Association of Bariatric Surgery With Rates of Kidney Function Decline Using Multiple Filtration Markers.

Introduction

Kidney function is typically assessed by estimated glomerular filtration rate using creatinine (eGFRcr), a marker heavily influenced by muscle mass.[1] Patients who undergo bariatric surgery have large changes in muscle mass, which may confound changes in eGFRcr over time. Other kidney filtration markers, such as cystatin C, β-2 microglobulin (β2m), and β-trace protein (βTP), are less affected by muscle mass but are rarely used clinically.[2] The objective of this study was to examine whether bariatric surgery was associated with a slower rate of kidney function decline, evaluating change in eGFR calculated using creatinine, cystatin C, β2m, βTP, and combination of creatinine and cystatin C, with the latter believed to be the most accurate GFR estimating equation in the general population and in individuals who have undergone bariatric surgery.[1,3]

Methods

This cohort study was approved by Geisinger Health System’s institutional review board. The informed consent requirement was waived per Geisinger policy, as participants had previously consented to a biobanking research study.[4] This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.

This observational, matched cohort study included adults aged 18 years or older at Geisinger Health System with body mass index (BMI; calculated as weight in kilograms divided by height in meters squared) 35 or higher with 2 biobanked serum samples, stored at −80 °C. Baseline evaluations were conducted from January 1, 2005, to December 31, 2009, and follow-up examinations were conducted from January 1, 2015, to December 31, 2017.

Patients who underwent bariatric surgery were matched 1 to 1 with participants who never underwent bariatric surgery based on sex, self-reported race, preoperative BMI (±10), age (±5 years), and eGFRcr (±10%). Each blood serum sample was measured for creatinine (enzymatic assay; Roche, isotope-dilution mass spectroscopy-traceable calibration; precision, 2.8%-2.9%), cystatin C (turbidometric assay; Gentian; precision, 3.2%-4.3%), β2m (latex agglutination; Roche; precision, 3.2%-5.1%), and βTP (immunonephelometric; Siemens; precision, 7.4%-10.6%) at the University of Minnesota. We used Chronic Kidney Disease Epidemiology equations to estimate GFR.[3,5]

The primary outcome was rate of decline in combined creatinine-cystatin C eGFR (eGFRcr-cyc). Secondary outcomes included rate of decline in eGFRcr, eGFRcyc, eGFRβ2m, and eGFRβTP. Mixed-effects models with random intercepts and slopes were used to compare eGFR trajectories between groups. We also examined whether the association differed by baseline eGFRcr-cyc, and conducted subgroup analyses by baseline eGFRcr-cyc levels. All analyses were completed using Stata/MP statistical software version 15.1 (StataCorp) from September 14, 2019, to June 7, 2020. We considered 2-sided P < .05 as statistically significant.

Results

Of 311 patients who underwent bariatric surgery and met inclusion criteria, a total of 144 patients (47.9 [10.2] years; 126 [87.5%] women) were matched with 144 participants (mean [SD] age, 48.5 [10.7] years; 126 [87.5%] women) who did not undergo bariatric surgery. Participants who underwent bariatric surgery vs those who did not were well-matched with no significant differences in mean (SD) eGFRcr-cyc (82.6 [19.9] mL/min/1.73 m2 vs 82.6 [18.2] mL/min/1.73 m2), hypertension (74 participants [51.4%] vs 71 participants [49.3%]), or diabetes (59 participants [41.0%] vs 59 participants [41.0%]) (Table). Baseline mean (SD) BMI was higher in the surgery group (46.2 [6.4] vs 44.1 [6.1]; P = .01). Mean eGFRcr-cyc decline rates were –0.41 (95% CI, –0.74 to –0.08) mL/min/1.73 m2 per year over a mean (SD) follow-up of 9.2 (1.4) years for the surgery group, and –1.44 (95% CI, –1.76 to –1.11) mL/min/1.73 m2 per year over a mean (SD) follow-up of 8.2 (1.1) years for the no surgery group (Table). Thus, bariatric surgery was associated with a 1.02 (95% CI, 0.56 to 1.49) mL/min/1.73 m2 per year slower decline in eGFRcr-cyc. Results were qualitatively similar using eGFRcr, eGFRcyc, eGFRβ2m, and eGFRβTP or when adjusted for baseline BMI (Table). Bariatric surgery was associated with a greater slowing of eGFR decline among those with lower baseline eGFR (P for continuous interaction = .04) (Figure).

Table.

Characteristics and Differences in Rate of eGFR Decline Between Individuals Who Underwent Bariatric Surgery and Those Who Did Not

Characteristics	Individuals, No. (%)		Difference (95% CI)	P value
	Surgery (n = 144)	No surgery (n = 144)
Age, mean (SD), y	47.9 (10.2)	48.5 (10.7)	NC	.62
Women	126 (87.5)	126 (87.5)	NC	>.99
Black race	1 (0.7)	1 (0.7)	NC	>.99
Weight, mean (SD), kg	126.2 (22.1)	118.6 (19.3)	NC	.002
BMI, mean (SD)	46.2 (6.4)	44.1 (6.1)	NC	.01
Hypertension	74 (51.4)	71 (49.3)	NC	.72
Diabetes	59 (41.0)	59 (41.0)	NC	>.99
eGFRcr-cyc, mean (SD), mL/min/1.73 m2	82.6 (19.9)	82.6 (18.2)	NC	>.99
Time from baseline to follow-up serum samples, median (IQR), y	9.17 (8.12 to 9.96)	8.20 (7.46 to 8.80)	NC	<.001
Weight change from baseline to follow-up, median (IQR), kg	–27.5 (–37.9 to –19.9)	–1.7 (–9.1 to 7.6)	NC	<.001
Rate of eGFR decline, mL/min/1.73 m2/y, mean (95% CI)a
eGFRcr-cyc	–0.41 (–0.74 to –0.08)	–1.44 (–1.76 to –1.11)	1.02 (0.56 to 1.49)	<.001
eGFRcr	–0.86 (–1.16 to –0.56)	–1.49 (–1.84 to –1.15)	0.63 (0.17 to 1.08)	.01
eGFRcyc	–0.06 (–0.41 to 0.29)	–1.40 (–1.73 to –1.07)	1.34 (0.85 to 1.82)	<.001
eGFRβ2m	–0.18 (–0.43 to 0.06)	–0.79 (–1.02 to –0.57)	0.60 (0.26 to 0.94)	.001
eGFRβTP	–0.18 (–0.41 to 0.05)	–0.97 (–1.19 to –0.76)	0.79 (0.47 to 1.10)	<.001

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); βTP, β-trace protein; β2m, β-2 microglobulin; cr, creatinine; cyc, cystatin C; eGFR, estimated glomerular filtration rate; IQR, interquartile range; NC, not calculated.

Calculated using Chronic Kidney Disease Epidemiology equations.

Figure.

Rates of Kidney Function Decline Stratified by Baseline Kidney Function

Discussion

To our knowledge, this cohort study is the first study to confirm a beneficial association of bariatric surgery with long-term kidney function trajectory using multiple filtration markers. The marked association between bariatric surgery and improved kidney function trajectory in patients at lower baseline kidney function was consistent with other studies that only examined eGFRcr.[6] Limitations of our study include a requirement to survive the interval period, potential selection bias, and other residual confounding. Since patients who undergo bariatric surgery must demonstrate sustained weight loss prior to the procedure, they may have had better adherence to healthy lifestyle behaviors and medication therapy. Strengths of this study include the use of multiple filtration factors to show consistency of the association and the long-term time interval. In conclusion, bariatric surgery was associated with long-term improvement in kidney function trajectory, assessed using multiple filtration markers.