Development and Psychometric Assessment of the Weight and Eating Quality of Life (WE-QOL) Scale in US Military Veterans

Abstract

Background

The Veterans Health Administration (VHA) is in need of population health approaches to address overweight and obesity-related diseases. BMI serves as a simple, blunt metric to monitor these efforts. However, emerging research has demonstrated that healthcare weigh-ins contribute to weight stigma which paraodoxically is associated with weight gain. An alternative metric is urgently needed for VHA’s MOVE!® Weight Management Program and other eating- and weight-related services.

Objective

To develop a brief population health metric called the Weight and Eating Quality of Life (WE-QOL) Scale and assess its psychometric properties.

Design

The literature was reviewed for relevant weight- and eating-specific QOL measures to identify unique and overlapping constructs. Eight items, representing these constructs, comprised the new brief WE-QOL Scale. A survey study was conducted with data analyzed in STATA.

Participants

A total of 213 consecutively evaluated US Veterans attending an orientation session for MOVE!.

Main Measures

The WE-QOL Scale, as well as a widely used generic health-related QOL measure, the European Quality of Life Screener (EQ-ED-5L), and relevant validated measures.

Key Results

WE-QOL descriptive findings demonstrated severe impacts on physical activity and physical discomfort for approximately 30% of the sample each; moderate-to-severe impacts on daily responsibilities, emotional distress, and shame and guilt for one-third of the sample each and public distress for one-fourth of the sample. The WE-QOL Scale performed as well as, or better than, the EQ-ED-5L for internal consistency (Cronbach’s alpha = 0.91) and associations to relevant constructs (BMI, eating pathology, and physical activity).

Conclusions

Findings support the reliability and construct validity of the WE-QOL Scale. The WE-QOL Scale has potential to provide a standardized population health metric that could be used as a screening tool and clinical reminder to identify, refer, and assess outcomes for Veterans with weight and disordered eating issues. Future research could be targeted at using this measure to improve patient care and quality of care.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11606-023-08132-4.

INTRODUCTION

The Veterans Health Administration (VHA) is the largest healthcare system in the USA and is challenged by addressing obesity-related diseases in a large segment of the Veteran population. The 2014 prevalence rates of overweight and obesity for VHA patients were 37% and 41%, respectively, in comparison to 2017-18 prevalence rates of 31.1% and 42.5% for the US general population.^1,2 Prevalence rates of obesity-related comorbidities are higher in Veterans compared to those in the civilian population. Rates for heart disease, coronary heart disease, hypertension, and stroke are 24.6%, 16.6%, 45.1%, and 6.5%, respectively, for the former and 10.5%, 5.3%, 24.8%, and 2.7%, respectively, for the latter.³

The MOVE!® program is VHA’s national evidence-based weight management and health prevention program designed to improve the lives of Veterans by encouraging healthy eating behavior and increasing physical activity. As a population health approach, the program strives to promote even small weight losses for a large number of Veterans to reduce health risks, prevent or reverse certain diseases, and improve quality of life (QOL).

While BMI is a useful screening tool to identify risk for obesity-related diseases and has potential to assess outcomes of VHA programs such as MOVE!, weigh-ins in healthcare settings have been identified as significant sources of weight stigma and paradoxically contribute to obesity.⁴ It is for this reason that alternative measures for weight management services are needed. Consistent with this line of research, the VHA State of the Art Conference on weight management identified an urgent need for a new population-level health metric that would ensure VHA-supported weight management interventions would not stigmatize patients.⁵ A measure of QOL that could be useful to this VHA program, as well as other weight- and eating-related services, has potential to reduce or ameliorate this problem. In VHA, weight and eating are also addressed across a wide range of clinics and settings including, but not limited to, primary care, nutrition services, pharmacology, bariatric surgery, whole health, and mental health/eating disorder care.

Health-related QOL (HRQOL) measures,^6,7 especially brief versions,^8,9 have potential use for population-level interventions. Disease-specific QOL measures, such as the Diabetes Quality of Life (DQOL)¹⁰ and the American Chronic Pain Association QOL Scale,¹¹ provide more nuanced data for specific patient populations than general HRQOL measures. Beyond reducing stigma, a QOL measure for weight- and eating-related services has potential to benefit patient care and quality of care. To date, a brief and specific QOL measure for weight^12,13 and disordered eating^14–17 does not exist.

This study had two aims. First, to address the need for a population health metric for weight and eating, we developed the Weight and Eating Quality of Life (WE-QOL) Scale. We intended for this new measure to be a brief (i.e., eight items) weight- and eating-specific QOL measure that could be used for a range of VHA programs or settings in which Veterans present with weight and disordered eating issues. Second, we aimed to assess the psychometric properties of the WE-QOL Scale and its construct validity in relation to a widely used general health-related QOL questionnaire, the European Quality of Life Screener (EQ-ED-5L),¹⁸ and other established weight- and eating-related measures to ensure that the measure is capturing QOL relevant for Veterans with weight and disordered eating issues.

METHODS

Participants

Participants were US Veterans (N = 213) who attended an orientation session for the VA National Weight Management Program, MOVE!, conducted at the VA Connecticut Healthcare System (VA CT) (see Table 1 for participant demographics).

Table 1.

Patient Demographics (N = 213)

	n (%)
Gender
Male	180 (84.51%)
Female	33 (15.49%)
Age (SD)	58.9 (SD = 12.93)
Race
White	132 (64.08%)
Black	53 (25.73%)
Multiple races	11 (5.34%)
Other	10 (4.85%)
Ethnicity
Hispanic	20 (10.31%)
Not Hispanic	174 (89.69%)
BMI (SD)	35.48 (SD = 6.04)
Binge eating (VA-BES)
Never	51 (23.94%)
Endorsed	162 (76.06%)
Employment
Employed	54 (25.47%)
Retired	107 (50.47%)
Student	9 (4.74%)
Unemployed	41 (19.34%)
Education
Less than high school	7 (3.32%)
High school graduate or GED	54 (25.59%)
Some college or associate degree	91 (43.13%)
Completed college	37 (17.54%)
Graduate school	22 (10.43%)

Procedure

To address the first aim, which was the development of the WE-QOL Scale, investigators with weight and eating disorder experience (i.e., first author RM, a postdoctoral fellow with relevant experience AB, and a research assistant AGM) searched for systematic reviews of obesity- and eating disorder (ED)–specific QOL measures to identify common domains for the new scale. No systematic reviews on this topic were found at the time of this study. One obesity-specific QOL non-systematic review¹⁹ identified four QOL domains (somatic, physical function, emotional state, and social interaction) among 11 widely used measures. Investigators found four ED-specific QOL measures^14–17 comprising 12 domains and considered all 16 domains (4 weight-related and 12 eating-related). Overlapping domains were consolidated and non-healthcare-related domains (e.g., leisure) were eliminated resulting in eight domains. These domains were converted to eight items, each prompting respondents to think about the impact of weight or eating for that domain using a 5-point scale from 0 (none) to 4 (could not be worse) (see Table 2). The eight items include as follows: physical activity, self-care, daily responsibilities, relationships, public distress, physical discomfort, emotional distress, and shame/guilt. A composite score was arrived at by adding scores for the eight items and dividing by eight. The measure was intentionally worded to limit concerns about health literacy and readability, and scores at a 5th grade reading level.

Table 2.

The Weight and Eating Quality of Life (WE-QOL) Scale

	None		Slight		Moderate		Severe		Could not be worse
	N	%	n	%	n	%	n	%	n	%
How much impact does your weight or eating have on your…
Physical activity (e.g., reaching, lifting bending, sitting, standing, exercising, walking, sexual activity)	33	15.49	40	18.78	79	37.09	53	24.88	8	3.76
Self-care (e.g., eating, bathing, dressing, toileting)	101	47.42	49	23.00	45	21.13	15	7.04	3	1.41
Daily responsibilities (e.g., work, study, housework, cooking, caring for family)	83	38.97	59	27.70	53	24.88	13	6.10	5	2.35
Relationships (e.g., family, friends, romantic partners)	85	39.91	54	25.35	37	17.37	29	13.62	8	3.76
Public distress (e.g., appearance in public, fear of ridicule/ discrimination, fitting in seats/aisles, eating in front of others)	118	55.40	31	14.55	43	20.19	12	5.63	9	4.23
Physical discomfort (e.g., shortness of breath, gastric distress, pain [back, arthritis, knee, or other])	31	14.55	44	20.66	66	30.99	49	23.00	23	10.80
Emotional distress (e.g., stress, anxiety, down/depressed, difficulty concentrating)	69	32.39	51	23.94	49	23.00	30	14.08	14	6.57
Shame/guilt (e.g., feeling self-conscious, unsure of self, do not like oneself)	92	43.19	41	19.25	42	19.72	21	9.86	17	7.98

To address the second aim, which was to assess the psychometric properties of the WE-QOL Scale, we collected paper-and-pencil survey data from self- or clinician-referred participants who attended one of 47 consecutive, in-person, bimonthly MOVE! weight management orientation sessions at VA Connecticut Healthcare System (VA CT) between January 1, 2018, and October 31, 2019. MOVE! is VA’s nationally available comprehensive weight management program that combines behavioral, dietary, and physical activity components. At the time of the study, Veterans targeted for the program had a BMI greater than 30, or a BMI between 25 and 29.9 with at least one obesity-related medical comorbidity. Not all Veterans who attended an orientation session went on to enroll in MOVE!. For those who did, Veterans attended sessions in a group format either in-person or by telehealth. Survey completion was a clinical requirement of the program for VA CT during this time period, and thus, written consent was waived and no incentives were provided. The VA CT Human Studies Institutional Review Board approved this study. All participants (N = 213) completed the following measures.

Main Measures

Demographics

Participants self-reported their age, race, ethnicity, education, employment status, height, and weight.

Body Mass Index (BMI)

We calculated BMI from self-reported height and weight. This method for calculating BMI has high concordance with measured BMI in Veterans.²⁰

The Weight and Eating Quality of Life (WE-QOL) Scale

The WE-QOL Scale was an 8-item measure of weight and eating QOL developed specifically for this study. A composite score is arrived at by adding scores for the eight items and dividing by eight. Higher scores represent worse QOL and range from 0 to 4.

The European Quality of Life Screener (EQ-ED-5L)¹⁸

The EQ-ED-5L is a measure of health-related QOL. For the purposes of this study, we used two methods for reporting results of the EQ-ED-5L: (1) the EQ-ED-5L Visual Analog Scale (VAS), which is a measure of overall self-rated health status, and (2) the EQ-ED-5L index value which provides population-based values for comparison. The VAS consists of a 100-point scale such that higher scores reflect better health. The index value was calculated based on value sets for the US population²¹ using the eg5d package²² in the statistical software R version 4.2.0. Index values range from less than zero (where zero represents a health state equivalent to dead and negative values represent health states worse than dead) to one (the value of full health), with higher scores indicating better health. The EQ-ED-5L is a valid, reliable, and widely used measure in the US population.²¹

The Questionnaire on Eating and Weight Patterns-5 (QEWP-5).²³

The QEWP-5 is a widely used and psychometrically valid self-report instrument to measure DSM-5 eating disorder diagnostic criteria.²³ This measure was adapted from the original Questionnaire on Eating and Weight Patterns (QEWP), which has been administered and validated in multiple clinical and community settings.^24,25

The VA Binge Eating Screener (VA-BES)²⁶

The VA-BES is a single-item, self-report screener for binge eating derived from an item on the MOVE!11, a screening tool for the MOVE! program. Participants respond to the following: On average, how often have you eaten extremely large amounts of food at one time and felt that your eating was out of control at that time? Response options are as follows: 0 (never), 1 (< 1 time/week), 2 (1 time/week), 3 (2–4 times/week), and 4 (5 + times/week). Participants were categorized by binge status: no binge eating (never) or any binge eating (1 or greater). The measure has been shown to have utility and validity as a single-item screener for binge eating in Veterans.²⁶

The Godin Leisure-Time Exercise Questionnaire (Godin)²⁷

The Godin assesses physical activity with three frequency items measuring strenuous, moderate, and mild physical activity (weighted nine, five, and three, respectively) over a typical week. We summed the items for an overall score in which higher scores reflect greater physical activity. The Godin is a widely used measure with good psychometric properties²⁸ and is validated for use in classifying activity levels in healthy adults.²⁹

RESULTS

Statistical Analysis

We analyzed the data with STATA version 14 and reported properties of the WE-QOL Scale, including its internal consistency (Cronbach’s alphas with > 0.7 considered good and > 0.9 considered excellent) as compared to the EQ-ED-5L. We used Pearson correlations for continuous variables and t tests for categorical variables to compare the strength of relationships between the WE-QOL composite score and the EQ-ED-5L index values and VAS, and demographic and clinical characteristics, as well as measures of binge eating, eating pathology, and physical activity.

The Weight and Eating Quality of Life (WE-QOL) Scale

The WE-QOL Scale had excellent internal consistency that was higher than that for the EQ-ED-5L (Cronbach’s alpha = 0.91 vs. 0.78). The physical activity and physical discomfort domains were severely impacted by eating and weight in approximately 30% of the sample each. All other domains (self-care, daily responsibilities, relationships, public distress, emotional distress and shame/guilt) were at least moderately impacted in over one-quarter of the sample each (see Table 2).

Analysis Comparing Quality-of-Life Measures

Overall, participants had a mean WE-QOL score of 1.30 (SD = 0.91; min, max = 0, 4), mean VAS score of 59.96 (SD = 21.95; min, max = 0, 95), and mean index value of 0.57 (SD = 0.32; min, max =  − 0.33, 1). Female Veterans had significantly worse QOL scores than men (M(SD) = 1.67 (1.07) vs. 1.23 (0.86); mean difference = 0.446, p = 0.009) for the WE-QOL, but not the VAS or index value. Table 3 displays findings for the relationships between the QOL measures and the measures of BMI, binge eating (VA-BES), eating behavior (QEWP-5), and physical activity (Godin). All three QOL measures were significantly related to BMI such that higher BMI was related to worse QOL (p’s ≤ 0.001). All three QOL measures were significantly related to binge frequency on the VA-BES (p ≤ 0.001), but not the QEWP-5 measure of binge frequency. All three QOL measures were significantly related to levels of upset by overeating (p’s ≤ 0.001) and upset over the loss of control eating (p’s ≤ 0.005) such that greater upset was related to worse QOL. Participants who responded affirmatively to four of the five eating pathology features on the WE-QOL (p’s < 0.002), three of the five features on the VAS (p’s < 0.05), and one of the five on the index value (p = 0.02) reported significantly worse QOL than those who denied those features. The WE-QOL and the VAS, but not the index value, were related to the Godin such that more physical activity was related to better QOL (p’s < 0.05).

Table 3.

Correlations of WE-QOL, VAS, and Index Value with Clinical Characteristics and Measures

	WE-QOL			EQ-ED-5L
				VAS			Index Value
	n	Test statistic*	p value	n	Test statistic*	p value	n	Test statistic*	p value
BMI (continuous)	207	0.329	< 0.001	204	− 0.229	0.001	202	− 0.234	0.001
Frequency of binge eating (VA-BES; 0–4)	210	0.351	< 0.001	206	− 0.222	0.001	205	− 0.225	0.001
QEWP-5
Binge episodes in last 3 months	62	0.100	0.440	62	− 0.108	0.401	61	0.143	0.27
Features
Feature: eaten rapidly (yes vs. no)	78	0.773	< 0.001	75	− 11.087	0.016	76	− 0.062	0.349
Feature: felt uncomfortably full (yes vs. no)	78	0.663	0.002	75	− 6.236	0.205	75	− 0.095	0.162
Feature: not physically hungry (yes vs. no)	77	0.363	0.087	74	− 12.336	0.007	76	− 0.025	0.704
Feature: embarrassment (yes vs. no)	78	1.025	< 0.001	75	− 10.529	0.076	76	− 0.152	0.072
Feature: disgust/guilty (yes vs. no)	77	0.748	< 0.001	74	− 14.953	0.001	75	− 0.150	0.020
Level of upset (0–4)
Level of upset overeating	190	0.498	< 0.001	186	− 0.256	< 0.001	185	− 0.270	< 0.001
Level of upset loss of control	190	0.408	< 0.001	186	− 0.259	< 0.001	185	− 0.204	0.005
Godin (continuous)	189	− 0.174	0.016	186	0.155	0.034	185	0.135	0.068

Higher scores for the WE-QOL indicate worse quality of life whereas higher scores for the VAS and index value indicate better quality of life

*Mean difference for binary variables/correlation for continuous variables

Bold values represent p-values significant at the 0.05 level n's differ across variables as some items are not applicable to all subjects

DISCUSSION

To address the need for a population health metric for weight- and eating-related services in the VHA and assess health-related QOL in Veterans with weight and disordered eating issues, we developed the brief, 8-item WE-QOL measure. The eight items reflect eight unique and overlapping constructs from established measures of QOL developed separately for weight and eating. This includes physical activity, self-care, daily responsibilities, relationships, public distress, physical discomfort, emotional distress, and shame/guilt. For descriptive purposes, the WE-QOL Scale provided a granular level of health-related information. Some degree of impairment was reported by approximately one-half or more of the sample in all eight domains except public distress. For physical activity and physical discomfort, approximately 30% of the sample reported severe or worse functioning in each of these domains, and over 10% reported that their physical discomfort could not be worse.

We also assessed the psychometric properties of the WE-QOL Scale and its construct validity in relation to a widely used general health-related QOL questionnaire, the EQ-ED-5L, and other established weight- and eating-related measures. The WE-QOL Scale demonstrated an excellent internal consistency in this sample that was superior to the EQ-ED-5L. The WE-QOL Scale was related to relevant outcomes such as BMI, binge eating, eating pathology, and physical activity. These associations were in line with, and in some cases stronger than, the more general QOL measures. Collectively these findings support the construct validity of this new brief, weight- and eating-specific QOL measure.

We note several limitations. First, we studied a convenience sample of Veterans seeking weight loss treatment at a single VHA facility. Thus, findings may not be representative of all Veterans and may not be generalizable to other clinical contexts. Analyses were correlational and as such directionality or causal implications cannot be implied. In addition, items in the proposed measure assessed the combined impact of weight and eating on QOL, but not their independent impact. However, this was specifically done for brevity and to have one measure for the VHA encompassing the range of eating (i.e., eating disorders) and weight issues. Finally, comparisons with long-form weight and eating disorder QOL measures would have been preferable, but we deemed these too burdensome for the clinical setting used in the present study.

Considering these findings, the WE-QOL Scale has potential for assessing QOL of Veterans in the context of weight management and perhaps other weight- and/or eating-related healthcare settings. Clinicians may use individual results from the WE-QOL Scale to tailor weight management and lifestyle change goals. Future research should assess the extent to which the WE-QOL Scale is sensitive to treatment change (for example, in the context of weight management) and its potential benefits for reducing weight stigma that is associated with weigh-ins. The WE-QOL Scale has potential for research and healthcare settings beyond VHA and for other conditions affecting obesity-related disease (e.g., cancer recurrence) for which a sole focus on change in BMI may be detrimental.³⁰

The current study supports the psychometrics of the WE-QOL Scale and the use of a brief and specific QOL measure for weight management settings. The WE-QOL Scale is well-suited for time-sensitive clinical settings and detects nuanced associations with specific QOL domains. The WE-QOL Scale has potential to standardize screening, referral, and evaluation across the range of VHA weight- and eating-related health services including, but not limited to, VHA’s weight management programs (MOVE!, pharmacotherapy, and bariatric surgery), nutrition services, whole health programming, and mental health/eating disorder care. Most importantly, the scale has potential to be used as a “pre-post” tool to see if after treatment and intervention, QOL improves, thus improving patient care and quality of care. Given the potential uses of the WE-QOL Scale, our findings provide support for the need for a more rigorous evaluation of its validity, reliability and sensitivity to treatment effects.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contribution:

All authors contributed important intellectual content to the manuscript and have approved the final version. Alison G. Marsh and Anastasia Bullock, PsyD, contributed to the manuscript.

Funding

This work was supported by the VA Health Services Research and Development Service (HSR&D) (grant number IIR 15–349) and HSR&D Pain Research, Informatics, Multimorbidities, and Education (PRIME) Center of (grant number CIN 13–407).

Data Availability:

The datasets generated and/or analyzed during the current study are not publicly available due to privacy protections regarding the use and distribution of VA data.

Declarations:

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee (Human Studies Subcommittee at the VA Connecticut Healthcare System) and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The article does not contain any studies with animals performed by any of the authors.

Conflict of Interest:

The authors declare that they do not have a conflict of interest.

Disclaimer:

The content of this research is solely the responsibility of the authors and does not necessarily represent the official views of the VA or the Veterans Health Administration.