Introduction
The worldwide prevalence of obesity has more than doubled since 1980, and approximately 1.9 billion people were obese in 2006. The obesity prevalence rates were 11.8%, 17.9%, and 22.1% in 1993–96, 2005–08, and 2013–14, respectively, according to the Nutrition and Health Survey in Taiwan. Obesity is associated with comorbidities, including hypertension, diabetes, cardiovascular disease, cancer, and chronic kidney disease (CKD). Compared with patients with body mass indexes (BMIs) <25 kg/m2, patients with severe obesity, that is, a BMI >35 kg/m2, have a 341% increased risk of developing end stage renal disease (ESRD). Potential factors underlying the increased risks for CKD and ESRD include obesity‐mediated hypertension, insulin resistance, obesity‐related glomerulonephropathy, renin–angiotensin–aldosterone system activation, inflammation, and dysregulated adipocytokine production.
Weight‐reduction therapy improves blood pressure, insulin resistance, dyslipidaemia, and obstructive sleep apnoea. A previous study's findings showed that compared with usual care, obese patients with type 2 diabetes benefitted more from bariatric surgery (BS), which included a lower incidence of nephropathy; however, the study was limited to patients with type 2 diabetes. Regarding CKD, a recent, large, long‐term cohort study's findings showed that the CKD risk categories improved in patients who underwent BS, but this study was limited by the absence of a control group. In addition, the obesity paradox suggests that obesity‐associated advantages in CKD patients could prevent inflammation and atherosclerosis. BS may induce kidney oxalate nephropathy and acute kidney injury. Furthermore, there are no guidelines regarding the use of BS to treat obese patients who are at risk of CKD and especially those at a high risk of CKD.
This study aimed to investigate the effects of BS or non‐surgical interventions on estimated glomerular filtration rates (eGFRs) and to determine whether BS can be recommended for renal function preservation based on BMI and eGFR changes in obese patients with CKD.
Materials and methods
The weight Reduction Intervention on GFR in Obese Patients with Renal Impairment‐Taipei Medical University (TMU) study was based on data from an institutional clinical database collected by Taipei Medical University Taipei Medical University Hospital (TMUH) and its affiliated hospitals, namely, Wan Fang Hospital and Shuang Ho Hospital, which have 3000 beds collectively. The database comprises three million patients' data that describe their demographic and clinical characteristics, outpatient and emergency room visits, hospital admissions, laboratory test results, and drug prescriptions; the data have been collected since 1997. All of the patients and physicians were anonymized, and the need for informed consent was waived. TMU's institutional review board exempted this study from a full review (TMU‐JIRB No. N201705035).
The Comprehensive Weight Management Center at TMUH provided guidance for efficient and healthy weight reduction. The BS teams at TMUH and its affiliated hospitals performed minimally invasive laparoscopic weight loss surgery and provided advice about lifestyle modifications, nutritional support, and professional physical training. Data from the subjects who registered to participate in the weight management programmes at TMUH and its affiliated hospitals were extracted from the clinical database at TMUH. The subjects' BMIs were recorded from electronic charts, and their eGFRs were calculated using the Taiwan Chronic Kidney Disease‐Epidemiology Collaboration (CKD‐EPI) equation.
Study subjects
We identified 4332 patients aged 20–70 years who visited the weight management centres at TMUH and Shuang Ho Hospital between 1 January 2008 and 31 August 2016. The exclusion criteria were patients who had never visited the weight management centres, were aged <20 or >70 years, and whose CKD risk was unknown (Figure ). The index dates were defined as the date of BS in the BS group and the cohort entry date, which was the first date on which the weight management centres were visited, in the non‐surgery group. The eGFR was calculated using the serum creatinine (sCr)‐based Taiwanese modification of the CKD‐EPI equation, that is, 1.262 × CKD‐EPI0.914; the CKD‐EPI equation follows: 141 × minimum (sCr/κ, 1)α × maximum (sCr/κ, 1)−1.209 × 0.993age × 1.018 (if female). The CKD risk groups were defined using the Kidney Disease: Improving Global Outcomes guidelines that consider the eGFR and the extent of proteinuria. The patients were assigned to three categories based on their CKD risks in the year before their index dates, as follows: low CKD risk: eGFR ≥90 mL/min·1.73 m2 and albuminuria <30 mg/g; moderate CKD risk: eGFR 60–89 mL/min·1.73 m2 or albuminuria ≥30 mg/g; and high CKD risk: eGFR <60 mL/min·1.73 m2. The interval between the collection of two samples for the sCr and urine protein measurements was >3 months; hence, a moderate CKD risk with proteinuria or a high CKD risk was considered to indicate pre‐existing CKD.
Interventions
Bariatric surgery was defined using the procedure codes in the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD‐9‐CM) and was based on (i) a laparoscopic adjustable gastric banding (ICD‐9‐CM: 44.99 and 54.21), (ii) a sleeve gastrectomy (ICD‐9‐CM: 43.89 and 54.21), (iii) a Roux‐en‐Y gastric bypass (ICD‐9‐CM: 43.7 and 54.21), or (iv) a bioenteric intragastric balloon placement (ICD‐9‐CM: 44.93).
Study endpoint
Based on a previous cohort study, a slower GFR decline, that is, 30% over 2 years, may be strongly associated with ESRD and all‐cause mortality. Our primary endpoint, which represented CKD progression, was defined as an eGFR decline ≥25% during the 1‐year follow‐up period. The follow‐up period began on the index date and continued until death, an eGFR decline ≥25%, 1 year after the index date, or until 31 August 2016, whichever came first.
Covariates
The medication prescriptions were extracted from the pharmaceutical data using the World Health Organization's Anatomical Therapeutic Chemical (ATC) classification system (
Statistical analyses
The risk estimates were computed based on the risk factors using univariate analyses and multivariate analyses with additional adjustments for potential confounders. Propensity scoring methods have been widely used to reduce confounding in observational studies. Each patient's probability of undergoing BS was derived from a logistic regression model that included baseline covariates, namely, age, sex, the CKD risk, hypertension, diabetes mellitus, coronary artery disease, nephrolithiasis, statins, diuretics, metformin, insulin, and the albumin, cholesterol, glutamic pyruvic transaminase, and triglyceride levels, the baseline BMI, smoking, and the cohort entry year. The propensity scores were matched using the nearest neighbour, 1:1 pair matching within 0.2 standard deviations (SDs) of the logit of the propensity score. By calculating the standardized differences of the means or proportions of each covariate after matching, baseline covariates with standardized differences <0.1 indicated a good balance between the BS group and the non‐surgery group.
Within the matched sample, the 95% confidence intervals (CIs) for incidence rates from the Poisson regression models, and the hazard ratios (HRs), and 95% CIs from the Cox proportional hazard regression models were used to estimate the risks of eGFR declines ≥25%. The times to eGFR declines ≥25% were determined using the Kaplan–Meier method and compared using the log‐rank test. We modelled the eGFR trajectories during the follow‐up period by using restricted cubic splines with five knots at the 5th, 27.5th, 50th, 72.5th, and 95th percentiles. Stratified analyses of the relative risks of the subjects' outcomes according to albuminuria, hypertension, and diabetes mellitus were performed. Spearman's correlation coefficient was used to determine associations between BMI declines and eGFR changes from baseline until 1 year, and scatter plots were drawn. All of the data management and analyses were performed using
Results
Demographic characteristics of the propensity score‐matched obese study population
The mean (SD) ages of the BS and non‐surgery groups (both n = 810) were 36.6 (10) years and 36.3 (9.8) years, respectively, and 36.0% and 35.3% of the patients in the BS and non‐surgery groups, respectively, had moderate risks of CKD with albuminuria ≥30 mg/day. The average BMIs in the BS and non‐surgery groups were 37.8 and 37.7 kg/m2, respectively. There were no differences between the groups (Table ). Supporting Information, Table S1 presents the raw data from the whole population (n = 4332).
Demographic data from patients in the weight management centres after propensity score matching| Overall (n = 1620) | Bariatric surgery (n = 810) | No bariatric surgery (n = 810) | ASD | ||||
| n | (%) | n | (%) | n | (%) | ||
| Age, years, mean (SD) | 36.5 | (9.9) | 36.6 | (10.0) | 36.3 | (9.8) | 0.04 |
| 20–29 | 407 | (25.1) | 203 | (25.1) | 204 | (25.2) | |
| 30–39 | 692 | (42.7) | 339 | (41.9) | 353 | (43.6) | |
| 40–49 | 336 | (20.7) | 171 | (21.1) | 165 | (20.4) | |
| 50–59 | 148 | (9.1) | 75 | (9.3) | 73 | (9.0) | |
| 60–70 | 37 | (2.3) | 22 | (2.7) | 15 | (1.9) | |
| Female, n (%) | 975 | (60.2) | 495 | (61.1) | 480 | (59.3) | 0.04 |
| CKD risk | 0.01 | ||||||
| Low | 985 | (60.8) | 488 | (60.2) | 497 | (61.4) | |
| Moderate | 578 | (35.7) | 292 | (36.0) | 286 | (35.3) | |
| High | 57 | (3.5) | 30 | (3.7) | 27 | (3.3) | |
| Comorbidities | |||||||
| Hypertension | 501 | (30.9) | 259 | (32.0) | 242 | (29.9) | 0.05 |
| Diabetes mellitus | 307 | (19.0) | 157 | (19.4) | 150 | (18.5) | 0.02 |
| Coronary artery disease | 58 | (3.6) | 35 | (4.3) | 23 | (2.8) | 0.08 |
| Nephrolithiasis | 23 | (1.4) | 12 | (1.5) | 11 | (1.4) | 0.01 |
| Medications | |||||||
| Statins | 91 | (5.6) | 44 | (5.4) | 47 | (5.8) | 0.02 |
| Diuretics | 70 | (4.3) | 37 | (4.6) | 33 | (4.1) | 0.02 |
| Metformin | 133 | (8.2) | 69 | (8.5) | 64 | (7.9) | 0.02 |
| Insulin | 84 | (5.2) | 42 | (5.2) | 42 | (5.2) | 0.00 |
| Laboratory | |||||||
| Albumin, g/dL, mean (SD) | 4.5 | (0.3) | 4.5 | (0.3) | 4.5 | (0.3) | 0.02 |
| Cholesterol, mg/dL, mean (SD) | 199.0 | (39.4) | 201.0 | (36.5) | 196.9 | (42.1) | 0.00 |
| GPT, IU/L, mean (SD) | 51.4 | (53.4) | 52.2 | (55.5) | 50.6 | (51.2) | 0.05 |
| TG, mg/dL, mean (SD) | 174.3 | (198.1) | 173.6 | (132.1) | 175.0 | (246.7) | 0.02 |
| Smoking | 18 | (1.1) | 10 | (1.2) | 8 | (1.0) | 0.02 |
| Baseline BMI, kg/m2, mean (SD) | 37.7 | (6.9) | 37.8 | (5.7) | 37.7 | (7.9) | 0.02 |
| Surgery type | |||||||
| Laparoscopic adjustable gastric banding | 75 | (4.6) | 75 | (9.3) | — | ||
| Sleeve gastrectomy | 722 | (44.6) | 722 | (89.1) | — | ||
| Roux‐en‐Y gastric bypass | 6 | (0.4) | 6 | (0.7) | — | ||
| Bioenteric intragastric balloon | 7 | (0.4) | 7 | (0.9) | — | ||
| Follow‐up frequency | |||||||
| Mean (SD) | 5.9 | (5.0) | 8.6 | (5.0) | 3.2 | (3.1) | |
| Median (Q1–Q3) | 5 | (2–8) | 8 | (5–11) | 2 | (1–4) |
Times to an estimated glomerular filtration rate decline of ≥25%
Overall, the eGFR was significantly less likely to decline ≥25% in the BS group than in the non‐surgery group (log rank P = 0.02) (Figure A). The likelihood of the eGFR declining ≥25% was lower in the BS group compared with that in the non‐surgery group among the patients with a low CKD risk, but the difference was not significant (Figure B). Among the patients with a moderate CKD risk, the eGFR in the BS group was significantly less likely to decline ≥25% compared with that in the non‐surgery group (log rank P < 0.001) (Figure C). For the patients with a high CKD risk, the eGFR in the BS group was significantly less likely to decline ≥25% compared with that in the non‐surgery group (log rank P = 0.039) (Figure D). Significantly more patients in the high CKD risk group than those in the low and moderate CKD risk groups showed eGFR decreases ≥25% (P < 0.001) (Figure E). After stratifying the patients into six groups according to the CKD risk and BS, the high CKD risk group that did not undergo BS had a significantly higher likelihood of an eGFR decrease ≥25% compared with the other groups (P < 0.001) (Figure F).
Estimated glomerular filtration rate trajectories in the patients who underwent bariatric surgery and matched control patients
The BS group's continuous eGFR curve showed an upward trend and that of the non‐surgery group showed a downward trend (Figure A). The upward trajectory was more pronounced in the groups of patients with eGFRs <90 mL/min·1.73 m2 who underwent BS (Figure B). In the BS group, the eGFR improvement began immediately and plateaued at an elevation of approximately 10% within about 3 months. A 10% decrease in the eGFR that became more stable after 3 months was observed in the non‐surgery group. Supporting Information, Figure S1 shows the stratification of the groups according to an eGFR of 60 mL/min·1.73 m2.
Changes in the estimated glomerular filtration rate and body mass index trajectories from baseline
Figure shows the relationships between time and the BMI and eGFR. The eGFRs did not improve in the patients with moderate or high risks of CKD who did not undergo BS (Figure A). In the patients with moderate or high risks of CKD who underwent BS, the eGFRs had improved at 3 months, the BMI reductions peaked at 6 months, and the eGFR improvements continued for up to 12 months (Figure B). In the patients at a low risk of CKD, the eGFR did not improve in those who did not undergo BS (Figure C), and it remained the same during the 12‐month study period in those who underwent BS (Figure D). Hence, patients with moderate and high CKD risks experienced the greatest eGFR improvements after BS.
Relationships between body mass index declines and estimated glomerular filtration rate changes according to the chronic kidney disease risk category and surgery
The BS group showed a greater BMI decline (2.5 kg/m2) at 12 months than the non‐surgery group (1.3 kg/m2), and similar BMI declines across the CKD risk categories (Table ). Overall, the eGFR increased by a mean (SD) of 4.4 (14) mL/min·1.73 m2 in the BS group and it decreased by a mean (SD) of 6.4 (16.0) mL/min·1.73 m2 in the non‐surgery group at 12 months. The greatest mean (SD) eGFR increase of approximately 10.2 (18.6) mL/min·1.73 m2 occurred among the patients with moderate or high CKD risks in the BS group (Table ). A BMI decline was significantly correlated with an eGFR change in the patients with moderate or high risks of CKD (Spearman's correlation −0.229, P < 0.001) (Figure A), but a correlation was not evident among the patients with a low risk of CKD (Spearman's correlation 0.02, P = 0.67) (Figure B).
Mean body mass index decline over the 1‐year follow‐up period stratified according to the baseline chronic kidney disease risk category| Visit (n) | Overall | Low risk | Moderate and high risk |
| Mean (SD) | Mean (SD) | Mean (SD) | |
| 3 months (494) | 1.1 (2.5) | 1.1 (2.4) | 1.2 (2.6) |
| 12 months (186) | 2.1 (3.7) | 2.1 (3.7) | 2.1 (3.7) |
| Bariatric surgery | |||
| 3 months (379) | 1.1 (2.5) | 1.0 (2.5) | 1.2 (2.6) |
| 12 months (170) | 2.5 (4.0) | 2.4 (3.9) | 2.7 (4.1) |
| No bariatric surgery | |||
| 3 months (115) | 1.2 (2.2) | 1.3 (2.0) | 1.1 (2.7) |
| 12 months (16) | 1.3 (2.9) | 1.5 (3.1) | 1.0 (2.6) |
| Visit (n) | Overall | Low risk | Moderate or high risk |
| Mean (SD) | Mean (SD) | Mean (SD) | |
| 3 months (494) | 0.6 (12.2) | −1.5 (6.5) | 3.9 (17.3) |
| 12 months (186) | 2.2 (15.2) | −0.4 (7.2) | 5.0 (20.5) |
| Bariatric surgery | |||
| 3 months (379) | 1.9 (10.2) | −0.8 (5.5) | 6.4 (14.0) |
| 12 months (170) | 4.4 (14.0) | 0.2 (7.2) | 10.3 (18.6) |
| No bariatric surgery | |||
| 3 months (115) | −3.9 (16.5) | −4.1 (8.8) | −3.5 (23.6) |
| 12 months (16) | −6.4 (16.6) | −4.7 (5.3) | −7.2 (19.8) |
Subgroup analysis
In the BS group, there was a 53% lower risk of an eGFR decrease ≥25% (HR 0.47, 95% CI 0.24–0.91, P = 0.03) compared with that in the non‐surgery group (Table ) after adjusting for the confounding factors, including the baseline eGFR and propensity scores. The adjusted multivariate model determined that a subgroup comprising obese patients with albuminuria ≥30 mg/g who had undergone BS had an 87% lower risk of an eGFR decrease ≥25% (HR 0.13, 95% CI 0.04–0.47, P = 0.002) compared with that in the non‐surgery group. The subgroup comprising obese patients with hypertension who had undergone BS had a 63% lower risk of an eGFR decrease ≥25% (HR 0.37, 95% CI 0.16–0.86, P = 0.02) compared with that in the non‐surgery group. The effect of BS was of borderline significance in the group of patients with type 2 diabetes mellitus.
Incidence rates of and hazard ratios for an estimated glomerular filtration rate decline ≥25%| n | Person month | Events, No. | Rate per 1000 person‐months (95% CI) | Model 1 | Model 2 | |||
| HR (95% CI) | P | HR (95% CI) | P | |||||
| Total | 1620 | 18 031 | 40 | 2.22 (1.58–3.02) | ||||
| No bariatric surgery | 810 | 8160 | 26 | 3.19 (2.08–4.67) | (Reference) | (Reference) | ||
| Bariatric surgery | 810 | 9871 | 14 | 1.42 (0.78–2.38) | 0.47 (0.25–0.90) | 0.02 | 0.47 (0.24–0.91) | 0.03 |
| Albuminuria | ||||||||
| <30 mg/g | 985 | 10 843 | 17 | 1.57 (0.91–2.51) | 1.49 (0.54–4.10) | 0.44 | 1.50 (0.54–4.15) | 0.43 |
| ≥30 mg/g | 635 | 7188 | 23 | 3.20 (2.03–4.80) | 0.14 (0.04–0.48) | 0.002 | 0.13 (0.04–0.47) | 0.002 |
| HTN | ||||||||
| No | 1119 | 12 644 | 13 | 1.03 (0.55–1.76) | 0.77 (0.25–2.36) | 0.65 | 0.77 (0.25–2.35) | 0.65 |
| Yes | 501 | 5387 | 27 | 5.01 (3.30–7.29) | 0.36 (0.16–0.85) | 0.02 | 0.37 (0.16–0.86) | 0.02 |
| DM | ||||||||
| No | 1313 | 14 715 | 22 | 1.50 (0.94–2.26) | 0.50 (0.21–1.21) | 0.16 | 0.50 (0.21–1.22) | 0.13 |
| Yes | 307 | 3316 | 18 | 5.43 (3.22–8.58) | 0.41 (0.15–1.08) | 0.07 | 0.41 (0.15–1.07) | 0.07 |
Discussion
This study's findings demonstrated the beneficial effect of BS on the eGFRs in obese patients at risk of CKD; hence, BS improved the patients' renal function, and the effect lasted for >1 year. BS, which mainly comprises laparoscopic sleeve gastrectomy, is a novel and effective obesity treatment for patients with CKD. BS has become standard therapy for patients with type 2 diabetes who are severely obese with BMIs >40 kg/m2 or >35 kg/m2 if their glucose levels are poorly controlled, because it reduces insulin resistance, proteinuria, and cardiovascular risks. Our study has generated evidence that will help determine whether all severely obese patients with renal disease should undergo BS and whether BS is appropriate for all CKD stages. This study's strength relates to our use of a large propensity score‐matched long‐term cohort from an electronic clinical database at TMUH. However, BS did not improve the eGFR in the group of patients with a low CKD risk. After BS, kidney hypertrophy/hyperfiltration may reverse, thereby normalizing the GFR in obese patients with a low risk of CKD; thus, the eGFR may also decrease. Our results that showed eGFR declines in the group with a low risk of CKD concur with these findings. Similar results have been reported from other studies. At 3 months after BS, the eGFR increased in the moderate and high CKD risk groups by up to 10 mL/min·1.73 m2. The findings from a study of 25 patients with CKD who underwent BS showed similar eGFR elevations of up to 20 mL/min·1.73 m2, and a larger cohort study's findings showed that BS had long‐term benefits for the eGFR that persisted for 5 years in obese patients with CKD. Although a large cohort study's findings showed that at 30 days after BS, each incremental CKD stage correlated with a 1.3‐times increase in complications, there were no long‐term follow‐up assessments. Therefore, despite increases in short‐term post‐operative complications, the long‐term benefit of BS‐induced weight reductions on CKD cannot be underestimated.
Overall, the mean (SD) decline in the BMI over 1 year in the study subjects who underwent BS was only 2.1 (3.7) kg/m2 or 6%. Two reasons may explain this small BMI decline. First, the follow‐up duration was only 1 year, and meta‐analyses have shown long‐term weight reductions of 5.2 kg/m2 or 32% after BS during follow‐up periods of up to 5 years. Second, the patients were Asian people whose mean (SD) BMI was 37.7 (6.9) kg/m2, and they were not very fat; thus, the effect of BS on weight reduction may not be pronounced. Interestingly, the BMI decline was proportional to the eGFR changes in the moderate and high CKD risk groups, but this was not observed in the low CKD risk group. After BS, the eGFR may have improved in the moderate and high CKD risk groups following a rapid decline in muscle mass during the weight‐reduction intervention that did not occur in the patients who did not have CKD. Patients with CKD are prone to muscle wasting caused by uraemia‐induced alterations, including increased energy expenditure and persistent inflammation, which lead to the excessive catabolism of muscle and fat. BS increases gastric emptying and intestinal transit as consequences of the activation of glucagon‐like peptide‐1, and ghrelin inhibition without weight changes may occur after BS. The findings from an animal study suggest that glucagon‐like peptide‐1 might increase the GFR by 30–50% by inhibiting tubular reabsorption, increasing renal blood flow/natriuresis, and inhibiting inflammation and reactive oxygen species production, which are induced by increased levels of markers of inflammation, including interleukin‐6 or tumour necrosis factor‐α, in adipose tissue. In addition, the findings from a study of obese patients with CKD who underwent BS showed that leptin and β‐2 microglobulin, rather than the BMI and the fat mass, correlated with kidney function. Similarly, the findings from a recent meta‐analysis showed that the GFR increased in a subgroup of patients with CKD and decreased in a hyperfiltration subgroup. Furthermore, compared with other forms of calorie restriction that induce transient visceral fat decreases, a larger loss of fat‐free mass accompanies BS. Although a high‐protein diet comprising 1–1.2 g/kg ideal body weight is recommended to patients who undergo BS, but not to CKD patients, many patients continue to consume low‐protein diets. While a low‐protein diet did not correlate with GFR changes in normal obese patients after BS, a low‐protein diet preserved renal function, particularly in patients with CKD. Hence, our study's findings showed that the eGFR improvements correlated with the BMI declines in the patients with CKD, but this was not observed in the patients who did not have CKD; this may have been caused by different rates of muscle wastage between the patients with and without CKD, improvements in the microenvironment, and the inhibition of inflammation.
Whether BS has more favourable effects on renal function is a matter of debate, because an accurate equation is not available that can determine the eGFR and the sCr and serum cystatin C levels are affected by extrarenal factors that are modulated by BS. Compared with measured GFRs, eGFRs determined using the original Modification of Diet in Renal Disease (MDRD) or CKD‐EPI equations are not accurately predicted for obese patients undergoing BS. A meta‐analysis of 23 cohort studies included measured GFRs in only one study, and this meta‐analysis showed that GFRs determined using the MDRD, CKD‐EPI, and Cockcroft–Gault equations were homogenous. Hence, it might be acceptable to use eGFRs to predict renal function in obese patients who are undergoing weight‐reduction therapy.
The subgroup analysis showed that BS might act independently to slow eGFR declines and, similarly, that patients with proteinuria or hypertension might benefit significantly more from weight‐reduction programmes than patients who do not have proteinuria or hypertension. Although the effect of BS on patients with type 2 diabetes was of borderline significance in relation to preserving the eGFR, the trend suggested a benefit. Recently, a Cochrane meta‐analysis indicated that even dietary interventions in patients with CKD increased the eGFR and lowered blood pressure. Therefore, clinicians may suggest that obese patients with CKD risks comprising either proteinuria or hypertension should enter weight‐reduction programmes to preserve the eGFR.
Limitations
Our study's limitations are described next. First, the follow‐up period was only 1 year, and the study involved Asian subjects whose levels of obesity are lower relative to those of Caucasian individuals. According to a study conducted by Navarro‐Diaz et al., eGFRs do not change significantly or remain stable in severely obese patients for 12–24 months after BS; therefore, a cohort study with a longer follow‐up period that includes other ethnic groups is warranted to corroborate the effects of BS on eGFRs and to generate generalizable data. Second, the eGFR may not accurately reflect measured renal function, especially with regard to predicting exogenous creatinine clearance. Indeed, an equation does not exist that can accurately estimate the GFR following significant body weight changes, and, in this situation, the MDRD equation is comparable with the CKD‐EPI equation for GFR predictions. Third, we did not record the lean body mass of the obese patients to prove that the effect of BS on the eGFR resulted from muscle wasting, which is commonly seen in advanced CKD patients, especially those who have been on long‐term dialysis.
In conclusion, the findings from this 1‐year large, propensity score‐matched cohort study showed that BS is beneficial for preserving the eGFR in obese patients and that it is associated with BMI declines, particularly among those with moderate and high risks of CKD. Clinicians may arrange weight‐reduction programmes for obese patients with CKD risks, including proteinuria or hypertension, to preserve their eGFRs. Further studies with longer follow‐up durations that evaluate changes in endocrine biomarkers or inflammatory markers and muscle wasting after BS are warranted to elucidate the relationship between muscle mass decreases and microenvironmental changes in the GFR. In addition, more data are needed that describe long‐term outcomes, for example, ESRD, hospitalization, or death, to gain a better understanding of the beneficial effects of weight‐reduction surgery and subsequent renoprotection in patients with CKD.
Acknowledgement
We thank the staff of Office of Information Technology, Taipei Medical University, for the technical support. The authors certify that they comply with the ethical guidelines for authorship and publishing of the Journal of Cachexia, Sarcopenia and Muscle.
Authors' contributions
Y.‐C.L. acted as the chief executor taking the whole responsibility of the study design, progress monitoring, data interpretation, and part of manuscript writing. Y.‐Z.L. performed the experiments and part of manuscript writing. M.‐S.W. discussed, analysed, interpreted the data, and was responsible for the discussion and trouble shooting. Y.‐C.L., M.‐S.W., K.‐C.C., and C.‐C.P. contributed reagents/materials. M.‐T.C. and T.‐H.C. contributed assistance in drafting the article and reformatting all plots in this manuscript.
Funding
This research is partially supported by Ministry of Science and Technology, Taiwan, R.O.C. under grant no. MOST 107–2221‐E‐038‐014‐MY2 and by Taipei Medical University (grant no. 106TMU‐AE1‐B19).
Conflicts of interest
The authors declare no conflicts of interest.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
© 2019. This work is published under http://creativecommons.org/licenses/by-nc-nd/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Abstract
Background
Weight‐reduction therapies, including bariatric surgery (BS), are standard treatments for severely obese patients with type 2 diabetes; however, the outcomes of these therapies are inconclusive for obese patients with chronic kidney disease (CKD). This study aimed to investigate the effects of BS or non‐surgical interventions on the estimated glomerular filtration rate (eGFR) and to determine whether BS can be recommended for renal function preservation based on body mass index (BMI) and eGFR changes in obese patients with CKD.
Methods
This study used data from the weight Reduction Intervention on GFR in Obese Patients with Renal Impairment‐Taipei Medical University (TMU) study, which was a large, long‐term, propensity score‐matched cohort study based on clinical data from patients who registered at weight‐reduction centres at TMU and its affiliated hospitals from 2008 to 2016. The patients were stratified according to whether they had undergone BS and into the mild, moderate, and high CKD risk groups using the Kidney Disease: Improving Global Outcomes guidelines. The primary outcome was the eGFR calculated using the Taiwan Chronic Kidney Disease‐Epidemiology Collaboration equation. Cox regression models were used to determine hazard ratios (HRs) for eGFR decreases ≥25%.
Results
A total of 4332 obese patients were enrolled in this study. After propensity score matching, 1620 patients, including 60.2% women, with a mean age of 36.5 (9.9) years were divided into BS or non‐surgery groups (n = 810 per group). The overall mean eGFRs increased by 4.4 (14) mL/min·1.73 m2 and decreased by 6.4 (16.0) mL/min·1.73 m2 in the BS and non‐surgery groups, respectively. The decrease in BMI in the BS and non‐surgery groups were 2.5 and 1.3 kg/m2, respectively. In the moderate/high CKD risk BS group, a significant correlation was evident between an increased eGFR and a reduced BMI (Spearman's correlation −0.229, P < 0.001). The Cox regression analysis showed that the BS group had a significantly lower risk of an eGFR decline ≥25% at 12 months [adjusted HR (aHR) 0.47, P = 0.03). After BS, obese patients with hypertension or albuminuria had significantly lower risks of eGFR declines ≥25% (aHR 0.37, P = 0.02 and aHR 0.13, P = 0.0018, respectively).
Conclusions
Bariatric surgery was associated with eGFR preservation in all obese patients and, particularly, in those with moderate‐to‐high CKD risks. A longer term outcome study is warranted to determine the benefits of BS for CKD patients.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
Details
1 Graduate Institute of Clinical Medicine, School of Medicine, College of Medine, Taipei Medical University, Taipei, Taiwan; Division of Nephrology, Department of Internal Medicine, Taipei Medical University Hospital, Taipei, Taiwan; Department of Internal Medicine, School of Medicine, College of Medicine, Taipei Medical University, Taipei, Taiwan
2 Department of Internal Medicine, School of Medicine, College of Medicine, Taipei Medical University, Taipei, Taiwan; Division of Cardiology, Department of Internal Medicine, Taipei Medical University Hospital, Taipei, Taiwan
3 Faculty of Medicine, National Yang‐Ming University, Taipei, Taiwan; Division of Endocrinology and Metabolism, Department of Medicine, Taipei Veterans General Hospital, Taipei, Taiwan
4 Graduate Institute of Clinical Medicine, School of Medicine, College of Medine, Taipei Medical University, Taipei, Taiwan
5 Graduate Institute of Clinical Medicine, School of Medicine, College of Medine, Taipei Medical University, Taipei, Taiwan; Division of Urology, Taipei Medical University‐Shuang Ho Hospital, New Taipei City, Taiwan
6 Graduate Institute of Biomedical Informatics, College of Medical Science and Technology, Taipei Medical University, Taipei, Taiwan; Office of Information Technology, Taipei Medical University, Taipei, Taiwan
7 Graduate Institute of Clinical Medicine, School of Medicine, College of Medine, Taipei Medical University, Taipei, Taiwan; Division of Nephrology, Department of Internal Medicine, Taipei Medical University‐Shuang Ho Hospital, New Taipei City, Taiwan
8 Graduate Institute of Biomedical Informatics, College of Medical Science and Technology, Taipei Medical University, Taipei, Taiwan; Clinical Big Data Research Center, Taipei Medical University Hospital, Taipei, Taiwan