Bariatric Surgery and Kidney Health

Medically complicated obesity is a clearly recognized public health concern in the United States and other Western countries. The underlying cause(s) are not entirely understood, nor are there any perfect solutions. Bariatric surgery is one approach to this vexing problem that has gained in popularity because it results in sustained weight loss and improvement in hard end points like diabetes, sleep apnea, and even death.¹ It is estimated that 50,000 malabsorptive procedures are performed yearly in North America, and 200,000 worldwide.² However, relatively little is known about long-term effects of bariatric surgery on kidney function.

In this issue of the Journal of the American Society of Nephrology, Friedman et al.³ analyzed data on a large cohort of patients followed for up to 7 years. They report that over this time span, improvement in the Kidney Disease Improving Global Outcomes CKD risk category was significantly more common than worsening. There are several caveats to keep in mind when analyzing these data. First, on average, patients lose not only fat but also muscle mass when they lose weight after bariatric surgery. In a recent study by Lieske et al.,⁴ creatinine generation fell an average 22.9% in 11 women followed for 1 year after roux en y gastric bypass (RYGB) (Table 1). In this Mayo cohort, measured GFR actually fell by 11%, whereas estimated GFR increased by 7% over the same period (because of a slight fall in serum creatinine).⁴ Other larger studies have also demonstrated a fall in 24-hour creatinine generation after bariatric surgery, ranging from 11% to 36% during the first postoperative year (Table 1).^5,6 This simple observation highlights the perils of using creatinine-based eGFR serially in any study involving bariatric surgical patients.

Table 1.

Patient demographics, blood, and urine creation before and after bariatric surgery

Clinical/Laboratory Feature
	Lieske et al.4			Getty et al.5		Serpa Neto et al.6
	Baseline	Postoperative		Baseline	Postoperative	Baseline	Postoperative
		6 mo	12 mo		6 mo		8 mo
Weight, kg	121.2±18.4	87.1±18.2a	75.7±9.2a	132.2±23	95.4±18.7a	122.76±18.98	83.77±13.53a
Body mass index, kg/m2	45.7±5.0	32.5±5.2a	28.4±2.0a	47.6±6.3	34.08±5.5a	46.17±5.44	31.47±3.71a
Serum creatinine, mg/dl	0.8±0.2	0.8±0.1	0.7±0.1	0.83±0.21	0.72±0.16a	0.83±0.20	0.69±0.13a
Urine creatinine,b mg/24 h	1342±44	1019±213	1035±255	1631.5	1445.3	1777.9	1130.7
Change in urine creatinine versus baseline, %	—	−25.1	−22.9	—	−11.4	—	−36.4
Creatinine clearance, ml/min	120±64	95±30	98±27	136.5±53	139.4±52	148.75±35.27	113.8±31.7a

Values are mean±SD. —, not available.

^aP value <0.001 versus baseline.

^b24-hour urine creatinine in Getty et al.⁵ and Serpa Neto et al.⁶ is calculated from the formula urine creatinine=plasma creatinine×creatinine clearance.

Estimating GFR in the presence of severe obesity poses a number of challenges. It is decidedly unclear how accurately creatinine-based eGFR equations perform among those with medically complicated obesity either before or after a significant weight loss, especially when looking at results normalized to a body surface area of 1.73 m². Use of cystatin C is one potential way around this because, unlike creatinine, it does not come strictly from muscle. However, cystatin C is not a well validated biomarker of GFR in the extremes of body mass. Compared with other cell types, adipose tissue appears to produce disproportionately more cystatin C,⁷ and cystatin C eGFR is biased toward increasingly lower values compared with measured GFR at higher body mass index levels.⁸ Further, the laboratory assay for cystatin C is not as standardized as creatinine, thus the cystatin C eGFR in the Friedman study may have been seriously confounded by systematic changes that have been reported in cystatin C results obtained from the Siemens assay between the early and late 2000s (around the time of the initial study visits).⁹ Overall, many factors in this study may have made the initial baseline cystatin C eGFR values falsely low compared with a measured GFR.

An issue not directly addressed in this study is the long-term risk of hyperoxaluria, kidney stones, and even oxalate nephropathy after bariatric surgery. Our population-based studies in Olmsted County suggest that that up to 50% of persons might be hyperoxaluric a year after RYGB, and the risk of new kidney stone events doubled compared with unoperated obese controls.¹⁰ Both risks are much higher in those that have undergone more weight loss procedures that cause a great amount of malabsorption (e.g., longer limb RYGB and biliopancreatic diversion with duodenal switch procedures), and new onset CKD was detectably higher in this subset as well.¹⁰ Indeed, cases of ESRD from oxalate nephropathy are being increasingly recognized.¹¹ Conversely, the risk of stones (or new onset CKD) does not appear to be increased in those who have had restrictive procedures.¹⁰

Nephrologists can contribute to the challenge of counseling patients about bariatric surgery, bringing a perspective on the balance of benefits and risks, starting with the surgical risks and the postoperative risk of AKI. By and large, those with medically complicated obesity will likely benefit from the weight loss that occurs after bariatric surgery. The net effect on long-term kidney health may be positive for many, on the basis the resolution of diabetes and the reduction in pathologic albuminuria as reported by Friedman et al.³ in this study and in earlier studies. Even if the eGFR is confounded by the above considerations, and true GFR is actually falling slightly, this may be positive in many cases (resolution of hyperfiltration). However, a relatively large number of patients that have malabsorptive bariatric procedures will likely suffer new stone events. A much smaller number may suffer oxalate nephropathy. Given the millions of people who have had these procedures over recent decades, it is important to recognize these complications because dietary manipulations can successfully reduce oxalate excretion and potentially reduce disease burden.

Clearly research is needed to develop additional treatments for this growing population of enteric hyperoxalurics who will require medical attention for at least the next generation, even if RYGB is supplanted by other weight loss strategies in the near future.

Disclosures

None.