| The Larocque Obesity Questionnaire: A New Measure of the Psychological Factors in Weight Control Published in The American Journal of Bariatric Medicine, Fall 2000,Vol.15, N.3. Maurice Larocque, M.D. & Stephen C. Stotland, Ph.D . Editor's Note: Dr. Larocque, an ASBP member, operates the MLA Obesity Clinic in Montreal. Dr. Stotland is on the staff of the Cognitive-Behavior Therapy Service of Montreal General Hospital. Introduction The treatment of obesity requires the alteration of underlying maladaptive behavior, namely excessive food intake and insufficient activity. The success of dietary and behavioral approaches to the treatment of obesity has, however, been quite limited. A disturbing percentage of patients either drop out of treatment at an early stage, fail to reach a healthy weight, or regain the weight that they do lose.1,2 Clearly, an effective treatment for obesity must address the psychobiological mechanisms responsible for the failure of behavioral self-management.3,4 However, to date research has failed to uncover any psychological variables consistently associated with obesity,5 although there is evidence that patients often show mood variability and personality changes during weight loss and maintenance that may have clinical significance.6,7 Furthermore, many clinicians believe that emotional factors are central to weight control, particularly as it relates to the long-term maintenance vs. relapse of behavior change.8 The present paper reports on a new questionnaire that measures a number of the behavioral and psychological variables that may be important in weight control. The Larocque Obesity Questionnaire was designed to be clinically useful - comprehensive, brief and easily administered by computer. These are extremely important considerations in clinical settings, and particularly in the primary care setting, where physicians see many patients suffering from obesity and where time and resources are scarce. The Larocque Obesity Questionnaire provides a time efficient and inexpensive tool for the assessment of problematic behaviors and emotions that may interfere with weight management efforts. The questionnaire also provides an individualized report indicating areas of concern and suggesting various targets for intervention. This questionnaire has been used for the past decade in a number of clinical settings with approximately 300,000 patients, but analyses of its reliability and validity have not been previously presented. We report here on a study that examined the statistical properties of the Larocque questionnaire and its relationship to other psychological measures. Method Subjects Participants included 78 female patients, 18 years and older, who were engaged in weight loss treatment provided by physicians specializing in the treatment of obesity. Treatment involved low-calorie diets and exercise plans, and regular contacts with physicians dealing with medical or psychological problems relevant to weight control. The average subject was 37.0 years of age (+/- 8.7), and 183.4 (+/- 37.8) pounds. Only patients with Body Mass Index above 25 were included in the study. Procedure Subjects completed the Larocque Obesity Questionnaire (LOQ) and a variety of other self-report scales at varying points in treatment. Some completed the questionnaire prior to their first treatment session, while others did so at later points in treatment. The LOQ is a 52-item self-report scale, computer administered and scored, requiring about 10 minutes to complete. The items were designed to measure a variety of behavioral, emotional and personality variables that may be associated with problems in weight control. Variables were chosen based on the authors' clinical experience in the treatment of obesity. Items inquire about problematic eating habits, emotional eating, eating as reward, weight loss goals, feelings of depression, boredom and guilt, stress responses, and traits such as aggressivity, passivity and paranoia. A previous factor analytic study with 680 obese women (Larocque & Stotland, 1992, unpublished data) suggested that the LOQ may be divided into four subscales, which we have labeled Habits, Weight Loss Motivation, Physical Stress Reactions, and Negative Emotions. Sample items from the four subscales are presented in Table 1.
The Eysenck Personality Questionnaire9
is a 90-item scale measuring Neuroticism, Extraversion and Psychoticism. The scale has been widely
used and has shown high levels of internal consistency and test-retest reliability. The Neuroticism and
Extraversion scales were used in the present study. Neuroticism is described as a measure of the tendency
to worry, moodiness and depression, and psychomotor complaints. Extraversion is thought to reflect sociability,
need for excitement and change, and a relaxed, easygoing manner. The 13-item Beck Depression Inventory10
measures symptoms and signs of depression. Correlations of the BDI-13 with the full length BDI range from
.89 to .97, and alpha coefficients consistently exceed .85. The Body Esteem Scale11
consists of three moderately intercorrelated dimensions related to self-appraisals of Sexual Attractiveness,
Weight Concern, and Physical Condition. A total score reflecting Global Body Esteem was used in the present
study. The Dutch Eating Behavior Scale12
is a widely used instrument assessing tendencies to diet (Cognitive Restraint) and to overeat in response
to emotional states (Emotional Eating). Both scales have high levels of internal consistency and show
expected relations with variables such as caloric intake and binge eating. To examine the discriminant
validity of the LOQ scales we looked at 1 intercorrelations of the LOQ subscales and 2 relations
between LOQ subscales and other variables. Results The reliability of LOQ
subscales was assessed using Chronbach's alpha, an indication of the internal consistency of
subscale items. The 17-item Habits subscale assesses a broad range of eating and exercise behaviors,
thus the observed alpha coefficient of .67 was expected. The 9-item Physical Stress Reactions subscale
had an alpha coefficient of .68, reflecting its broad coverage of stress-related symptoms. The 19-item
Negative Emotions subscale had an alpha level of .85, suggesting that it measures a fairly coherent construct.
The 5-item Weight Loss Motivation subscale had a relatively low alpha coefficient of .46. This low level of
internal consistency may reflect the small number of items, or the multidimensionality of the weight loss motivation construct.
However, the sum of the motivation items was considered a useful measure of global weight loss motivation, and was retained
for further analyses. The correlations among LOQ subscales are presented in Table 2.
The Habits subscale was significantly related to Weight Loss Motivation (r=.37, p.01), indicating that patients reporting healthier
habits also reported higher weight loss motivation, and to Negative Emotions (r=-.29, p.01), suggesting that better habits were associated
with lower levels of negative affect. The strongest correlation was between Negative Emotions and Physical Stress Reactions (r=.57, p.001),
consistent with typical findings linking self-reports of negative affect and stress reactions. In general, the LOQ subscales were weakly
or moderately correlated, apart from the emotions/stress association just mentioned. It therefore appears that the LOQ subscales represent
different psychological constructs. This was explored further in the analysis of relations between LOQ scales and other psychological measures.
Table 2
Intercorrelations among Larocque Obesity Questionnaire Subscales
* p<.05** p<.01***p<.001 The Habits subscale was significantly related to a number
of the psychological variables, including body esteem (r=.31, p.01), cognitive restraint of eating (r=.46, p.001), emotional eating
(r=-.41, p.001) and BDI scores (r=-.32, p.01). These results suggest that subjects with higher Habits scores tended to have higher
body esteem and cognitive restraint of eating, and lower emotional eating and depressive tendencies. Weight Loss Motivation was
related to BDI scores (r=-.26, p.01) and weight (r=-.26, p.01), indicating that weight loss motivation scores were associated with
lower levels of depression and lower weight. Physical Stress Reactions were related to body esteem
(r=-.39, p.001), BDI scores (r=.63, p.001) and neuroticism (r=.62, p.001), suggesting that reports of
stress reactions were strongly related to negative affectivity and negative self-perceptions. Finally,
Negative Emotions showed a similar pattern of associations as the Physical Stress Reactions subscale,
with significant correlations to body esteem (r=-.27, p.01), BDI scores (.66, p.001) and neuroticism (r=.76, p.001).
Negative Emotions was also associated with higher scores on Emotional Eating (r=.42, p.001).
Table 3
Relations between Larocque Obesity Questionnaire Subscales and Other Measures
* p<.05** p<.01***p<.001 Discussion The LOQ appears to be a reliable measure
of several psychological and behavioral variables associated with obesity. This conclusion was supported by
the internal consistency and convergent validity analyses. Results suggest that the LOQ measures a number of
coherent and discriminably different variables, with predictable associations with other measures. Whether or
not the variables measured by the LOQ are related to important clinical outcomes (e.g., weight loss or maintenance)
is the subject of an ongoing investigation 13. One of the LOQ factors, the Weight Loss
Motivation subscale, appears to have rather low internal consistency. As suggested above, this may be interpreted
as a reflection of the breadth of the weight loss motivation construct, since not everyone wants to lose weight for
the same reasons, and it is not clear that the various sources of motivation are additive. In any case, it is clear
that further work in the exploration and measurement of this variable is urgently needed, given its theoretical and
potential practical importance. The variables measured by the LOQ are not novel.
Validated measures of most of these factors exist elsewhere.6 However, the real value of the LOQ is that it measures
these psychological factors in a convenient and thus clinically accessible format. The study of processes mediating long-term weight
control is of great importance, given the health consequences of obesity and the failure of currently available treatments.
While the obstacles to effective weight control are certainly complex, involving a variety of biopsychosocial variables,
there is little doubt that psychological factors feature prominently in the equation. The results of the present study
indicate that the Larocque Obesity Questionnaire may be a useful tool in the continuing effort to understand the psychological
processes in weight control. REFERENCES
1. French S, Jeffery R, Murray D. Is dieting
good for you? Prevalence, duration and associated weight and behavior changes for specific weight loss
strategies over four years in US adults. Int J Obes. 1999;23:320-327.
2. Wing R. Behavioral approaches to the treatment of obesity. In G. Bray, C. Bouchard, & P. James (Eds.),
Handbook of Obesity. 1997; New York: Marcel Dekker.
3. Foster, G., Wadden, T., Swain, R., Stunkard, A., Platte, P., & Vogt, R. The eating inventory
in obese women: Clinical correlates and relationship to weight loss.
Int J of Obes and Rel Metab Dis. 1998;22, 778-785.
4. McGuire M, Wing R, Klem, Lang W, Hill J. What predicts weight regain in a group of successful weight losers?
J of Consult and Clin Psych . 1999;67, 177-185.
5. Friedman M, Brownell K. Psychological correlates of obesity: Moving to the next research generation.
Psych Bulletin . 1995;17, 3-20.
6. Wadden T, Foster G. Behavioral assessment and treatment of markedly obese patients. In T. Wadden & T. VanItallie (Eds.)
Treatment of the seriously obese patient. 1992; New York: Guilford.
7. Williams P, Surwit R, Babyak M, McCaskill C. Personality predictors of mood related to dieting.
J of Consult and Clin Psych. 1998;66, 994-1004.
8. Brownell K. Relapse and the treatment of obesity. In T. Wadden & T. VanItallie (Eds.)
Treatment of the seriously obese patient. 1992. New York: Guilford.
9. Eysenck H, Eysenck S.
Manual of the Eysenck Personality Questionnaire. 1975, San Diego: EdITS.
10. Beck A, Ward C, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression.
Arch of Gen Psychiat.1961;4, 561-571.
11. Franzoi S, Shields J. The body esteem scale: Multidimensional structure and gender differences in a college population.
J of Personality Assessment.1984
12. Van Strein T, Frijters J, Bergers G, & Defares P. (1986). Dutch eating behavior questionnaire for the assessment of
restrained, emotional and external eating behavior.
Int J of Eat Dis. 1986;5, 295-315.
13. Larocque M, Gougeon R. (1999). Primary care treatment of obesity: Strategy for long-term weight maintenance.
American Journal of Bariatric Medicine, 1999;14, 16-20.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||