Introduction

Although living kidney donation is characterized as a safe procedure with low risks for the donor, potential living kidney donors are thoroughly screened to ensure that the risk of adverse outcome of donation is as minimal as possible. Given the health risks that accompany obesity, most centers exclude individuals with a high body mass index (BMI), indicating obesity, from donation. However, with the rising obesity epidemic, an increasing number of individuals presenting themselves as potential living kidney donors are overweight or obese. Consequently, the criteria for living kidney donation are gradually becoming more liberalized to address the ongoing demand for donor kidneys. Screening guidelines for living kidney donation mostly rely on BMI1,2. However, BMI is not reliable in accurately assessing body composition and identifying a central body fat distribution, which has been associated with an increased risk of kidney function impairment3,4,5. This may result in falsely low estimation of risk for future decline of renal function in donors with a pathological body composition. Waist circumference and waist-to-hip ratio provide more information about body shape and fat distribution6,7, but cannot differentiate between different fat tissue compartments. This distinction can have important health implications, since especially visceral and intramuscular adipose tissue have been associated with kidney risk and metabolic risk factors such as high blood pressure and insulin resistance8,9,10,11. Computed tomography (CT) imaging can accurately visualize and quantify fat distribution, which can aid in identifying donors at risk of developing kidney failure after donation. To be able to provide optimal donor care, there is a need for a novel approach to identify which donors have a body composition type that puts them at increased risk of developing disorders such as kidney failure after donation.

This study aimed to determine the association between abdominal fat area and kidney function before and after donor nephrectomy using CT imaging. We hypothesized that higher levels of abdominal fat are associated with lower kidney function.

Methods

This retrospective cohort study included 550 living kidney donors. All donors underwent donor nephrectomy between 2002 and 2019 at the University Medical Center Groningen (UMCG), Groningen, the Netherlands. Potential donors were excluded based on the Dutch guidelines for screening for living kidney donation: BMI > 35 kg/m2, inability to provide informed consent, manifested Diabetes Mellitus, major cardiovascular risk factors, prior kidney disease or glomerular filtration rate (GFR) of < 60 mL/min/1.73 m2, monokidney, pregnancy, recent or active malignancies, chronic/active infection, hypertension with end-organ damage, inadequately regulated hypertension, proteinuria, microscopic hematuria, and bilateral nephrolithiasis on CT scan. The study-specific exclusion criteria were significant interfering artifacts on CT imaging and when subcutaneous fat tissue was not fully visible on CT imaging. An overview of the participants and available data is provided in Fig. 1.

Fig. 1
figure 1

Overview of the participants and available data. Measured Glomerular Filtration Rate, mGFR (mL/min)

Full size image

All clinical and radiological data were retrieved from the TransplantLines Biobank and Cohort Study (ClinicalTrials.gov identifier: NCT03272841). All solid organ transplant donors (aged ≥ 18 years) from the UMCG were invited to participate and all participants provided written informed consent on enrolment. A detailed description of the study design, inclusion and exclusion criteria has been described previously12. The study protocol has been approved by the local Institutional Ethical Review Board (‘Medisch Ethische Toetsingscommissie UMC Groningen’, METc 2014/077) and all experiments were performed in accordance with relevant guidelines and regulations12. In 2020, historical data of living kidney donors were included in the TransplantLines biobank and cohort study and underwent a renewed ethical review in accordance with the current ethical guidelines. With its approval, the use of historical clinical and biological materials of living kidney donors in research and publications was approved alongside newly collected data of donors. The application for access to the necessary retrospective data for this study was approved by the TransplantLines working group, as the analyses in the present study fall under the scope of the METc. The clinical and research activities being reported are consistent with the Principles of the Declaration of Istanbul as outlined in the ‘Declaration of Istanbul on Organ Trafficking and Transplant Tourism’. All clinical and biochemical measurements were performed as described previously12. Measurements were performed during donor screening and 3 months after donor nephrectomy. Kidney function was assessed using measured GFR (mGFR, 125I-Iothalamate) as described in a previous study13. Body surface area (BSA, m2) was calculated using the DuBois formula14. All living kidney donors underwent CT imaging at the UMCG (n = 545) or Dutch non-academic referring hospitals (n = 5) as part of donor screening. A total of 546 scans were contrast-enhanced (n = 4, portal venous phase; n = 13, arterial phase; n = 529, late phase) and 4 were unenhanced. Scan matrix was 512 × 512, slice thickness varied between 0.75 and 5.0 mm. Median tube voltage and mean current were 100 kVp and 90.5 mAs, respectively. A soft tissue reconstruction kernel was used for all scans.

The cross-sectional plane of a CT slice at vertebral level L3 was analyzed using the semi-automatic program Sarcomeas (version 0.54, UMCG, Groningen). The slices were imported anonymously from the picture archiving and communication system (PACS) in the native Digital Imaging and Communications in Medicine (DICOM) format. The fat area was manually delineated by an experienced radiologist and rechecked by a second radiologist. Fat was defined as densities ranging from − 190 to -30 Hounsfield units (HU). Measurements of each abdominal fat compartment (subcutaneous adipose tissue (SAT), visceral adipose tissue (VAT), and intramuscular adipose tissue (IMAT)) were performed, as well as for the total abdominal adipose tissue area (TAT = VAT + SAT + IMAT). These measurements were indexed for the height2 (m2) of the donor. The height-indexed tomographic fat measurements included the visceral abdominal adipose tissue index (VATi), subcutaneous abdominal adipose tissue index (SATi), intramuscular abdominal adipose tissue index (IMATi), and total abdominal adipose tissue index (TATi).

Normally distributed variables are presented as means (standard deviations) and were tested for normality using histograms and quantile-quantile plots. Linear regression analyses were performed to assess the association between tomographic fat measurements and kidney function before and after donation, adjusting for age15, sex16, and body size17. Primary analyses were performed with BSA-standardized mGFR (mL/min per 1.73 m2), a common practice in literature. Secondary analyses were performed using crude mGFR (mL/min). Age at CT had a quadratic relationship with mGFR, and a quadratic and centered term was added. Multiplicative interaction terms were tested, but no significant interactions were found. The multivariable regression model with kidney function consisted of tomographic fat measurement indexed for height (cm2/m2), age at CT scan (absolute term, years), age at CT scan2 (quadratic term), mGFR at screening (mL/min, in models with outcome mGFR at 3 months post-donation), and weight at screening (kg, in models with outcome variable crude mGFR). Homoscedasticity, collinearity, and normal distribution and autocorrelation of the residuals were assessed. Two-tailed p-values were used, with significance set at p < 0.05. Statistical analyses were performed using RStudio (PBC, Boston, MA, USA, 2021) and SPSS version 28.0 (IBM, Armonk, USA). In additional analyses, linear regression analyses were performed with available mGFR data at 5 years post-donation, and the associations between pre-donation BMI and mGFR prior to and 3 months after donation were assessed.

Results

Characteristics of the study population

A total of 550 living kidney donors were included in this study, of whom 56% were male. The mean age at the time of CT was 53 ± 11 years. Male donors had significantly higher BMI (male: 25.6 ± 3.01 versus female: 24.7 ± 3.12 kg/m2, p = 0.001), TAT (male: 306.2 ± 129.6 versus female: 282.7 ± 117.8 cm2, p = 0.03), VAT (male: 153.0 ± 82.1 versus female: 83.1 ± 51.5 cm2, p < 0.001), and IMAT (male: 14.6 ± 7.60 versus female: 11.9 ± 6.49 cm2, p < 0.001) values than female donors. Male donors also had significantly higher baseline mGFR values than female donors (male: 120.9 ± 20.3 versus female: 102.1 ± 17.4 mL/min, p < 0.001) (Table 1).

Table 1 Baseline characteristics of the study population at screening for donation.
Full size table

Tomographic fat measurements and mGFR at screening

In the unadjusted linear regression analyses, VATi, SATi, IMATi, and TATi were significantly associated with BSA-standardized mGFR at screening (Table 2). After adjusting for sex and age, higher levels of all tomographic fat measurements were significantly associated with lower BSA-standardized mGFR at screening (VATi: Β = -0.11, p < 0.001, SATi: Β = -0.10, p < 0.001, IMATi: Β = -1.18, p < 0.001, TATi: Β = -0.07, p < 0.001) (Table 2). In age- and weight-adjusted secondary analyses with crude mGFR, all tomographic fat measurements were significantly associated with mGFR at screening, with higher fat areas being associated with lower mGFR (Supplemental Table S1). After dividing the study population into quartiles based on their fat index levels, one-way ANOVA analyses with mGFR at screening showed a significant difference between the groups for all tomographic fat indexes (Supplemental Table S2). Although the significant differences in mGFR levels between the groups differed for each fat index, the quartile with the lowest fat index levels in general had significantly higher mGFR levels than the other three quartiles.

Table 2 Uni- and multivariable linear regression analyses of tomographic fat measurements with measured glomerular filtration rate standardized for BSA.
Full size table

Tomographic fat measurements and kidney function after living kidney donation

In primary unadjusted linear regression analyses with BSA-standardized mGFR at 3 months after donation, all tomographic fat measurements were significantly associated with mGFR (Table 2). In linear regression analyses adjusting for sex, age, and pre-donation mGFR, the negative associations between all of the tomographic fat measurements and BSA-standardized mGFR levels at 3 months after donation were significant (VATi: Β = -0.07, p < 0.001, SATi: Β = -0.13, p < 0.001, IMATi: Β = -1.56, p < 0.001, TATi: Β = -0.09, p < 0.001) (Table 2). Secondary regression analyses showed no significant associations between tomographic fat measurements and crude mGFR at 3 months post-donation (Supplemental Table S1).

Donors with higher pre-donation tomographic abdominal adipose tissue index levels experienced a significantly higher decrease in kidney function between screening for donation and 3 months after donation (Table 3).

Table 3 Linear regression analyses with delta BSA-standardized measured glomerular filtration rate screening – 3 months follow-up.
Full size table

Additional analyses with pre-donation BMI did not show a significant association with BSA-standardized mGFR prior to donation, but higher BMI levels were significantly associated with lower BSA-standardized mGFR at 3 months after donation (Β = -0.66, p < 0.001) (Supplemental Table S5).

Kidney function long-term after living kidney donation

For 314 donors (57% of the total study population, 170 male and 144 female), kidney function data at 5 years after donation were available. In analyses adjusted for age, sex, and pre-donation GFR higher levels of all tomographic fat measurements were significantly associated with lower BSA-standardized mGFR levels 5 years post-donation (Supplemental Table S3). Age-, pre-donation mGFR, and pre-donation weight-adjusted linear regression analyses with crude mGFR 5 years after donation did not show significant associations between tomographic fat levels and kidney function (Supplemental Table S4).

Discussion

We had hypothesized that higher levels of CT-determined abdominal fat are associated with lower kidney function. In this study, we found a significant association between CT-determined abdominal fat area and kidney function prior to and (in analyses with BSA-standardized mGFR) after donation, with a higher fat area associated with lower kidney function. The results of this study aid in elucidating the effects of body composition (measures) on kidney function.

Prior to donation, all abdominal fat compartments showed a significant association with kidney function, with higher CT-derived abdominal fat area values resulting in lower mGFR values. In studies among healthy individuals, visceral obesity was independently associated with kidney function impairment in all ages and both sexes, except for males < 45 years old18. VAT, SAT, and IMAT have been associated with kidney function decline and/or chronic kidney disease in the general population19, and VAT area independently affected estimated GFR (eGFR) levels in a study among healthy women20. Furthermore, baseline visceral fat area was associated with proteinuria in healthy individuals21. Interestingly, especially donors in the lowest fat index quartiles had higher mGFR levels compared to donors with higher levels of fat indexes in the present study.

After donation, at 3 months follow-up, the tomographic measured abdominal fat areas continued to be strongly associated with lower (BSA-standardized) mGFR levels. In addition, donors with higher pre-donation tomographic abdominal adipose tissue index levels experienced a significantly higher decrease in kidney function between screening for donation and 3 months after donation. A study among Japanese living kidney donors showed that kidney function 12 months after donor nephrectomy was significantly lower in donors with higher VAT area than in donors with lower VAT area22. Retroperitoneal adipose tissue measurement was also significantly correlated with a decrease in eGFR at the first and sixth month after donation in comparison with eGFR before donation23. Thus, it seems that donors with higher levels of abdominal fat areas have a lower kidney function level at which they start their post-donation kidney function trajectory. It is of interest to know how this association between pre-donation abdominal fat levels and post-donation kidney function develops long-term after donation. In the present study, 57% of the study population had available kidney function data 5 years after donation. Although analyses with these long-term kidney function data suggest a trend indicating that higher levels of tomographic abdominal fat areas are associated with lower kidney function at 5 years post-donation, further studies on long-term outcomes of living kidney donation are essential to completely elucidate this association.

The negative association between IMAT and kidney function was an interesting finding. IMAT is an indicator of myosteatosis, fat infiltration into skeletal muscle, and a predictor of a deviant course in numerous patient populations24,25,26,27. Most studies use skeletal muscle radiation attenuation or muscle density measured in HU as a surrogate marker for myosteatosis. The measurement of adipose tissue area in skeletal muscle on CT scans is a novel technique for assessing myosteatosis and may be a promising method in myosteatosis research. To the best of our knowledge, this is the first study to demonstrate a relationship between myosteatosis and kidney function in healthy living kidney donors. Further research that includes long-term clinical outcomes, such as the development of kidney disease, is needed to establish cut-off values for intramuscular adipose tissue and myosteatosis.

The association between abdominal fat area and kidney function before and after donor nephrectomy was most apparent when using mGFR standardized for BSA. In the secondary analyses with unstandardized mGFR, the association between abdominal fat area and mGFR was uncovered when weight was added to the regression model. Weight was added to account for body size, which is known to influence kidney function17. Standardizing GFR for BSA removes the effects of body size and although it is a known methodological practice, it is currently under discussion for its substantial consequences in populations with extreme body sizes28. Due to the screening guidelines for living kidney donation in our center (e.g., advice to lose weight if BMI > 30 kg/m2 and exclusion if BMI ≥ 35 kg/m2), the study population of the present study consisted of individuals with relatively ‘normal’ body sizes. BSA standardization has little consequences on GFR levels in individuals with such ‘normal’ body sizes and may enable more specific assessment of the relationship between abdominal fat area and kidney function in those individuals, independent of overall body size29.

The possible mechanisms underlying the negative association between abdominal adipose tissue and kidney function include direct and indirect effects. The direct interaction between adipose tissue and the kidney is referred to as the ‘adipo-renal axis’ and plays an important role in maintaining normal kidney function. Adipose tissue secretes a large number of factors which play an active role in the endocrine system30,31. Normal levels of these adipose tissue factors are important in preserving kidney function31. Excess caloric intake results in hypertrophy or hyperplasia of adipocytes, leading to several processes resulting in high levels of adipose-derived molecules and dysregulated metabolites, leading to oxidative stress, (chronic) inflammation, and kidney fibrosis, eventually causing kidney injury31.

There is also an indirect effect: an increased amount of visceral adipose tissue is associated with conditions including metabolic syndrome and diabetes32,33,34, which are risk factors for developing CKD35,36,37. Other possible underlying mechanisms could be related to mechanical stress on the kidney due to, for example, pressure exerted by perirenal adipose tissue. An increase in perirenal adipose tissue can cause kidney damage by a direct obstruction of the parenchyma and vessels, followed by an increase in sodium reabsorption and subsequently the development of high blood pressure38,39. Additionally, the compression of kidney parenchyma leads to an increase in interstitial hydrostatic pressure, resulting in reduced kidney blood flow and possible kidney disease progression37,38,39.

Much remains unknown about the possible mechanisms underlying the relationship between abdominal adipose tissue and kidney function, and future research, potentially incorporating investigations of (pathological) parameters such as interstitial fibrosis, tubular atrophy, and perirenal adipose tissue in protocol biopsies, is needed to further elucidate these connections.

Strengths of this study include its relatively large cohort size, kidney function measurements (as opposed to estimates), and limited missing data (< 5%). Limitations are the lack of severely obese donors, limited generalizability due to the majority of participants being of European descent, single institution and retrospective design of the study. The earliest time point after donation at which mGFR was measured was 3 months post-donation. This impaired analysis of change in single-kidney mGFR of the remaining kidney during the first months following donation.

This study found that higher levels of abdominal fat, as measured by CT analysis, were associated with lower pre- and post-donation kidney function. Although the magnitude of the impact of pre-donation abdominal fat area on short-term post-donation kidney function may seem small in comparison with factors such as age and pre-donation mGFR, body composition is one of the few factors that may be optimized prior to donation. This study also raises questions about the use of BMI as a gold standard measure3,4,5 for screening donors, as it does not provide information on abdominal fat areas. BMI did not show an association with mGFR prior to donation, where the tomographic fat indexes did. CT analysis, which is routinely performed in most centers as part of screening for donation, may be a more effective way to assess body composition and associated health risks. Further research is needed to translate this knowledge to clinical practice, determine the effectiveness of CT-derived body composition measurements in donor screening guidelines, and investigate the potential benefits of abdominal fat reduction in living kidney donors.

This study shows that a higher abdominal fat area, measured by CT analysis, is associated with lower kidney function at the time of screening and after living kidney donation. The results of this study aid in elucidating the effect of body composition on kidney function in living kidney donors. Further investigations are imperative to explore the association between adipose tissue measured via CT and kidney function in various (living kidney donor) populations, particularly those affected by obesity. Additionally, radiological assessment of for example perirenal adipose tissue and its influence on kidney function may be of considerable interest in expanding our understanding of living kidney donation and kidney outcomes.