Obesity Surgery, 13, 684-692

The Validation of the Moorehead-Ardelt Quality of

Life Questionnaire II

Melodie K. Moorehead, PhD1; Elisabeth Ardelt-Gattinger, PhD2; Hans

Lechner, PhD2; Horacio E. Oria, MD, FACS3

1
Drs. Moorehead, Parish & Associates, associated with the Center for Severe Obesity and US
Bariatric, Ft. Lauderdale, FL, USA; 2Institute for Psychology, University of Salzburg, Austria;
3
Spring Branch Medical Center, Houston, TX, USA

Background: The Moorehead-Ardelt Quality of Life females, mean BMI=50), participants of gastric
Questionnaire was originally developed as a disease- bypass support groups. Reliability of the M-A QoLQII
specific instrument to measure postoperative out- was determined using Cronbach’s alpha coefficient.
comes of self-perceived quality of life (QoL) in obese Construct validity was measured by conducting a
patients. 5 key areas were examined: self-esteem, series of Spearman rank correlations.
physical well-being, social relationships, work, and Results: A Cronbach’s alpha coefficient of 0.84 indi-
sexuality. Each of these questions offered 5 possible cated satisfactory internal consistency. The M-A
answers, which were given + or - points according to QoLQII was found to be significantly correlated
a scoring key. The questionnaire was used independ- (P<0.01) to 7 of the 8 SF-36 scales: Physical Role
ently or incorporated into the Bariatric Analysis and (r=0.357), Bodily Pain (r=-0.486), General Health
Reporting System (BAROS). The instrument is simple, (r=0.413), Vitality (r=0.588), Social Functioning
unbiased, user-friendly and can be completed in <1 (r=0.517), Emotional Role (r=0.480), and Mental Health
minute. It has been found useful, reliable and repro- (r=0.489). The questionnaire also significantly corre-
ducible in numerous clinical trials in different coun- lated (P<0.01) to the Beck Depression Inventory-II (r=-
tries. Further research and feedback from some of its 0.317), as well as to the ‘Disinhibition’ (r=-0.307) and
users prompted refinements, now included in the ‘Hunger’ (r=-0.254) factors of the Stunkard and
Moorehead-Ardelt Quality of Life Questionnaire II (M- Messick Eating Inventory.
A QoLQII). This study tested the validity of the Conclusions: The M-A QoLQII correlates well with
improved instrument. other widely used health and well-being indicators
Methods: The wording of the questions was such as the SF-36, Beck Depression Inventory II and
changed, to make them less suggestive and allow for the Stunkard and Messick Eating Inventory. The study
the use of the survey before and after medical inter- established the validity and reliability of this
vention, and with control groups. A 6th question, ana- improved disease-specific instrument for QoL meas-
lyzing eating behavior, was added. The ±1 point given urement in the obese population.
to the evaluation of self-esteem was split with this
new question, thus maintaining the consistency of Key words: Quality of life, questionnaire, validation, SF-
the scores. The drawings were simplified. Finally, the 36, bariatric surgery, obesity, morbid obesity, outcome
scoring key was changed to a 10-point Likert scale, to assessment, weight loss
improve response-differentiation. To validate the M-A
QoLQII, we examined its concordance with other
health and well-being indicators, specifically the MOS
36-Item Short-Form Health Survey (SF-36), the Beck Introduction
Depression Inventory-II (BDI-II) and the Stunkard and
Messick Eating Inventory. The study population
included 110 morbidly obese patients (20 males, 90 The original Moorehead-Ardelt Quality of Life
Questionnaire (M-A QoLQ) was created to be part
Reprint requests to: Melodie K. Moorehead, PhD, 1201 East
Broward Blvd., Fort Lauderdale, FL 33301, USA.
of the Bariatric Analysis and Reporting Outcome
Fax: 954-779-7994; e-mail: psydrmm@aol.com System (BAROS).1 The BAROS was initially devel-

684 Obesity Surgery, 13, 2003 © FD-Communications Inc.

 Quality of Life Questionnaire II Validation

oped by the senior author (HEO) in response to the MOOREHEAD-ARDELT: Quality of Life Questionnaire
1991 National Institutes of Health (NIH) Consensus Self Esteem and Activities Levels
Please print form: Make a check to show how your life has
Conference Statement, which identified the lack of changed after your weight loss.
standards for comparison of outcomes in the surgi- 1. Compared to the time before my weight loss treatment I feel. . . .
cal treatment of severe obesity as a key problem:
“Better statistical reporting of surgical results is
urgently needed for better assessment of outcomes”.
Furthermore, this panel stated, “Quality of life con- 2. I am able to participate physically in activities. . .
siderations in patients undergoing surgical treatment
of obesity must be addressed . . ”.2
The BAROS, shown in Figure 1, consists of a
scoring table that includes three columns with the 3. I am willing to be involved socially. . .
main areas of analysis: Weight Loss, Improvement
of Medical Conditions and Quality of Life. Points
are added or subtracted according to changes in 4. I am able to work . . .
these domains. A maximum of three points is given
to each domain to evaluate changes after medical or
surgical intervention. Points are deducted for com-
plications or reoperations. The total number of 5. I am interested in sex . . .
points defines five outcome groups from ‘failure’ to
‘excellent’. The original M-A QoLQ, seen in Figure
2, was designed on a single page, using simple
drawings to offer answer options in each of five Figure 2. The original Moorehead-Ardelt Quality of Life
important quality of life (QoL) domains: self- Questionnaire.
esteem, physical activity, social life, work condi-
tions, and sexual activity. This questionnaire’s scor-
ing key is shown in Figure 3. Using a European-
based study population, the questionnaire was found
to be both valid and reliable.3
The BAROS has been used by surgeons in the
U.S.A., as well as by many centers in Europe,
Brazil, Canada, Mexico, and other countries. This
outcome system has been adopted as a standard for
reporting results in Austria, Germany and all
German-speaking countries, as well as in Spain and
Brazil. The BAROS has been found to be easy to
use and effective in reporting standardized surgical
outcomes.4-11 Prestigious bariatric surgeons have
recommended that the system be adopted as a stan-
dard for evaluation of outcomes in obesity sur-
gery.12-14
In an attempt to remain open and responsive to
suggestions made from members of the surgical
community, the M-A QoLQ has been improved.15
The new questionnaire can now be used for pre, as
well as for post-intervention assessment (see
Figure 1. Bariatric Analysis and Reporting Outcome Appendix 1 at end, P. 691). It also allows for com-
System (BAROS). 1 parison with control groups. A question related to

Obesity Surgery, 13, 2003 685

Moorehead et al

Moorehead-Ardelt Quality of Life Questionnaire can be completed in less than 1 minute, a fact that
SCORING
contributes to a high response rate. Furthermore, it
1. SELF ESTEEM
does not require a structured interview or any assist-
ing or coaching. These points are particularly
important, because one of the most serious prob-
-1.0 -.50 0 +50 +1.0 lems in the field of bariatric surgery research is poor
2. PHYSICAL long-term follow-up data, even in the simple report-
ing of weight loss.16
To validate the M-A QoLQII, we examined its
concordance with other widely-used health and
-50 -.25 0 +.25 +.50
3. SOCIAL
well-being indicators, specifically the SF-36, the
Beck Depression Inventory-II, and the Stunkard and
Messick Eating Inventory.17-19 The study was under-
-.50 -.25 0 +.25 +.50 taken to establish the construct validity of the new
4. LABOR questionnaire. This means the degree to which an
instrument measures the concepts that it purports to
measure, and that the correlation with other tests
which measure the same construct is high.
-.50 -.25 0 +.25 +.50
5. SEXUAL

Methods
-.50 -.25 0 +.25 +.50

SCORING KEY
Study Population
-3 -2.5 -2 -.75 -.5 0 +.5+.75 +2 +2.5 +3 The validation study was conducted in patients par-
ticipating in gastric bypass support groups from two
surgical practices in Fort Lauderdale, FL, USA. A
Figure 3. Scoring key for the original M-A QoLQ.
total of 110 patients were included. All participants
were either preparing for or were interested in
food perception was added. Hence, six items are bariatric surgery. The tests were administered on a
now used for measuring a patient’s subjective monthly basis for a 6-month period following regu-
impression of QoL in the areas of: 1) general self- larly scheduled support group meetings, as well as
esteem, 2) physical activity, 3) social contacts, 4) in the privacy of their surgeons’ offices, during
satisfaction concerning work, 5) pleasure related to office appointments. Demographics of the study
sexuality, and 6) focus on eating behavior. The participants are shown in Table 1.
wording of the questions was also improved. This
change was intended to make the questions less sug-
Reliability
gestive and reduce any tendency towards “socially
desirable” responses. All the questions are now The reliability of the new instrument was deter-
equally weighted and a 10-point Likert scale is used mined by calculating the Cronbach’s alpha coeffi-
for scoring, as shown at end in Appendix 2, P. 692. cient. This coefficient is considered useful at a level
This change makes the instrument more sensitive 0.70, but a coefficient of 0.80 is statistically
and improves response-differentiation. Lastly, some stronger. The correlations are attenuated very little
of the color-illustrated graphic symbols were by measurement errors beyond this level.
changed to promote fewer “culture-sensitive” Comprehensive test-retest reliability studies were
responses, minimizing the cross-cultural and lin- conducted in Austria by two of the authors (EA-G
guistic factors that can influence the reliability of and HL). In these trials, the M-A QoLQII also
any instrument. The M-A QoLQII, like the original, demonstrated satisfactory reliability.

686 Obesity Surgery, 13, 2003

 Quality of Life Questionnaire II Validation

Table 1. Demographic data of the 110 patients, 20 sists of 21 items and is the most widely used instru-
males and 90 females ment for identifying depression. It uses depression
criteria in line with the Diagnostic and Statistical
Variable Mean Range
Manual of Mental Health Disorders - Fourth
Age (yrs) 42 19 to 65 Edition (DSM-IV).20
Weight (kg) 139 95-328 4) The MOS 36-Item Short-Form Health Survey
BMI (kg/m2) 50 32-92
(SF-36) was designed for use in clinical practice and
research, health policy evaluations, and general pop-
Protocol ulation surveys. It is a multi-item scale that assesses
eight health concepts: limitations in physical activi-
The research subjects were highly motivated to par-
ties because of health problems, limitations in social
ticipate in the study. All respondents were guaran-
activities because of physical or emotional prob-
teed anonymity, and the participants were provided
lems, limitations in usual role activities because of
with the following explanation:
THANK YOU for participating in our International study
physical health problems, bodily pain, general men-
helping to advance knowledge in the field of Bariatric tal health (psychological distress and well-being),
Surgery Psychology. By honestly and openly answering vitality (energy and fatigue) and general health per-
the following questions, you will help us learn more ception. It has been widely-used for decades and is
about the life-threatening disease of morbid/super obe- often considered in the U.S. to be the gold-standard
sity. This new information will help us to better provide to evaluate QoL by the medical community.21-24
sensitive services to our patients and recommendations
to their surgeons. We ask that you answer each and
every question as it reflects your thoughts or feelings Statistical Analysis
RIGHT NOW. Thank you for helping us give a voice to
those who suffer from this debilitating disease. The Construct validity was measured by conducting a
more the people of the world understand about this series of Spearman rank correlations, comparing the
medical condition, the less suffering others will have to new questionnaire to the above-mentioned health
endure. and QoL instruments. Correlations between the test
instruments were considered significant at P<0.01.
Questionnaires Utilized
1) The M-A QoLQII. This questionnaire is a one-
dimension-structure instrument that evaluates a Results
patient’s self-perception of QoL in six key areas.
The tool was specifically designed for use with A Cronbach’s alpha coefficient of 0.84 indicated
overweight, morbidly obese and super obese patient satisfactory internal consistency and demonstrated
populations that are seeking medical or surgical the instrument’s reliability. The intra-class correla-
intervention. As previously mentioned, the ques- tions were between 0.54 and 0.69, a satisfactory
tionnaire is easy to understand and can be complet- level. Table 2 displays the concordance of the M-A
ed in less than 1 minute. It is user-friendly, easy to QoLQII scores with the eight health concepts of the
complete and score, cross-cultural and, like the orig- SF-36. Scores were highly and significantly
inal, has the added value of generating a high (P<0.01) correlated with seven of the eight health
response rate. scales of the SF-36: limitations in social activities
2) The Stunkard and Messick Eating Inventory is an because of physical or emotional problems, limita-
instrument used to help recognize and treat eating tions in usual role activities because of physical
disorders. It assesses three dimensions of eating health problems, bodily pain, general mental health
behavior: cognitive control of eating, disinhibition, (psychological distress and well-being), vitality
and hunger. The Inventory consists of 51 questions (energy and fatigue) and general health perception.
and is designed to take approximately 15 minutes to No significant correlation was found, however, with
administer. the physical functioning scale of the SF-36.
3) The Beck Depression Inventory-II (BDI-II) con- The concordance of the questionnaire with other

Obesity Surgery, 13, 2003 687

Moorehead et al

Table 2. Correlation of the Moorehead-Ardelt Quality of scales. The ‘physical functioning’ concept was the
Life Questionnaire II and the Short Form 36 only scale of the SF-36 that failed to correlate sig-
nificantly. This failure may be explained by the dif-
SF-36 Scales Spearman r ficulty in measuring physical functioning by paper
and pencil testing. Furthermore, the M-A QoLQII is
Social Function 0.517*
Emotional Role 0.480* designed to measure the delight experienced when
Mental Health 0.489* engaged in physical functions rather than the actual
Physical Functioning 0.023 ability to function.
Physical Role 0.357* The SF-36 is a lengthy and generic health-related
Bodily Pain -0.486*
QoL measurement. Similar drawbacks exist with
General Health 0.413*
Vitality 0.588* other surveys, such as the 112-item Gothenburg
Quality of Life Scale.25 Researchers have developed
*P<0.01 disease-specific instruments to better study particu-
lar pathologies or populations.26-30 One of these, the
health and QoL indicators, the Stunkard and Impact of Weight on Quality of Life Questionnaire
(IWQoL), is dedicated to the obese.31 However, this
Messick’s Eating Inventory and the BDI-II, is
questionnaire includes many questions in different
reported in Table 3. The Eating Inventory demon-
areas and is time-consuming. The M-A QoLQII, on
strated significant (P<0.01) negative correlations on
the other hand, is a specially developed disease-spe-
two of the three factor scales, ‘Disinhitition’ and cific instrument to study the obese population, inde-
‘Hunger.’ A non-significant correlation was found pendently or in conjunction with the BAROS. It is
on the ‘Cognitive Restraint’ factor of the Eating simple, concise, easy to understand and answer, and
Inventory. There was a significant negative correla- requires minimal time, human and material
tion between the M-A QoLQII and the BDI-II, indi- resources to complete. These benefits increase the
cating that the greater the depression, the lower the response rate. The evaluation and scoring is simple,
individual’s perceived QoL. objective and short, reducing interviewer depend-
ence and eliminating evaluator’s bias. Therefore, the
instrument is very cost-effective. Furthermore, the
use of colored illustrations to assess the patient’s
Discussion perceived QoL in the six principal domains, makes
the questionnaire user-friendly, fun to complete, and
The high Cronbach’s alpha coefficient obtained in not culture-influenced. These are some of the rea-
these tests proved the internal consistency and relia- sons the questionnaire was rapidly accepted in dif-
bility of the questionnaire. The results of this study ferent countries and cultures.
also showed construct validity between our instru- Significant correlations were also demonstrated
ment and the SF-36. This was demonstrated by sig- with two of the three scales of the Stunkard and
nificant correlations with seven of the eight SF-36 Messick’s Eating Inventory, i.e. ‘Disinhitition’ and
‘Hunger’. There was, however, no significant corre-
lation between the QoL instrument and the
Table 3. Correlation of the MA QoLQII with the
Stunkard and Messick Eating Inventory and the Beck ‘Cognitive Restraint’ factor of this inventory. QoL is
Depression Inventory-II an emotional variable and may have little to do with
cognitive control. This lack of correlation with the
Scale Spearman r ‘Cognitive Restraint’ item may be because of the
fact that the M-A QoLQII is an instrument that
Stunkard & Messick ‘Disinhibition’ -0.307*
Stunkard & Messick ‘Hunger’ -0.254*
measures “affect”, or what one feels, rather than
Stunkard & Messick ‘Cognitive Restraint’ 0.173 “cognition”, or what one thinks. Further, this lack of
Beck Depression Inventory-II -0.317* correlation may also suggest that the ‘Disinhibition’
and ‘Hunger’ factors are the driving forces that pro-
*P<.01 mote inability to control eating behavior, thereby

688 Obesity Surgery, 13, 2003

 Quality of Life Questionnaire II Validation

lowering QoL. These findings may explain, in part, References

why non-surgical weight-loss programs based on
behavioral modifications (i.e. cognitive control) are 1. Oria HE, Moorehead MK. Bariatric Analysis and
doomed to fail in the morbidly obese. Reporting Outcome System (BAROS). Obes Surg
The Beck Depression Inventory-II demonstrated a 1998; 8: 487-499.
very significant negative correlation to the M-A 2. National Institutes of Health Consensus Development
QoLQII. This statistically significant negative corre- Conference Draft Statement. Gastrointestinal surgery
lation helps explain what we already know in clini- for severe obesity 25-27 March 1991. Obes Surg
cal work, that as a person loses control over their 1991; 1: 257-66.
QoL, depression scores go up. We know from many 3. Ardelt E, Moorehead MK. The validation of the
studies presented at the American Society for Moorehead-Ardelt Quality of Life Questionnaire.
Bariatric Surgery, that 90% or more of our morbid- Obes Surg 1999; 11: 13 (abstr 30).
ly obese patient population interested in bariatric 4. Favretti F, Cadière GB, Segato G et al. Bariatric
surgery, present with depression. It would be useful Analysis and Reporting Outcome System (BAROS).
to have a standarized tool for bariatric patients that Obes Surg 1998; 8: 500-4.
measures the increase or decrease in QoL as depres- 5. Hell E, Miller KA, Moorehead MK et al. Evaluation
sion scores rise or fall in preparation for or follow- of health status and quality of life after bariatric sur-
ing medical/surgical interventions. gery: comparison of standard Roux-en-Y gastric
bypass, vertical banded gastroplasty and laparoscopic
adjustable silicone gastric banding. Obes Surg 2000;
10: 214-9.
Conclusions 6. Wolf AM, Falcone AR, Kortner B et al. BAROS: an
effective system to evaluate the results of patients
The results of this clinical investigation indicate that after bariatric surgery. Obes Surg 2000; 10: 445-50.
the M-A QoLQII correlates well with a longtime 7. Oria HE, Brolin RE. Performance standards in
and widely-used QoL assessment instrument, the bariatric surgery. In: Deitel M, Cowan G, eds. Update:
SF-36, and other health indicators such as the Beck Surgery for the Morbidly Obese Patient. Toronto,
Depression Inventory-II and the Stunkard and Canada: FD-Communications Inc, 2000: 85-93.
Messick Eating Inventory. We have established the 8. Nguyen NT, Goldman C, Rosenquist CT et al.
internal consistency and reliability of the improved Laparoscopic versus open gastric bypass: a random-
questionnaire and completed the validation process ized study of outcomes, quality of life, and costs. Ann
initiated in previous European research. While con- Surg 2001; 11: 265-70.
ducting this study, the simplicity, ease-of-use and 9. Victorzon M, Tolonen P. Bariatric Analysis and
cost-effectiveness of this disease-specific instru- Reporting Outcome System (BAROS) following
ment for the evaluation of QoL outcomes in the laparoscopic adjustable gastric banding in Finland.
obese population was again noticed. Finally, these Obes Surg 2001; 11: 740-3.
results reasserted the wisdom of having incorporat- 10.Kalfarentzos F, Kechagias MD, Soulikia K et al.
ed the M-A QoLQ into the Bariatric Analysis and Weight loss following vertical banded gastroplasty:
Reporting Outcome System (BAROS), the unique Intermediate results of a prospective study. Obes Surg
instrument for the objective and complete evalua- 2001; 11: 265-70.
11.Nini E, Slim K, Scesa JL et al. Evaluation of laparo-
tion of outcomes in bariatric surgery.
scopic bariatric surgery using the BAROS score. Ann
We thank Drs. Norman Samuels (Center for Severe Obesity) and Chir. 2002; 127: 107-14.
Robert Marema (US Bariatric), for participation of their patients in 12.Hell E. IFSO – and the future of surgery. Presidential
this study. Thanks to Drs. George Cowan, Mervyn Deitel, Michael Address, 5th Congress of International Federation for
Parish, Cynthia Buffington and Anna Wolf for input and encour-
the Surgery of Obesity, Genoa, September 22, 2000.
agement, and to Drs. Hell, Miller, Favretti, Scopinaro, Garrido,
Baltasar, and others for promoting standardization of outcome Obes Surg 2000; 10: 495-7.
studies and enthusiasm in the acceptance of BAROS. 13.Cowan GSM Jr. Obligations of the bariatric surgeon.

Obesity Surgery, 13, 2003 689

Moorehead et al

Presidential Address, International Federation for the 24.O’Brien PE, Dixon JB, Brown W et al. The laparo-
Surgery of Obesity, Salzburg, September 24, 1999. scopic adjustable gastric band (Lap-Band® ): a
Obes Surg 2000; 10: 498-501. prospective study of medium-term effects on weight,
14.Scopinaro N. Outcome evaluation after bariatric sur- health and quality of life. Obes Surg 2002; 12: 652-
gery (Invited Commentary). Obes Surg 2002; 12: 253. 60.
15.Baltasar A. On BAROS (Correspondence). Obes Surg 25.Isacsson A, Frederiksen SG, Nilsson P et al. Quality
1999; 9: 288. of life after gastroplasty is normal: a controlled study.
16.Renquist KE, NBSR Data Contributors. Surgical Eur J Surg 1997; 163: 181-6.
treatment of obesity in America: Data according to the 26.Velanovich V. Comparison of generic (SF-36) vs dis-
National Bariatric Surgery Registry. Probl Gen Surg ease-specific (GERD-HRQL) quality-of-life scales
1992; 9: 231-8. for gastroesophageal reflux disease. J Gastrointest
17.Ware JE Jr, Sherbourne CD. The MOS 36-Item Short- Surg 1998; 2: 141-5.
Form Health Survey (SF-36). Med Care 1992; 30: 6. 27.Freys SM. Tigges H, Heimbucher J et al. Quality of
18.Stunkard AJ, Messick S. Eating Inventory Manual. life following laparoscopic gastric banding in patients
Toronto: The Psychological Corporation, 1986. with morbid obesity. J Gastrointest Surg 2001; 5: 401-
19.Beck, AT, Steer, RA, Brown, GK. Beck Depression 7.
Inventory – II. Toronto: The Psychological 28.Fernando HC, Schauer PR, Rosenblatt M et al.
Corporation, 1996. Quality of life after antireflux surgery compared with
20.Diagnostic and Statistical Manual of Mental nonoperative management for severe gastroe-
Disorders, 4th Edn. Washington, DC: American sophageal reflux disease. J Am Coll Surg 2002; 194:
Psychiatric Association, 1994. 23-7.
21.Temple PC, Travis B, Sachs L et al. Functioning and 29.Arcila D, Velazquez D, Gamino R et al. Quality of life
well-being of patients before and after elective surgi- in bariatric surgery. Obes Surg 2002; 12: 661-5.
cal procedures. J. Am Coll Surg 1995; 181: 17-25. 30.Schwimmer JB, Burwinkle TM, Varni JW. Health-
22.Fontaine KR, Cheskin LJ, Barofsky I. Health-related related quality of life of severely obese children and
qualityof life in obese persons seeking treatment. J adolescents. JAMA 2002; 289: 1813-9.
Fam Pract 1996; 43: 265-70. 31.de Zwann M, Mitchell JE, Howell LM et al. Two
23.Choban P, Onyejekwe J, Burge JC et al. A health sta- measures of health-related quality of life in morbid
tus assessment of the impact of weight loss following obesity. Obes Res 2002; 10: 1143-51.
Roux-en-Y gastric bypass for clinically severe obesi-
ty. J Am Coll Surg 1999; 188: 491-7. (Received June 21, 2003; accepted July 18, 2003)

690 Obesity Surgery, 13, 2003

 Quality of Life Questionnaire II Validation

Appendix 1.

Moorehead-Ardelt Quality of Life Questionnaire II

Self Esteem and Activity Levels
Please make a check in the box provided to show your answer.

Obesity Surgery, 13, 2003 691

Moorehead et al

Appendix 2. The M-A QoLQII scoring key.

Self Esteem and Activity Levels
SCORING KEY
1. Usually I feel . . .

-.50 -.40 -.30 -.20 -.10 +.10 +.20 +.30 +.40 +.50

2. I Enjoy Physical Activities . . .

-.50 -.40 -.30 -.20 -.10 +.10 +.20 +.30 +.40 +.50
3. I Have Satisfactory Social Contacts . . .

-.50 -.40 -.30 -.20 -.10 +.10 +.20 +.30 +.40 +.50
4. I Am Able to Work . . .

-.50 -.40 -.30 -.20 -.10 +.10 +.20 +.30 +.40 +.50

5. The Pleasure I get Out of Sex Is . . .

-.50 -.40 -.30 -.20 -.10 +.10 +.20 +.30 +.40 +.50
6. The Way I Approach Food Is . . .

-.50 -.40 -.30 -.20 -.10 +.10 +.20 +.30 +.40 +.50

692 Obesity Surgery, 13, 2003