Assessment of self‐rated health 5 years after Roux‐en‐Y gastric bypass for severe obesity

Background Patients' perceptions of health change after bariatric surgery are complex. The aim of this study was to explore whether self‐rated health (SRH), a widely used tool in public health research, might be relevant as an outcome measure after Roux‐en‐Y gastric bypass (RYGB) for severe obesity. Methods This was a single‐centre retrospective study of a local quality registry. SRH score was registered at baseline and 5 years after RYGB. SRH, one of the 36 items in the quality‐of‐life Short Form 36 (SF‐36®) questionnaire, is the answer to this single question: ‘In general, would you say your health is excellent (1), very good (2), good (3), fair (4) or poor (5)?’ Change in SRH was analysed in relation to change in weight, co‐morbidities and quality of life after 5 years. Results Of a total of 359 patients who underwent RYGB between September 2006 and February 2011, 233 (64·9 per cent) reported on SRH before and 5 years after surgery. Of these, 180 (77·3 per cent) were women, and the mean(s.d.) age was 40(9) years. Some 154 patients (66·1 per cent) reported an improvement in SRH, 60 (25·8 per cent) had no change, and SRH decreased in 19 patients (8·2 per cent). SRH in improvers was related to better scores in all SF‐36® domains, whereas SRH in non‐improvers was related to unchanged or worsened scores in all SF‐36® domains except physical function. Conclusion Two‐thirds of patients reported improved SRH 5 years after RYGB for severe obesity. In view of its simplicity, SRH may be an easy‐to‐use outcome measure in bariatric surgery.


Introduction
The patient's experience of improvement in general health is the ultimate goal for all medical treatment. The perception of health has several aspects, and the WHO defines health as 'a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity' 1 . As severe obesity and bariatric surgery affect all of these aspects, and the most important factors motivating patients to consider seeking bariatric surgery are physical health and longevity 2,3 , measuring weight change alone seems insufficient to evaluate the global effect of this treatment.
Generic as well as disease-specific tools have been used to evaluate change in quality of life (QoL) after bariatric surgery 4 . Generally, these measures are comprehensive and time-consuming, and more useful in research than in clinical settings. An association between improvement in QoL and objective improvement in health has not been documented.
Self-rated health (SRH) is a person's subjective evaluation of their general health, expressed as the answer to the question: 'In general, would you say your health is excellent, very good, good, fair, or poor?' 5 -8 . In public health surveys and sociological research, SRH has been the most widely used health indicator since the 1950s 9 . Owing to its simplicity, SRH has proved to be a more valid and powerful predictor of morbidity, mortality and healthcare use than more comprehensive self-reporting instruments and objective biometric measures predicting future health 10 -12 . Interpreted as a spontaneous subjective assessment of a person's health status, SRH is regarded as the most precise measure of actual experienced health 13 . Public health surveys from different countries and social contexts have documented a relationship between SRH and genetic factors, inflammation and allostatic load, indicating a connection between SRH and biological processes 14 -19 . People suffering from severe obesity report lower SRH than the non-obese, even in the absence of chronic disease 20 . However, despite many advantages, bariatric surgery also has some adverse effects 21 -24 . As a general measure of perceived health, SRH might express the sum of positive and negative aspects of life as experienced by patients in the aftermath of the surgical procedure 25 . Nevertheless, there appear to be no publications on change in SRH after bariatric surgery.
The present study explored whether SRH, a patientreported, simple and robust instrument from public health research, is applicable as an outcome measure in bariatric surgery. The primary aim of the study was to evaluate change in SRH from before to 5 years after  Roux-en-Y gastric bypass (RYGB) for severe obesity. The secondary aim was to explore the relationship between change in SRH to weight loss, co-morbidity and change in QoL.

Methods
This study is a retrospective analysis of patients who had RYGB at Aalesund Hospital, a public, non-academic, secondary referral centre covering a population of 260 000 in Norway. The indication for RYGB was a BMI above 40 kg/m 2 or a BMI above 35 kg/m 2 with obesity-related co-morbidity in an adult population. The SRH response was collected as part of the Short Form 36 (SF-36 ® ; QualityMetric, Lincoln, Rhode Island, USA) questionnaire about 1 month before the operation, at the end of a preoperative education day 26 . Answers had no influence on the decision regarding whether the patient would have the operation or not. SRH is the first question of the SF-36 ® , and the version used in this study was the Norwegian translation of the question and alternative answers: 'In general, would you say your health is (1) excellent, (2) very good, (3) good, (4) fair or (5) poor?'.
Data for all patients who had RYGB at Aalesund Hospital between September 2006 and February 2011 were collected prospectively in a local quality registry, and data from routine visits at 6 weeks and 6, 12, 18, 24, 36, 48 and 60 months after surgery were updated to January 2018. Participation in postoperative support groups, adverse events, plastic surgery and new symptoms related to the bariatric procedure were also registered.
The difference between baseline SRH scores and scores at 5 years was calculated, and the change in SRH was categorized as improvement, no change, or a decrease.
Weight development from baseline through 5 years was reported by standard measures: percentage excess weight loss (%EWL), percentage excess BMI loss (%EBMIL), percentage total weight loss (%TWL) and change in BMI 27 . Weight regain, from nadir weight occurring between 1 and 2 years after surgery to 5 years, was reported
as change in BMI and percentage of maximum weight loss 28 . The study was approved by the Regional Ethics Committee (REK 2016/331) and by the local Data Protection Officer.

Statistical analysis
Categorical variables are given as proportions. All but one of the continuous variables (SF-36 ® physical function sum-score) were normally distributed and are given as mean(s.d.) values. SRH acts as a categorical as well as a continuous variable. Pearson's χ 2 test was performed for comparison of categorical variables, and independent and paired t tests were performed for comparison of continuous variables. Multiple logistic regression analysis was used to explore whether baseline variables could predict changes in SRH. P < 0⋅050 was considered statistically significant for all analyses. All analyses were performed using IBM SPSS ® version 23 (IBM, Armonk, New York, USA).

Results
A total of 359 patients underwent laparoscopic RYGB as a primary bariatric procedure between September 2006 and February 2011. At baseline, 339 patients completed the SF-36 ® questionnaire. After the operation, 322 patients (89⋅7 per cent) attended the 5-year follow-up visit, of whom 242 completed an identical questionnaire. There were complete baseline and postoperative SF-36 ® data, as well as clinical information on weight, co-morbidity, complications and blood test results, for 233 patients, representing 64⋅9 per cent of patients undergoing RYGB at this hospital in the study period.

Fig. 3 Change in SF-36 ® domains among non-improvers and improvers from baseline to 5 years after Roux-en-Y gastric bypass for severe obesity
In terms of individual changes in SRH, 154 patients (66⋅1 per cent) had a better SRH score at 5 years, 60 (25⋅8 per cent) had no change, and 19 (8⋅2 per cent) had a decrease in SRH score (Fig. 1). As the number with decreased SRH was low, the variable 'change in SRH' was dichotomized to improvers and non-improvers by merging the no change and decrease categories.
One of the success criteria for bariatric surgery is the achievement of a postoperative BMI of less than 35 kg/m 2 . In total, 179 patients (76⋅8 per cent) had a BMI of 35 kg/m 2 or less at 5 years. There was no significant relationship between BMI below or above 35 kg/m 2 and change in SRH (P = 0⋅377).
Another criterion of success is %EWL of 50 per cent or more, which occurred in 124 (80⋅5 per cent) of improvers and 58 (73 per cent) of non-improvers at 5 years (P = 0⋅215). In multiple logistic regression analysis, none of the baseline variables predicted change in SRH (data not shown).
At baseline, improvers had worse sum-scores than non-improvers for all SF-36 ® domains. However, this difference was not significant for physical function (P = 0⋅065), bodily pain (P = 0⋅149) or role emotional (P = 0⋅137). At 5 years, the opposite relationship was found, as sum-scores for improvers were significantly better (P < 0⋅050) than those for non-improvers for all domains. For improvers, sum-scores at 5 years were better than baseline scores for all eight SF-36 ® domains (P < 0⋅005). Non-improvers had better scores for physical function (P < 0⋅001) and worse scores in mental health (P = 0⋅017) at 5 years compared with the baseline, but no change in the other domains. Details on the relationship between changes in SRH and the eight domains in SF-36 ® are given in Tables 2 and 3, and Fig. 3.
In terms of co-morbidity, none of the 19 patients with decreased SRH at 5 years had type 2 diabetes mellitus (T2DM) before surgery. Of the 60 with no change in SRH, ten (17 per cent) had T2DM at baseline and six of these patients did not require medication at 5 years. Of the 154 patients with improved SRH, 28 (18⋅2 per cent) had T2DM at baseline and 21 did not require medication at 5 years ( Table 1).
Abdominal surgery for suspected internal herniation was more common among the improvers, but there was no difference between improvers and non-improvers for cholecystectomy, abdominal excess skin removal or births ( Table 1).

Discussion
Before RYGB, patients with severe obesity in the present cohort reported SRH far below that in the general population 29 , but after 5 years their scores were similar, with 81⋅5 per cent reporting SRH as good, very good or excellent.
QoL scores in SRH improvers were worse than those of non-improvers at baseline, but they were better at 5 years. Although improved SRH was related to better scores in all SF-36 ® domains, non-improvement was related to unchanged or worsened scores in all domains except physical function. None of the baseline characteristics predicted in which patients perceived health would improve. In a clinical context, these findings may indicate that patients with severe obesity who perceive their health as poor have more to gain from bariatric surgery than patients who perceive their health as good. Moreover, in the long run SRH can be interpreted as the result of the patients' continuous negotiation between the positive and negative effects of the RYGB procedure on all aspects of life.
It terms of the relationship of SRH with weight loss, the study found that the difference between SRH improvers and non-improvers depended partly on the formula used: %EWL and %EBMIL were better for improvers than for non-improvers, but %TWL and change in BMI were not different; and the proportion of patients attaining a BMI of 35 kg/m 2 or less, or %EWL above 50 per cent at 5 years, was similar for improvers and non-improvers. However, non-improvers regained 3⋅3 kg more than improvers from nadir to 5 years after RYGB, a significant difference. Whether this 'marginal' weight regain reduced SRH, or whether other health issues led to increased weight among non-improvers, could not be explored further from the available data.
Considering long-term outcomes, a meta-analysis 30 reported that health-related QoL improved in the first year after bariatric surgery, declined after 2 years and stabilized at a level below that in the general population at 5 years and, compared with control groups with obesity, improvement in both physical and mental health was reported more than 5 years after surgery 31 . Long-term observational studies 32 -34 of adults with severe obesity have reported that, compared with usual care, bariatric surgery is associated with a reduced rate of cardiovascular events and deaths, but still with a higher mortality rate than in the general population. The sample size in the present study was too small and the observation time too short to explore whether improved SRH after RYGB had an effect on mortality and future morbidity.
The strengths of this study are the close follow-up and complete registration for many variables from baseline to 5 years after the RYGB, and that patients reported on SRH when they had long-term experience of the positive and negative effects of the surgery on their general health status. Among the limitations of the study are the small sample size, and that the SF-36 ® questionnaire was not given to all patients who attended the 5-year follow-up visit. In addition, the study did not consider socioeconomic factors or life events that may have affected SRH at baseline or during follow-up after the bariatric procedure.
SRH, expressed by the answer to one single question, seems relevant and valid as an outcome measure for bariatric surgery, and in this observational study RYGB for severe obesity resulted in improved SRH in two-thirds of the patients. Focusing not only on weight, but also on health in general, might reduce the stigma experienced by people with severe obesity considering or undergoing bariatric surgery. The increased knowledge on what to expect from bariatric surgery will be useful for patient education, their choice of treatment, and their view of life after treatment for severe obesity. In clinical use, SRH might replace more comprehensive QoL tools, and SRH scores can be used to identify patients in need of closer follow-up after surgery.