who screen positive for an eating disorder for follow-up and counseling.” Response options ranged from strongly disagree (=1) to strongly agree (=5). Only questions pertaining specifically to screening (i.e., a and d above) were entered as a predictor of self-reported screening percentages; the other survey items provide descriptive indication of provider perceived behavioral control to provide follow-up services. Provider attitudes towards routinely screening for obesity and EDs were assessed with two survey items: “In your opinion, is it a bad idea or a good idea to routinely screen for or ask about the following with every female student who visits the college health center … obesity? …eating disorders?” Response options ranged from very bad idea (1) to very good idea (5). These questions were adapted to measure the constructs of perceived behavioral control and behavioral beliefs/attitudes from the TPB following guidance from Fishbein and Ajzen (2010).²⁹

Organization-level variables.

Questions about the college/university in which the health center was located included state (coded into midwest, northeast, west, and south for analyses), college type (Response options: public/state, private religious, private secular), and urbanicity (Response options: rural, suburban, urban, don’t know). These characteristics provide insight into the CFIR outer setting and may tie to the perceived patient needs and resources. Presence of a screening policy and recent trainings in screening (i.e., inner setting determinants of provider-reported screening rates) were also assessed. To determine presence of a policy, providers were asked, “Does your college health center have a policy to routinely ask about or screen all female students for … obesity? eating disorders?” Trainings for both obesity and EDs were assessed with the following question “During the past year, have there been any trainings or in-services at your college health center about how to ask about or screen for each of the following?” with obesity and eating disorders included in the response options. For policy and training questions, response options included yes, no, and don’t know. “Yes” responses were coded as (1); other responses were coded as (0). Similar questions have been used previously in Sutherland and Hutchinson (2019).¹⁸

Statistical analyses

Demographic frequencies for providers and schools were calculated as were descriptives of responses to survey questions. Medians were used for questionnaire responses that did not follow a normal distribution. A Wilcoxon signed-rank test was used to compare screening frequencies for obesity and EDs and Chi-Square tests were used to compare responses regarding policies, training, and attitudes by obesity and EDs. Random-intercept multilevel models nested within school and estimated with restricted maximum likelihood were used to examine the association between organization-level and provider-level predictors with reported screening frequency of obesity and EDs. Models predicting screening frequency were built in three stages: 1) school-level demographic characteristics, 2) provider reports of policies and trainings were then added to models, and finally 3) provider attitudes and provider perceived behavioral control were added to the models. Predictors pertaining to eating disorders were only used in eating disorder models and predictors pertaining to obesity were only used in obesity models. SPSS (IBM version 28) was used to download and analyze data.

Results

As reported elsewhere,¹⁴ most respondents were female, Caucasian/white, held NP licenses and were employed as providers in the college health centers of state colleges/universities [Table 1]. Regional location, school type, and urbanicity of the schools had diverse representation in the sample. The number of providers representing a school ranged from 1 to 18 with a mode of 1 (n=117), a median of 2 and a mean of 3.02.

Table 1.

Demographics of College Healthcare Provider Respondents to the National College Healthcare Provider Survey

	Percentage of Sample (n=1159)
Gender:
Female	84.20%
Male	15.40%
Non-binary or Other	0.40%
Age:
20 – 39 years old	23.00%
40 – 49 years old	26.60%
50 – 59 years old	31.50%
60 + years old	18.90%
Race:
African-American/Black	4.50%
Asian	8.50%
AI/AN	0.60%
Caucasian/White	84.40%
Other	2.00%
Hispanic Ethnicity	4.50%
College/University Type:
State	73.30%
Private/religious	10.10%
Private/secular b	16.60%
College/University Enrollment: c
< 5,000	17.70%
5,000 – 9,999	19.60%
10,000 – 19,999	25.00%
≥ 20,000	37.70%
College/University MSI or HBCU c
Yes	10.90%
No	89.10%
Location: c
Urban	44.20%
Suburban	36.60%
Rural	19.20%
Region:
Northeast (12 states + DC)	28.80%
Midwest (12 states)	21.70%
South (13 states)	27.00%
West (11 states)	22.50%
Provider Degree:
NP	53.70%
MD/DO	34.10%
PA	12.20%
Survey Completed Online	64.60%

Descriptives of obesity and ED screening variables

Descriptive statistics for survey responses are presented in Table 2. Approximately one-third (33.9%) and two-thirds (59.1%) of respondents indicated that their health center had a policy to routinely screen for EDs and obesity, respectively. Notably, providers from the same college health centers frequently offered different responses as to whether such screening policies were in place. Inconsistent reports were noted in 37% and 45% of schools with multiple respondents for EDs and obesity, respectively (i.e., some providers indicated a policy was in place while others indicated no policy was in place). The degree of disagreement did not suggest a clear pattern (i.e., the results did not suggest most responding providers believed there was a policy with few believing there was not one or vice versa).

Table 2.

Descriptives and Comparisons of Eating Disorder and Obesity Survey Items

	Eating Disorders	Obesity	Test Statistic	p-value
Percent reporting center-wide screening policies	33.9%	59.1%	X2 = 80.2	p<0.001
Median percent of female students screened in Spring 2022	30%	70%	z=14.9	p<0.001
Percent of providers who believe screening is a good or very good idea	82.5%	75.6%	X2= 3.5	p=0.06
Percent of providers offered trainings in previous year	25%	15%	X2= 29.0	p<0.001
Behavioral control ratings (from 1–5) to… (M ± SD/Median/Mode)
Regularly screen for obesity	-	4.26 ± 0.94/4/5
Discuss obesity status with eligible students	-	4.09 ± 0.95/4/4
Provide guidance around weight loss	-	4.12 ± 0.95/4/4
Regularly screen for EDs	3.78 ± 1.00/4/4	-
Refer positive ED screens	4.47 ± 0.80/5/4	-

Providers reported screening greater proportions of female patients for obesity rather than for EDs, z = 14.9, p <0.001 [Figure 1]. The median percent of female students reportedly screened for obesity was 70% compared to 30% for EDs. The majority of providers indicated that they believed routinely screening for EDs (82.5%) and obesity (75.6%) was a good or very good idea. Recent in-service trainings were not common; only 25% of providers reported ED screening trainings and 15% reported obesity screening trainings were offered during the previous year. Providers reported relatively high behavioral control to screen for and intervene in obesity and EDs. Perceived behavioral control for obesity screening was significantly higher than perceived behavioral control for ED screening, t(1110)=14.6, p<0.001.

Figure 1.

Percentage of College Healthcare Providers who Reported Screening Female Students for Eating Disorders and Obesity at Low (0–30% of students), Medium (40–70% of students), and High (80–100% students) Frequencies in Spring 2022.

Organization- and provider-level predictors of ED and obesity screening variables

Results of models for predictors of provider-reported screening frequencies are presented in Table 3.

Table 3.

Predictors of Provider-Reported Screening Rates for ED and Obesity in Female Students

	Eating Disorders		Obesity
	Percent screened	F-Statistic	Percent screened	F-Statistic
Region		0.24		6.65***
Northeast	-		-
Midwest	2.38 (3.50)		15.08 (4.01)*
South	1.55 (3.33)		14.40 (3.82)*
West	2.68 (3.52)		6.68 (4.04)
College Type		0.72		16.91***
State/Public	-		-
Private Religious	−3.74 (3.65)		−24.07 (4.19)*
Private Secular	1.18 (3.42)		−8.76 (3.93)***
Urbanicity		2.68		2.34
Urban	-		-
Suburban	1.89 (2.32)		−0.73 (2.69)
Rural	−5.04 (3.00)		−7.23 (3.46)
Screening Policies a		505.18***		337.65***
No	-		-
Yes	39.25 (1.75)		37.79 (2.06)
Trainings a		58.13***		0.24
No	-		-
Yes	17.33 (2.27)		6.14 (3.41)
Provider Perceived Behavioral Control	9.61 (0.81)	142.37***	12.30 (1.03)	140.56***
Provider Attitudes	8.26 (1.01)	66.88***	10.93 (0.99)	122.82***

^aModels were created in 3 stages: 1) region, college type, and urbanicity, 2) screening policies and trainings controlling for previous variables; 3) provider perceived behavioral control and attitudes controlling for previous variables;

^*p<0.05;

^**p<0.01;

^***p<0.001

Obesity.

Providers reported higher screening rates of female students for obesity at schools in the South (56%) and Midwest (57%) compared to the schools in the Northeast (42%). Screening rates were also higher at state schools (62%) compared to private religious (38%) and private nonsecular schools (53%). Belief that a center-wide policy for screening was in place was a significant predictor of percentage of female students screened for obesity (policy = 70% vs. no policy = 32%), but trainings were not significantly associated with screening frequency. More positive provider attitudes towards screening for obesity were associated with more frequent screening. Specifically, an increase of 1-point on the 5-point Likert scale was associated with increases of 10.9% in reported frequency of screening for obesity. A similar pattern was observed for perceived behavioral control of screening; a one-point increase in perceived behavioral control of screening was associated with an increase in reported screening frequency of 12.3% for obesity. Urbanicity did not predict percentage of female students screened.

Eating disorders.

The belief that there was a center-wide screening policy (policy = 63% vs. no policy = 24%) and access to trainings (trainings = 49% vs. no trainings = 31%) were significantly associated with the percentage of female students screened for EDs. More positive provider attitudes towards screening for EDs were associated with more frequent provider-reported screening. Specifically, an increase of 1-point on the 5-point Likert scale was associated with increases of 8.3% in reported frequency of screening for EDs. Moreover, a one-point increase in perceived behavioral control of screening was significantly associated with an increase in reported screening frequency of 9.6% for EDs.

Discussion

The current study elaborates on previously published findings indicating that providers in college health settings self-reported personally screening approximately 2/3 of female student patients for obesity and approximately 1/3 of female student patients for an ED in Spring 2022. Belief that a clinic policy existed for screening was a key predictor of provider reported screening rates for both obesity and EDs, indicating the utility of a policy to increase screening behavior. Reported clinic policies for routine screening were more common for obesity than for EDs, partially explaining higher screening rates for obesity. EDs are mental health disorders whereas obesity is a physical condition.^1–3 Thus, obesity may be considered within the scope of practice in health centers, whereas ED screening and treatment may be considered better suited to counseling centers (counseling center providers were not assessed in the current sample), which may help explain the lower rates. Indeed, in primary care settings, USPSTF guidelines indicate obesity screening is part of routine care¹¹ whereas eating disorders often go unnoticed in primary care settings.³⁰ This premise would align with the finding that providers reported greater behavioral control to screen for obesity than to screen for EDs. Nevertheless, a large majority of providers endorsed positive attitudes towards screening for both conditions, indicating potential interest in increasing screening practices in health centers.

ED screening can be more complex than obesity screening, potentially explaining the limited use of ED screening in health services centers. Height and weight measurements are frequently part of the workflow of medical appointments, medical providers receive training and practice in taking height and weight measurements, and electronic medical records can calculate BMI and diagnose obesity, whereas ED screening requires identification, administration, scoring, and interpretation of a screening questionnaire that may fall outside of traditional workflows even though relatively simple screening tools exist.¹³ Considering when and where ED screening could be incorporated in workflows and redesigning accordingly, for instance including a screening questionnaire with automatic scoring in the electronic medical records, may be one effective implementation strategy to increase provider screening behavior.³¹ Trainings in ED diagnosis, assessment, and management improve clinician confidence, knowledge, and skills in managing patients with EDs³² and may be another valuable pathway to improve ED screening rates and provider comfort and skills, as providers receive limited exposure to and instruction in working with patients with EDs during their own education.^{33, 34} Findings from the current study support training as a way to improve screening rates for EDs in college health centers.

Lastly, clinical guidelines for ED screening are more ambiguous than screening for obesity. The United States Preventive Services Task Force (USPSTF) recommends universal screening for obesity in adults and considers it part of routine care,¹¹ whereas a recent USPSTF review concluded that there is not enough information on the risks and benefits associated with EDs to suggest universal screening³⁵ (though the AAP recommends ED screening at well-child visits).¹² It may be that providers are screening for EDs in a targeted way, for instance, only if they believe a patient to be suffering from an ED. While this may be effective to identify EDs in some individuals with physical symptoms (e.g., low weight characteristic of anorexia nervosa), it may neglect identifying EDs in others with primarily psychological or behavioral symptoms, such as binge eating disorder, bulimia nervosa, or atypical anorexia nervosa. Indeed, studies have indicated that practitioners have difficulties detecting bulimia nervosa in their routine clinical practice.³⁶ If this is the case, particularly given the high rates of EDs in college students, routine screening would be one way to more consistently and comprehensively identify ED symptomology in students. Trainings in recognizing ED symptomatology, particularly in disorders that may be overlooked, such as atypical anorexia or binge eating disorders, may complement this practice.

Notably, the majority of providers were not screening all students for obesity, as the USPSTF guidelines would recommend. College health centers often provide acute care (e.g., respiratory illness, injury)³⁷ and calculation of BMI may not be considered relevant for these visits. It is possible that BMI calculation of a student of healthy weight may not be considered to be obesity screening by some providers, leading to underreporting. Additionally, providers have also noted discomfort discussing weight due to its sensitive nature and have concerns about stigmatizing students or promoting disordered eating, leading them to shy away from taking weight measurements.³⁸ Relatedly, others may operate from a weight-neutral approach in which body weight alone is believed to have little impact on one’s health and thus not see the need to identify BMI.³⁹ Part of the hesitancy in calculating BMI measurements may exist due to studies of universal screening programs, such as BMI report card programs in elementary and secondary schools. BMI report cards, which involve BMI measurements and reports sent home to parents, have shown little benefit and potential harm in children.⁴⁰ In line with the USPSTF obesity screening and management guidelines, a positive screen for obesity (i.e., a calculated BMI ≥30 m/kg² based on height and weight measurements) should be followed by counseling and referral to intensive multicomponent behavioral intervention.¹¹ Most schools do not offer this type of obesity care and obesity treatment programs have limited availability in other healthcare settings.^{41, 42} Providers should be thoughtful in screening for obesity and discussing obesity status with students, particularly if they are unable to identify resources. Indeed, college healthcare providers have indicated lack of resources for treatment and referral as one of the reasons obesity is not discussed with students.³⁸

The region and type of higher education institution were found to be associated with obesity screening. Specifically, the results indicated that providers from Southern and Midwestern schools compared to schools in the Northeast as well as public schools compared to private religious and secular schools were more likely to have providers who reported greater obesity screening frequency. A recent study of college campuses found obesity rates to be higher at Midwestern schools compared to Northeastern schools and at public schools compared to private schools (a category that likely encompassed religious schools).²¹ As indicated by the CFIR, outer setting contexts can impact clinical practices.²⁰ Caring for students in larger body sizes and related health conditions may prompt more frequent obesity screening to provide appropriate care, though this was not explicitly assessed. Neither region, college type, nor urbanicity were associated with ED screening frequency. Prior studies have indicated ED prevalence and access to resources can vary by these characteristics,^43–45 though research is scant. Additional research into outer setting factors may provide greater insight into variations in ED screening practices.

Additionally, in support of the TPB,¹⁵ more positive individual provider attitudes towards and greater perceived behavioral control to screen predicted greater provider-reported screening frequency for both EDs and obesity in this study. Other research also supports that attitudes and perceived behavioral control influence the use of evidence-based practices in clinical settings,⁴⁶ including screening of various health conditions.^{17, 18} Training in best practices for screening can impact these factors. For example, a study of clinical trainees reported greater confidence in caring for individuals with EDs following an online training.²⁵ Notably, attitudes and behavioral control do not reflect the skill level with which providers are screening and intervening with students who screen positive for these conditions. Future work may explore the quality of interactions students are having with providers around these topics, including how many students receive appropriate treatment and experience symptom improvement. Patient engagement with treatment following a positive screen and/or referral is limited for both eating disorders¹³ and obesity.⁴⁷

This study has many strengths, including the large and representative sample of college health centers in the United States. However, a primary limitation is the focus on only female students. The data collection occurred as part of a larger study focused on screening rates of interpersonal violence and sexual violence, which is more frequently directed towards female-identified individuals and as such, so did many survey questions. EDs occur at approximately double the rates in females compared to males,⁴⁸ whereas obesity rates are more similar across these two genders.⁴⁹ Future research should consider how screening may differ by gender of the patient, including in gender minority individuals (e.g., transgender). Moreover, the survey questions were an initial assessment of provider screening practices and attitudes. The scope of the project did not allow for a comprehensive assessment of all possible factors that could influence screening.

There was also some ambiguity in the survey language. In particular, questions asked about “routine screening”, which was left undefined. Different providers may have different interpretations of the words “routine” and “screening” for each condition. For example, screening of height and weight without calculation of BMI or calculation of BMI without categorizing it by weight category may be perceived as obesity screening by some but not others. Similarly, for EDs, an informal inquiry into concerns about eating without use of a validated tool may be perceived as screening by some but not others. This may have contributed to inconsistency in reporting on the existence of screening policies across providers in the same health center. Relatedly, data were all self-reported by providers and as such, some data, such as percentage of students screened or the existence of a center-wide policy, reflect providers’ perceptions and are not objectively measured data. College health centers frequently treat acute conditions (e.g., illness, injury) during which it may not be expected or appropriate to engage in screening efforts for unrelated conditions. Future work should focus on providing greater insight into how and when screening for obesity and EDs is used within college health centers and how the information is utilized (e.g., patient referral, in-house treatment, etc.).

In conclusion, this study suggests that it is more common to screen for obesity than EDs in college health centers. Both conditions can pose physical and mental health risks for students and screening is a necessary first step to connect students with these conditions to evidence-based care. More research is needed on optimal processes and timing of screening as well as to consider how to minimize potential unintended harms, particularly in relation to EDs as minimal research exists in this area. The ability to make appropriate referrals following a positive screen is an integral follow-up step that should be in place prior to screening initiation.

Data Availability:

The study and analysis plan were not pre-registered. De-identified data and analytic code from this study are not available in a public archive. De-identified data and analytic code from this study may be made available (as allowable according to institutional IRB standards) by emailing the corresponding author. Materials used to conduct the study are not publicly available.