Cardiovascular diseases are still the main cause of death after kidney transplantation. In addition to the classical risk factors (obesity, diabetes mellitus, hypertension, hyperlipidemia, smoking, etc.), dialysis periods before transplantation, graft function after transplantation, proteinuria, acute rejection episodes, post-transplant diabetes, hyperhomocysteinemia and immunosuppressive drugs are also important risk factors [20].

An increase in body weight is commonly observed after transplantation [21]. The main factors underlying the post-transplant weight gain are the recovery of uremic elements, decreased diet restrictions, and increased appetite after the transplantation and immunosuppressive treatments [22]. In our cohort, weight gains were comparable in the CsA and Tac groups (median 3.75 and 4.15 kg in the 12th month, 5.60 and 4.95 kg in the 24th month, 6.25 and 8.95 kg in 36th month and 7.0 and 10.25 kg in the 48th month, respectively). Localization of adipose tissue plays a role in the development of many diseases [23]. High BMI (≥ 30 kg/m2) and BF% (> 25% in men, > 35% in women) indicate general obesity [18]. However, both do not provide information about where adiposity is localized in the body. BMI does not help differentiate muscle and fat tissue, nor does it supply adequate information about peripheral and central adiposity. Therefore, WaC (> 102 cm in men, > 88 cm in women), HC and WHR (> 0.9 in men, > 0.8 in women) have become more commonly used anthropometric indices in recent years [19]. Abdominal obesity is associated with insulin resistance and metabolic syndrome and determines cardiovascular risk. WaC and WHR are widely used as abdominal (visceral) obesity indicators. A study showed that the WHR was indicated to be superior to BMI and WaC in the cardiovascular risk assessment [24]. In other study in 122 patients with chronic kidney disease, a strong correlation was reported between WaC and visceral adiposity [14]. HC is an indicator of gluteo-femoral adiposity, which is more common in women and less hazardous for health compared to visceral adiposity [25].

In the present study, mean body weight, BMI, WaC and HC values increased significantly in both treatment groups starting from 3 months post-transplant. WHR increased significantly in the first three months in the CsA group and the first six months in the Tac group compared to the baseline values but did not change significantly in the following months. WaC measurements in the early post-operative period may be misleading due to post-operative oedema in the abdomen and waist regions, especially in the skin and subcutaneous areas. Thus, WHR is overestimated in the first three-six months after surgery. Gradual improvement in oedema and increase in adiposity over the following months may increase WaC and HC, although there is no change in WHR. In our cohort, median BF% values in the pre-operative period were 22.8% in the CsA group and 23.0% in the Tac group. The BF% increased significantly after three months in the CsA group and 24 months in the Tac group up to the 48-month post-transplant period. Our findings provide little explanation for the differences between treatment groups regarding BF% change after transplantation since there is insufficient data about our patient’s dietary habits and lifestyle changes during the post-transplant period. In a past study conducted with transplant patients, glucocorticoid and Tac treatment were reported to be associated with increased BF% and post-transplant diabetes mellitus [26].

Basal body weight, BMI, WaC, WHR and BF% values of both groups were similar, and there was no difference between percentage changes in body weight, BMI, WaC, WHR and BF% throughout the study. The increase in mean HC values in the CsA group was significantly higher than in the Tac group (median 2.80% vs. 0.98% in the 1st month, 7.84% vs 4.96% in the 12th month, and 8.10% vs. 3.00% in the 24th month, respectively). Although the increase in HC values in the 36th (median 8.69% vs. 5.18%) and 48th (median 8.13% vs. 5.24%) months was higher in the CsA group, the difference did not reach statistical significance. Both CNIs increased BMI and overall adiposity. Since the increase in WHR was similar in both groups, we can conclude that the higher HC increase in the CsA group compared to the Tac group did not affect the metabolic parameters. The earlier increase of BF% in the CsA group indicates that CsA increased overall adiposity, and the accumulation was predominantly in the femoral region. On the other hand, general body adiposity in the Tac group increased in a later period.

Our previous retrospective study showed that CsA use was one of the independent predictors of weight gain 12 months post-transplant [27]. Weight gain may be associated with the continued increase in adipose tissue due to hyperlipidemia and higher water-sodium retention after CsA treatment, alterations in steroid treatment dose and duration, and poor patient compliance with dietary recommendations [28, 29]. Our present study’s anthropometric assessment included WrC and NC measurements. WrC values showed a significant early increase at 3 and 6 months in the Tac group compared to baseline values but decreased significantly at 48 months in the CsA and Tac groups. Compared to baseline values, NC increased significantly between the 1st and 24th months in the CsA group and between the 1st and 12th months in the Tac group but did not change in the following months. No significant difference was observed between WrC and NC percentage changes in the CsA and Tac groups. The early increase in WrC in the Tac group may indicate that Tac-related weight gain results in more homogeneous adiposity. In recent years, wrist-ankle measurement has been used to determine cardiovascular risk factors such as insulin resistance and hypertension [30,31,32]. WrC measurement can give information about body bone structure and peripheral fat distribution. Insulin shows anabolic effects by binding to insulin-like growth factor-1 (IGF-1) receptors in osteoblasts and can change bone mass and density [33, 34]. In the pre-transplant period, many factors, especially uremia, deterioration in calcium-parathormone levels and a diet poor in protein, cause insulin resistance [35]. Glucocorticoids used after transplantation cause dose-dependent peripheral and central insulin resistance and suppress insulin-related lipolysis. They also cause less accumulation of adipose tissue in peripheral areas compared to the central visceral adipose tissue of the body. During the follow-up period, insulin sensitivity increases, insulin resistance decreases with decreasing doses of drugs, increasing urea elimination and improving calcium-parathormone levels. There was no significant difference between the percentage changes in NC values ​​in the two groups after transplantation. Also, concomitant steroid therapy rather than CNIs may be responsible for the increase in NC. As seen in patients with Cushing’s syndrome, increased NC values ​​may be associated with hypercortisolemia and trunk obesity due to decreased insulin sensitivity.

Dyslipidemia is common among kidney transplant recipients, and new-onset or worsening dyslipidemia has been associated with the use of sirolimus, CNIs (especially cyclosporine), and glucocorticoids. Increases in total- and LDL-cholesterol levels are more common [36]. CsA specifically causes an increase in very low-density lipoprotein (VLDL) cholesterol, LDL-cholesterol and triglyceride levels by inhibiting LDL-cholesterol receptor synthesis and lipoprotein lipase activity and increasing apolipoprotein C-III and proprotein convertase subtilisin/kexin type 9 levels [37]. Compared with pre-transplant values in our study, there was a significant increase in total-cholesterol and HDL-cholesterol values in both groups. Compared with pre-transplant values, there was an increase in LDL-cholesterol values up to the 24th month in the CsA group and all months (except the 12th month) in the Tac group. No significant change was observed in triglyceride values. In our study, the changes in the percentage of total cholesterol were significantly higher in the CsA group in the 1st and 12th months than in the Tac group, suggesting that the hyperlipidemic effect of CsA was more prominent in the first year. Percentage changes in other lipid parameters of the groups were also comparable. In our study, the proportion of patients with post-transplant dyslipidemia in the CsA group was significantly higher than in the Tac group. Increases in total-cholesterol, LDL-cholesterol and triglycerides appear to correlate with anthropometric measurement increases during both CsA and Tac treatments. Therefore, especially the hyperlipidemic effect of CsA may be related to adiposity.

The increase in WaC and insulin resistance is expected to lead to an increase in TG levels and a decrease in HDL-cholesterol levels. HDL-cholesterol levels increased in both groups, whereas both CNIs did not change TG levels. This status can be explained by reducing post-transplant insulin resistance in uremic patients. However, our study did not measure insulin resistance. LDL cholesterol is the lipid fraction most closely associated with atherogenicity. Increasing WHR and WaC values ​​may increase atherogenicity and thus LDL-cholesterol levels. Many factors may have affected our study's analysis results regarding lipid parameters. During follow-up, physicians may have reduced doses or discontinued the drug in some patients using statins, mainly due to concerns about its interaction with CsA and graft dysfunction due to the risk of myopathy. An antilipidemic drug may be added to the treatment afterwards. In addition, these patients use many drugs that may affect insulin resistance.

In our study, Tac was not preferred in diabetic patients due to its higher diabetogenic potential than CsA, especially if the recipient has low immunological risk. Pre-transplant diabetes ratios were 18.3% in the CsA group and 1.5% in the Tac group. The difference in diabetic ratios of the groups may explain the higher basal serum glucose levels in the CsA group compared to the Tac group. When compared with baseline values, serum glucose levels decreased in the 12th months in the CsA group and increased in the 1st month in the Tac group. There were no significant changes at other time points. On the other hand, glucose percentage changes at 6 (median 5.40% vs. –7.78%, p = 0.049), 12 (4.41% vs. –7.76%, p = 0.030) and 24 (4.72% vs. –4.47%, p = 0.045) months were significantly higher in the Tac group than in the CsA group, respectively. CsA and Tac increase the risk of post-transplant diabetes [38]. However, Tac is more diabetogenic than CsA because Tac causes more severe swelling and vacuolization of islet cells [39]. Both drugs cause reversible toxicity in pancreatic beta cells, especially in the early post-transplantation period, directly affect the transcriptional regulation of insulin expression, cause glucose intolerance and inhibit lipolysis [40,41,42,43,44].

The relatively higher percentage of diabetes patients in the CsA group may also have been a factor in weight gain. However, the less diabetogenic effect of CsA in these patients, the use of lower doses of steroids and faster dose reduction for glycemic control may have facilitated glycemic control. In diabetic patients in the CsA group, impaired insulin-glucose homeostasis leads to hypertrophy of adipocytes and fat accumulation, especially in the lower part of the body [40, 45]. Tac and CsA can inhibit glucose entry into the cell by increasing the internalization of glucose transporter-4 (GLUT4) at different rates on the cell surface of adipocytes and by causing phosphorylations on insulin receptors at various points.

Post-transplant diabetes mellitus is a multifactorial condition that occurs in 4–25% of kidney transplant patients and within the first three months of transplantation in most cases [46, 47]. In a meta-analysis of three large-scale studies including 980 transplant patients, Tac therapy was associated with a 5.03-fold higher risk of post-transplant diabetes mellitus development [48]. However, post-transplant diabetes developed in more patients in the Tac group (19.1%) than in the CsA group (8.3%) during follow-up. After 36 months, weight gain in the whole cohort tended to increase more rapidly in the Tac group, which may support the continued diabetogenic effect of Tac in the late period. Although patients often received a similar corticosteroid regimen, we administered different doses to some patients. The steroid dose they were exposed to could have been very important in anthropometric measurements. However, we did not detect any difference in cumulative steroid doses between the groups in all months throughout our study. Depending on the use of CNIs at different doses and durations, fat distribution may also differ due to its diverse effects on glucose uptake [49, 50]. In our study, heterogeneous distribution of diabetic patients in the groups, insulin doses and cumulative steroid doses might have affected our results. Nephrologists preferred CsA or Tac treatment for some diabetic patients due to medical concerns. Transplantation is associated with an increased need for insulin. While most patients received post-transplant insulin therapy, very few patients used only oral antihyperglycemic drugs. Insulin has significant anabolic effects and could have affected study parameters, including weight gain. Even the insulin dose requirements were different during follow-up in patients with type 1 and type 2 diabetes and de novo diabetes mellitus.

Two subgroup analyses evaluated the effects of factors such as heterogeneity in fat distribution and the anabolic effect of insulin therapy on anthropometric measurements in diabetic and de novo diabetic patients. When compared anthropometric measurements in diabetic and non-diabetic patients, only baseline BF% values were significantly higher in patients with diabetes than in non-diabetics. The rates of increase in NC in the 12th month in diabetic patients and BF% at the 24th and 36th months in non-diabetics were significantly higher than the other group. When anthropometric measurements in insulin-using and non-insulin-using patients were compared, the baseline anthropometric measurements were comparable. The increase rates in NC from the 3rd month to the 24th month and WaC in the 12th month were significantly higher in patients who used insulin than in those who did not. Insulin treated patients are under the anabolic effect of the drug which can cause fat accumulation. Diabetogenic drugs given for the prevention of transplant rejection, mainly steroids, can cause fat accumulation in the central parts of the body like neck and waist. This accumulation increases insulin resistance further, requiring a higher dose to overcome poor glycemic control. This vicious circle can explain the difference between diabetic and non-diabetic subjects and insulin and other antihyperglycemic drug users among the diabetic group.

Metabolic disorders that can lead to hyperglycemia, insulin resistance, hypertension, dyslipidemia after kidney transplantation increase the risk of overweight and obesity [10]. A recent study revealed a significant association between greater weight gain and the youngest age, female gender, lower pre-transplant BMI, living kidney donor, and fewer post-transplant hospitalizations [51]. Another study showed that shorter pre-transplant dialysis time, a living kidney donor, and being obese at baseline increased the risk of weight gain [52]. Weight gain is expected in all recipients, especially in the early post-transplant period. In the present study, we determined obesity risk factors according to different anthropometric parameters beyond weight gain in a more extended period (4-year follow-up). We investigated the factors that were the primary statistically determinant of the development of obesity according to the patient’s BMI, BF%, WaC and WHR values at the end of the 48th month. These risk factors included baseline BMI, young age, and lack of exercise for BMI; baseline BMI, baseline WrC, eGFR level at one month, baseline NC, and no exercise for BF%; baseline WaC, BMI and WrC values for WaC and baseline WaC, history of dyslipidemia and no exercise for WHR. After adjusting for other factors, a high baseline BMI increased the risk of being obese (based on BMI, BF% or WaC) by 1.49–2.09 times. Lack of regular exercise increased the risk of obesity (based on BMI, BF% or WHR) by 4.75–30.93 times. High 1-month eGFR increased the risk of obesity 1.04-fold (based on BP) and high baseline NC 1.49-fold. High baseline WaC increased the risk of obesity based on WaC by 1.18 times and the risk based on WHR by 1.11 times. In those with a history of dyslipidemia, the probability of developing obesity based on WHR was 4.66 times higher than those without a history of dyslipidemia. A recipient with a high baseline BMI, NC and WaC may have a higher risk of developing obesity due to hyperplasia of adipose cells. Dyslipidemia and not regular exercise can also increase the risk of obesity by increasing insulin resistance or as an indicator of high insulin resistance. Good early graft function (a higher eGFR) may trigger an increase in appetite and weight gain due to the improvement of uremic symptoms and metabolic changes, possibly with the contribution of immunosuppressive drugs. Age increase was associated with a 10% reduction in the risk of obesity according to BMI. Kidney recipients paying more attention to their health or reducing their food intake in old age may be associated with a decrease in BMI. High baseline WrC was associated with a 70% and 50% reduction in obesity risk according to BF% and WaC, respectively. WrC with less adipose tissue accumulation may not be a predictor of obesity in these patients.

Nephrotoxicity is considered the most critical side effect of CNI treatment. CNIs lead to graft dysfunction via dose-dependent renal vasoconstriction and the development of tubular atrophy and interstitial fibrosis in the chronic course [53]. In the current study, while 6th and 48th-month eGFR values were significantly lower than the 1st-month levels in the Tac group, no significant difference was noted between Tac and CsA groups in terms of serum creatinine and eGFR levels during the entire 4-year post-transplant period. Similarly, in a past study by Alghamdi et al. [54], in CsA and Tac-treated patients, no significant difference was reported between treatment groups regarding 2-year follow-up data on serum creatinine levels. In two studies with 5-year follow-up of CNI-treated patients by Kaplan et al. [55] and Vincenti et al. [56], the authors noted significantly higher eGFR in the Tac group. The decrease in eGFR values in our patients receiving Tac therapy within the 6th month of transplantation could be explained by the CNI toxicity, urinary tract infection and inadequate oral intake. However, we did not detect any difference between the episodes of urinary infection in the two groups in the first year after transplantation. Implementation of serum urea and creatinine measurements only at a single time point is an essential limitation of the current study, given the likelihood of fluctuations in serum urea and creatinine levels between measurement dates as well as the changes in clinical condition with possible impact on the graft function such as the development of acute kidney injury.

The main limitations of this single-centre study are the relatively small number of patients, the inability to extend the study over a longer period, and the lack of detailed information about patients' dietary adherence, water consumption, and other medications patients take. One hundred twenty-eight patients were included in the study, and 111 completed the study after four years. The younger age of Tac patients, which may be related to the physicians’ preference of prescribing Tac in younger patients to avoid the remarkable cosmetic side effects of CsA therapy (i.e. hirsutism, gingival hypertrophy) in this age group, might have also affected the findings achieved in the current study.