Obesity Surgery

, Volume 21, Issue 10, pp 1585–1591

A 5-Year Prospective Quality of Life Analysis Following Laparoscopic Adjustable Gastric Banding for Morbid Obesity

Authors

    • Department of SurgeryTurku University Hospital
  • Paulina Salminen
    • Department of SurgeryTurku University Hospital
  • Harri Sintonen
    • Hjelt Institute/Department of Public HealthUniversity of Helsinki
  • Jari Ovaska
    • Department of SurgeryTurku University Hospital
  • Mikael Victorzon
    • Department of SurgeryVaasa Central Hospital
Clinical Research

DOI: 10.1007/s11695-011-0425-y

Cite this article as:
Helmiö, M., Salminen, P., Sintonen, H. et al. OBES SURG (2011) 21: 1585. doi:10.1007/s11695-011-0425-y

Abstract

Background

In addition to actual weight loss and the possible resolution of obesity-related co-morbidities following bariatric surgery, another widely recognized important outcome measure is the improvement of quality of life (QOL).

Methods

Disease-specific quality of life (DSQOL) and general health-related quality of life (HRQOL) were measured preoperatively and at 1 and 5 years postoperatively following laparoscopic adjustable gastric banding (LAGB) for morbid obesity. The Moorehead–Ardelt questionnaire was used for DSQOL assessments and a generic 15-dimensional questionnaire (15D) was used for HRQOL measurements. In addition, HRQOL was compared with that of the age- and gender-standardized general population.

Results

DSQOL scores were significantly improved on all domains after 1 year from the operation and this improvement was maintained at 5 years. This improvement was also seen in the total HRQOL scores. Despite this improvement, the HRQOL after LAGB remained worse compared to the age- and gender-standardized general population.

Conclusions

DSQOL and HRQOL improve both significantly after LAGB. This QOL improvement is maintained at 5-year follow-up although QOL does not reach the level of the general population.

Keywords

Morbid obesityDisease-specific quality of lifeHealth-related quality of lifeBariatric surgeryLAGBGastric banding15D

Introduction

Laparoscopic adjustable gastric banding (LAGB) was first introduced in 1993 [1] and has since then become a widely used procedure in bariatric surgery. Traditionally, weight loss and its effect on obesity-related co-morbidities have been the main postoperative outcome measures. However, in recent years, the importance of the improved quality of life (QOL) has been more acknowledged. QOL is likely to be impaired in obese patients [2, 3], and thus from their perspective, it is arguably one of the most important outcome measures of a weight-reducing procedure.

When measuring QOL, it is important to distinguish between the disease-specific quality of life (DSQOL) and the generic health-related quality of life (HRQOL) [4, 5]. The chosen generic HRQOL instrument should be well validated in the applied populations enabling the comparison of outcomes between different populations and interventions, particularly regarding cost-effectiveness studies. A DSQOL instrument can be a useful addition to general HRQOL assessment as it may be more sensitive in detecting and determining certain small changes in specific diagnostic groups that are important to clinicians or patients [6]. We have previously reported DSQOL outcome at 2 years after LAGB [7] and both HRQOL and DSQOL outcomes at 12 and 28 months after LAGB and compared the HRQOL with that of Finnish age norms [8]. The aims of this study were to prospectively assess the effect of LAGB on both DSQOL and HRQOL at 1 and 5 years and to compare the HRQOL score with that of the age- and gender-standardized general population.

Materials and Methods

From March 2000 to October 2003, 101 consecutive patients were enrolled for operation for morbid obesity at Vaasa Central Hospital in Finland by LAGB. The indications and contraindications for operative treatment were set according to standard recommendations and binge eating was considered a contraindication for LAGB. A prospective QOL study was started using the Moorehead–Ardelt questionnaire [9] as a DSQOL instrument. Approximately 1 year later, the 15D questionnaire [10] was added as an HRQOL instrument. The patients’ QOL was assessed preoperatively and postoperatively at 1 and 5 years.

Patient Selection

Permission for this study was granted from the ethical committee of Vaasa Healthcare District and informed consent was obtained from all of the 101 consecutive patients. The Moorehead–Ardelt questionnaire was offered to all patients, of whom 95 patients agreed to participate (94% response rate). In addition, the 15D questionnaire was offered to the last 79 patients, of whom 75 agreed to participate (95% response rate). The postoperative response rates for the Moorehead–Ardelt questionnaire were 71% and 62% and for the 15D questionnaire 73% and 85% at 1 and 5 years, respectively. Comparable 15D data, both at baseline and at 5 years follow-up, were available in 49 patients (65%).

The Moorehead–Ardelt Questionnaire

The Moorehead–Ardelt QOL questionnaire is a disease-specific one-page analysis that focuses on self-esteem, social, sexual and physical activity, and work capacity [9]. It is simple and patient-friendly and is frequently used in the evaluation of QOL following operative treatment for morbid obesity. It is, however, not well validated in the Finnish population.

The 15D Instrument

The 15D [10] is a generic, comprehensive, 15-dimensional, standardized, self-administered measure of HRQOL that can be used both as a profile and single index score measure. This instrument subscribes conceptually to the definition of health by the World Health Organization as being composed of physical, mental, and social well-being. It includes the following 15 dimensions: breathing, mental function, speech (communication), vision, mobility, usual activities, vitality, hearing, eating, elimination, sleeping, distress, discomfort and symptoms, sexual activity, and depression. Each dimension is divided into five levels, by which more or less of the attribute is distinguished. The valuation system of the 15D is based on an application of the multi-attribute utility theory [10]. A set of utility or preference weights, elicited from the Finnish adult population through a three-stage valuation procedure, is used to generate the level values on a 0–1 scale for all dimensions as well as in an additive aggregation formula the 15D score (single index number) over all the dimensions, also on a 0–1 scale (1 = full health, i.e., no problems of any dimension and 0 = being dead). A change of >|0.03| in the 15D score is clinically important in the sense that people can on average feel the difference. The reliability, validity, sensitivity, discriminatory power, and responsiveness to change have been well tested in the Finnish population [10, 11] and found to be very good. In terms of these important properties, the 15D is superior or at least equal to existing generally used single index score instruments such as EQ-5D, HUI3, and SF-6D [10, 1214].

General Population Sample

Another reason for choosing the 15D was that there are excellent data on the HRQOL of the Finnish population measured by the 15D, which came from the National Health 2000 Health Examination Survey [15]. This survey covered a representative sample of the Finnish population aged 30 years and over. Individuals in the same age range as our patients were selected for this analysis, and the population sample was weighted to reflect the age and gender distribution of the patients (n = 4,612 at 1 year, n = 5,320 at 5 years).

Statistical Analyses

The continuous variables are described as means ± SD. The Moorehead–Ardelt scores were analyzed using a repeated measurements analysis of variance and the Tukey–Kramer method was used to adjust the p values of pairwise comparisons of time points. The difference between the patients and the population sample in the mean 15D dimension level values and scores were tested using a two-tailed independent sample t test. The few missing items of data on any dimension of the 15D questionnaire were replaced by predictions from regression models with the other dimensions and age as explanatory variables. A p < 0.05 was considered statistically significant.

Results

The mean preoperative age of the patients was 42.4 years (range 23–66, SD 10.7) and 75% were female. The mean preoperative BMI was 46.3 kg/m2 (range 36.3–66.6, SD 6.3) and excess weight 60.0 kg (min 32.6, max 141.9; SD 22.2). Seventy-one of the patients (70%) suffered from at least one of the most common co-morbidities associated with morbid obesity (type 2 diabetes, hypertension, joint disorders, and sleep apnea). Four operations (4%) had to be converted from laparoscopic to open approach. Seven patients (7%) had revision laparoscopies mainly due to band leakage and two of them had revision operations twice. In addition to these reoperations, a total of 13 patients (13%) had to have their band removed during this 5-year follow-up. The mean excess weight loss (EWL) was 37.3% (SD 16.5) at 1 year and 57.9% (SD 31.1) at 5 years postoperatively.

DSQOL scores were significantly improved on all five domains of the Moorehead–Ardelt questionnaire at 1 year from the operation. The DSQOL scores did not change significantly after that on any domains at 5 years postoperatively (Table 1). There was a significant improvement in HRQOL on the dimensions of moving, breathing, sleeping, usual activities, depression, distress, vitality, and sexual activity and a significant worsening on the dimension of eating at 1-year follow-up compared to preoperative values. At 5-year follow-up, these significant improvements remained despite the dimensions of sleeping and distress. The significant worsening of eating was not present at 5-year follow-up. At 1 year after surgery, the HRQOL had improved in a statistically significant and clinically important manner as indicated by the total 15D score. This improvement was maintained at 5 years postoperatively, and there was no statistically significant difference in the total 15D scores between the 1- and 5-year assessments. Despite these improvements, the HRQOL remained in the operated group all along at a lower level compared with the age- and gender-standardized general population (Table 2, Fig. 1). Due to differences in the composition and size of the groups at different time points, the 15D scores slightly differ at baseline and 5 years (Table 2).
Table 1

Moorehead–Ardelt scores

Dimensions

Score range

Preoperative group (n = 95)

1-year postoperative group (n = 73)

5-year postoperative group (n = 63)

p valuesa

Self-esteem

−1 to +1

−0.01 (0.39)

0.44 (0.37)

0.40 (0.43)

<0.01, <0.01, 0.43

Physical

−0.5 to +0.5

0.14 (0.23)

0.29 (0.19)

0.30 (0.19)

<0.01, <0.01, 0.96

Social

−0.5 to +0.5

−0.23 (0.23)

0.07 (0.23)

0.13 (0.25)

<0.01, <0.01, 0.22

Labor

−0.5 to +0.5

−0.04 (0.26)

0.15 (0.25)

0.19 (0.25)

<0.01, <0.01, 1.00

Sexual

−0.5 to +0.5

−0.03 (0.30)

0.09 (0.24)

0.08 (0.33)

0.04, 0.05, 0.87

Total score

−3 to +3

−0.17 (1.08)

1.04 (0.90)

1.10 (1.16)

<0.01, <0.01, 0.94

Values are means (SD), 95% CI

ap values derived from the differences between the preoperative group and the 1-year postoperative group, between the preoperative group and the 5-year postoperative group, and between the two postoperative groups, in that order

Table 2

15D scores before and at 1 and at 5 years after LAGB

Patients (n = 55) preoperative (=0), 1 year (=1), and 5 years (=5) postoperative

Pairs

Group

Number

15D score, mean

SD

SE, mean

p valuesa

Move 0

   

0.8138

0.15434

0.02081

<0.001

Move 1

   

0.9234

0.14528

0.14528

 

Move 0

   

0.8310

0.15170

0.02167

<0.001

Move 5

   

0.9248

013749

0.01964

 

See 0

   

0.9068

0.13117

0.01769

0.557

See 1

   

0.9186

0.12926

0.01743

 

See 0

   

0.9279

0.11711

0.01673

0.347

See 5

   

0.9102

0.12120

0.01731

 

Hear 0

   

0.9727

0.07875

0.01062

0.622

Hear 1

   

0.9674

0.10050

0.01355

 

Hear 0

   

0.9847

0.06063

0.00866

0.083

Hear 5

   

0.9694

0.08290

0.01184

 

Breath 0

   

0.6475

0.23352

0.03149

<0.001

Breath 1

   

0.8275

0.19792

0.02669

 

Breath 0

   

0.6891

0.20999

0.03000

<0.001

Breath 5

   

0.8911

0.18843

0.02692

 

Sleep 0

   

0.7433

0.20923

0.02821

0.008

Sleep 1

   

0.8217

0.19059

0.02570

 

Sleep 0

   

0.7657

0.20328

0.02904

0.747

Sleep 5

   

0.7751

0.19169

0.02738

 

Eat 0

   

1.0000

0.00000

0.00000

0.044

Eat 1

   

0.9743

0.09272

0.01250

 

Eat 0

   

1.0000

0.00000

0.00000

0.159

Eat 5

   

0.9856

0.07073

0.01010

 

Speech 0

   

0.9946

0.04001

0.00539

0.568

Speech 1

   

0.9892

0.05605

0.00756

 

Speech 0

   

0.9939

0.04239

0.00606

(a)

Speech 5

   

0.9939

0.04239

0.00606

 

Elimination 0

   

0.8643

0.19435

0.02621

0.957

Elimination 1

   

0.8660

0.20545

0.02770

 

Elimination 0

   

0.8975

0.16099

0.02300

0.073

Elimination 5

   

0.8412

0.19985

0.02855

 

Usual acts 0

   

0.7838

0.21874

0.02950

<0.001

Ususal acts 1

   

0.9112

0.18418

0.02483

 

Usual acts 0

   

0.8319

0.19452

0.02779

<0.001

Ususal acts 5

   

0.9476

0.14016

0.02002

 

Mental function 0

   

0.9222

0.14864

0.02004

0.853

Mental function 1

   

0.9173

0.16178

0.02182

 

Mental function 0

   

0.9418

0.13317

0.01902

0.208

Mental function 5

   

0.9072

0.16878

0.02411

 

Discomfort 0

   

0.7500

0.19002

0.02562

0.259

Discomfort 1

   

0.7917

0.23237

0.03133

 

Discomfort 0

   

0.7624

0.17314

0.02473

0.625

Discomfort 5

   

0.7768

0.19149

0.02736

 

Depression 0

   

0.8257

0.18748

0.02528

0.009

Depression 1

   

0.8841

0.13618

0.01836

 

Depression 0

   

0.8480

0.16906

0.02415

0.047

Depression 5

   

0.8942

0.12835

0.01834

 

Distress 0

   

0.8424

0.20282

0.02735

0.022

Distress 1

   

0.9061

0.14857

0.02003

 

Distress 0

   

0.8643

0.17380

0.02483

0.565

Distress 5

   

0.8789

0.17241

0.02463

 

Vitality 0

   

0.7390

0.20771

0.02801

<0.001

Vitality 1

   

0.8791

0.13956

0.01882

 

Vitality 0

   

0.7760

0.18202

0.02600

0.002

Vitality 5

   

0.8566

0.16736

0.02391

 

Sexual activity 0

   

0.7173

0.21549

0.02906

<0.001

Sexual activity 1

   

0.9058

0.15605

0.02104

 

Sexual activity 0

   

0.7469

0.20667

0.02952

0.001

Sexual activity 5

   

0.8729

0.20717

0.02960

 

D15 total score 0

  

55

0.8362

0.09541

0.01286

<0.001

D15 total score 1

  

55

0.8998

0.08269

0.01115

 

D15 total score 0

  

49

0.8590

0.08119

0.01160

<0.001

D15 total score 5

  

49

0.8986

0.08714

0.01245

 

D15 total score 1

  

40

0.9090

0.08258

0.01306

0.262

D15 total score 5

  

40

0.8957

0.09190

0.01453

 

aTwo-tailed t test. (a) The correlation and t cannot be computed because the standard error of the difference is 0

https://static-content.springer.com/image/art%3A10.1007%2Fs11695-011-0425-y/MediaObjects/11695_2011_425_Fig1_HTML.gif
Fig. 1

15D profiles of patients before and at 1 and at 5 years after LAGB compared with the age- and gender-standardized general population

Discussion

As previously shown, patients preoperatively were significantly worse off than the age- and gender-standardized general population on 11 of the 15 dimensions [16]. The mean 15D score among patients on the waiting list for bariatric surgery was 0.844 compared to 0.935 in the age- and gender-standardized general population presenting a clinically important difference. Surgery is, beyond any doubt, the only effective treatment for severe and morbid obesity [17], and many studies have demonstrated that the QOL will improve significantly after surgical treatment [1823]. At least one study has demonstrated superior QOL improvements to conservative treatment even 10 years later [21]. Our results confirm these earlier findings and show that both DSQOL and HRQOL improve significantly 1 year after LAGB. This HRQOL improvement is maintained at 5-year follow-up although the HRQOL does not reach the level of the age- and gender-standardized general population (Fig. 1). On the other hand, QOL did not continue to improve anymore between 1 and 5 years postoperatively even though excess weight loss proceeded. The weak correlation between EWL and changes in QOL has been demonstrated before [8, 24]. This has been attributed to a decline in the frequency of medical consultations and visits as more time from the operation passes [2528]. Although there was no significant difference in the total 15D score at 1- and 5-year follow-up (Table 2), significant improvements on the dimensions of sleeping and distress seen at 1-year follow-up could not be shown at 5 years. Patients were also worse off on the dimension of eating at 1-year follow-up compared to patients at baseline, a difference no longer present at 5 years, perhaps indicating decreasing functioning of the band or merely adaption to the circumstances by the patient. At 5 years, the difference in the HRQOL scores between patients with and without surgery would be actually greater than shown here (Fig. 1). In this patient group, a 5-year aging would decrease the 15D score by 0.018. If the effect of aging is taken into account, the comparison in the 15D score at 5 years would be between 0.841 and 0.899 (Table 2).

One of the main drawbacks of this study is the missing data on the possible improvement or resolution of the obesity-related co-morbidities postoperatively. This is mainly due to insufficient recording of the information at the postoperative follow-up, and unfortunately, this information could not be obtained retrospectively from the hospital records. To what extent the improvements in QOL seen in this study are a result of resolution or improvement in co-morbid conditions or merely due to EWL cannot be determined based on our data. The percentages of excess weight loss at 2 and 5 years are comparable with what has been published before [29, 30]. As EWL generally has a well-known positive impact on associated co-morbidities and metabolic disorders, one can assume that the improvements in QOL seen in this study are a result of both factors. However, the HRQOL did not reach the level of that of the age- and gender-standardized general population. We know now that with longer follow-up more and more bands will eventually fail. Bariatric surgery was started at Vaasa Central Hospital in 1996. As seen from the small numbers in this study, the volume in the beginning was modest, and mainly laparoscopic gastric bandings were performed. Since 2006, the numbers have risen to 150–200 operations per year, and a clear shift to more invasive procedures, mainly laparoscopic gastric bypass, has occurred. The main goal of any bariatric operation would be no less than to improve QOL to the same level as that of the general population. In conclusion, our results show that laparoscopic gastric banding significantly improves HRQOL up to 5 years postoperatively, but HRQOL remains at a lower level than that of the age- and gender-standardized general population.

Conflict of Interest

Harri Sintonen is the developer of the 15D. Otherwise, the authors declare that they have no conflict of interest.

Copyright information

© Springer Science + Business Media, LLC 2011