Variable Selection for Assessing Risk Factors for Weight and Body fat Gain During the First Year After Kidney Transplantation

Background: Body fat and overall weight gain are common after kidney transplantation and are associated with poor clinical outcomes. Therefore, identification of at-risk patients is relevant for preventive interventions. Clinical Question: What variables influence weight and fat gain in patients in the first year after kidney transplantation? Literature Search Prospective and retrospective cohort studies published in or after 2001 naming fat and/or overall weight gain during the first year after kidney transplantation as outcome variable(s) were systematically searched in Medline/Pubmed in November 2018 and March 2022. Clinical Appraisal: We identified 16 studies examining a wide variety of potential factors influencing weight and fat gain over the first posttransplant years. These included genetic, socio-demographic, behavioral, biomedical, psychological and environmental factors. For a number of variables, study results were contradictory: some studies indicated preventive impacts on weight or fat gain; others concluded that the same factors increased it. Cases were discussed with 2 clinical experts. We eventually agreed on 13 potentially relevant risk factors for post-transplant weight/fat gain: age, gender, genes, income, ethnicity, education, eating habits, physical activity, smoking cessation, baseline BMI, baseline fat, depression and perceived overall wellbeing. Integration into Practice Before integration into clinical practice, a critical evaluation of all potential risk factors’ suitability for assessment will be necessary. In addition to feasibility, operational definitions and measurement methods must also be considered. Evaluation: To reduce the list of risk factors to the most relevant, a first testing within a prospectively collected data set is planned.


Background / Significance of the Problem
Body fat and overall weight gain are common after kidney transplantation, and are known risk factors for poor clinical outcomes such as mortality. 1 Even in patients with normal pre-transplant weight, unmonitored weight gain can lead to overweight or obesity. 2 Therefore, early identification of patients with a high risk for post-transplantation weight or body fat gain is important, as it allows timely initiation of preventive interventions. To identify risk factors for weight/fat gain in the first year after kidney transplantation, an evidence-based approach was chosen.

Clinical Question
Our clinical question was "What variables influence body fat and overall weight gain during the first year after transplantation?" To ensure that causality could be inferred, only longitudinal studies were included.

Search of the Literature
Medline / Pubmed were searched in November 2018 that included longitudinal prospective and retrospective cohort studies that investigated risk factors for weight gain, body fat gain, or body mass index (BMI) outcomes. Our initial returns were updated in March 2022 using the following search strategy: Age Controversy: Younger age as risk factor, 4-8 no effect [9][10][11] An important moderating factor Ethnicity Controversy: Black ethnicity (vs all other ethnicities) as risk factor for weight gain, 5, 7 no effect of black ethnicity (vs all others) on weight gain 6,12 Risk factor Income Controversy: Lower income as a risk factor for weight gain, 7 higher income as risk factor for weight gain 6 Risk factor in general population Gender Controversy: Female gender as risk factor for weight gain at month 12, 5, 7, 8 no effect of gender (either female or male) on weight gain 6,9,12 Risk factor in general population

Education
No evidence in kidney transplantation Risk factor in general population

Behavioural factors
Eating habit Controversy: Self-reported carbohydrate consumption was a risk factor for weight gain in the multivariate model, 6 other indicators such as baseline kilocalories of energy, total fat, total carbohydrates and total protein on weight gain were identified as risk factors only in simple correlation. 6 Self-reported percentage of fat intake, percentage carbohydrate, percentage protein intake was not identified as risk factor in either simple or multivariate analyses. 6 Self-reported intake of mono-& disaccharides, intake of energy-rich drinks/dairy, and vegetable intake were identified as risk factors for fat gain only in mean difference analysis. 3 Eating habits are known to be the main factor for weight gain, but the applied self-report methods in those studies may not be specific enough.

Risk factor
Physical activity Identified as risk factor: Lower number of steps per day (assessed with pedometer) and lower self-reported physical activity (time × intensity) as risk factors for fat gain. 3 Lower number of minutes of moderate-to-vigorous intense activity or more time spent sedentary (assessed with pedometer) were not significantly associated with fat gain. Self-reported physical activity, days of activity and time sleeping were not identified as risk factors for weight gain in a simple correlational design. 6

Risk factors in general population
Smoking cessation Lack of evidence in kidney transplantation Risk factor in general population, eg, Hu 2018 13

Biomedical factors
Creatinine during first year / creatinine clearance Not identified as risk factor: Creatinine clearance during year 1 was not identified as risk factor. 11 Lack of evidence: Other studies investigated creatinine only at month 12. 5,10 No risk factor Rejection episodes Controversy: Only one study reported having "no rejection episode" as a risk factor for weight gain, 7 whereas others did not detect a significant association 5,10,11 No risk factor, having no rejection may be linked to higher wellbeing Hospitalization episodes Weak: Lower number of hospitalizations is a risk factor for weight gain 8

No risk factor
Baseline BMI (at day of transplant) Identified as risk factor: Higher BMI 5 or obesity (BMI > 30) 14 as risk factor; lower BMI 8,10 or BMI below 25 14 as risk factor Risk factor Prednisone dose Controversy: Cumulative dose 11, 15 or withdrawal after 100 days 16 were not identified as risk factors in some studies, whereas cumulative dose 10 and steroid withdrawal after 7 days 14,17 were identified as risk factor in other studies.

No risk factor
CyA/Tac Controversy: Type of immunosuppressive medication has effect on weight gain, 10 whereas others do not 5 No risk factor Oxidative stress Weak: Identified as risk factor in one study (only two group comparison, not regression model) 18 No risk factor Dyslipidemia Controversy: effect of total cholesterol, whereas other parameters (egHigh density lipoprotein) were not 6,10 No risk factor, but associated with metabolic syndrome Higher trunk fat Weak: Identified as risk factor 6 (continued) The literature search yielded 533 articles, the reference lists of which led to 5 more, resulting in a total of 538 articles. Studies were excluded if any of the following criteria applied: No prospective or retrospective cohort design with body weight as outcome (N = 500); review or protocol for future study (N = 4); reporting inconsistencies (N = 1); no measurement at month twelve (N = 10); influencing variables not measured before month twelve (N = 3); not restricted to kidney transplantation (N = 1); or the operationalization of the outcome variable(s) does not allow the drawing of inferences regarding the influencing variable (N = 2); language of publication neither English nor German (N = 1).

Results of the Literature Review
Of the 538 studies returned by the search, 16 fulfilled the criteria for inclusion in the narrative review. Of those, only 1 focused on body fat gain 3 ; the other 15 explored risk factors for weight gain. All included studies were published between 2001 and 2020; the majority (N = 11) were conducted in the United States. Their common strength was their prospective (N = 11) or retrospective (N = 5) cohort design, which clarified the causality both of factors and of outcomes. Their most common limitation was a small sample size: 7worked with fewer than 100 transplant recipients. The selected studies investigated a wide variety of risk factors on weight gain over the first years posttransplant. These included genetic, socio-demographic, behavioral, biomedical, psychological and environmental variables. The study results were synthesized narratively. Based on the evidence presented, the risk factors were divided into 5 groups: Identified risk factors: Genes, lower levels of physical activity, living donation and baseline BMI. Not identified as risk factors: Creatinine clearance, pretransplant diabetes, dialysis modality, the presence of calcineurin inhibitors, and haemoglobin levels. Risk factors with controversial results, ie, variables for which some studies reported effects, while others did not or even noted contrary effects: age, ethnicity, income, gender, eating habits, rejection episodes, prednisone dose, immunosuppressive regimen, dyslipidemia, hypertension, food availability, depression. Risk factors with very weak evidence, ie, those for which only one study found an effect: number of hospitalization episodes, perceived overall well-being, oxidative stress, and trunk fat.

Risk factors with no evidence in kidney transplantation:
For education, smoking cessation and creatinine before month 12, no evidence exists regarding kidney graft recipients.

Expert opinions
All factors, but especially those with controversial or weak evidence, were discussed with 2 clinical experts independently. Both experts had broad clinical and research expertise in weight gain and solid organ transplantation, held university degrees (PhD or MD) and had published in this field of research. Any variables that at least 1 expert associated with Body composition parameters are more specific than BMI, therefore, must be included.
Risk factor, ATM may be more sensitive than BMI Donor type Weak: Living donation identified as risk factor 5,8 No risk factor, wellbeing is probably the moderating factor here. Pretransplant diabetes / glucose Not identified as risk factor [5][6][7]10 No risk factor, but is associated with metabolic syndrome Dialysis modality Not identified as risk factor: Peritoneal dialysis may be protective for weight gain, 11, 15 but due to differences in baseline weight (peritoneal dialysis patients have higher baseline weight), no risk factor (implicitly) 10 No risk factor Hypertension Controversy: Hypertensive diastolic value as risk factor, whereas systolic value is not a significant risk factor 10 No risk factor, but is associated with metabolic syndrome Calcineurin inhibitor Not identified as risk factor: No effect 15 No risk factor Haemoglobin Not identified as risk factor: No effect 10 No risk factor Psychological factors Depression Weak: Center for Epidemiologic Studies Depression Scale (CES-D) score had no effect on weight gain 6,19 Risk factor in general population Well-being Weak: Identified as risk factor 6 Risk factor, may be an important moderating factor Environmental factors Food availability Controversy: Effect of number of grocery stores within 1-mile radius on BMI change at month 12, no effect of availability of fast food restaurant or convenience stores on BMI change at month 12 9 No risk factor, may be culture-specific increased risk were counted as potential risk factors. Of the risk factors identified in the literature or those not previously identified as risk factors, the experts assessed only living donation differently from the published evidence: both saw a living donation as associated with better well-being; therefore, the experts recommended not to treat it as a risk factor. The results of the literature review and expert opinion are displayed in Table 1. Age, gender, genes, income, ethnicity, education, eating habits, physical activity, smoking cessation, baseline BMI, baseline body fat, depression and well-being were identified as potentially relevant factors for posttransplant weight and/or body fat gain.

Integration into Practice
Based on the literature review and expert opinions, a number of potential risk factors have been identified. Before these can be integrated into clinical practice, further steps are necessary. One of these is to critically appraise each identified factor's suitability for clinical use. To do that, the affordability, practicability, effectiveness and cost-effectiveness, acceptability, side-effects and safety, and equity criteria (APEASE) can provide a valuable framework. 17 While age, gender, ethnicity, education, smoking status and BMI fulfill all APEASE criteria and may be ready for adoption into clinical practice, others do not. As a result, those gauged less suitable will be more challenging to integrate into clinical settings.
For example, while assessments of depression and wellbeing were administered as part of each prospective kidney recipient's assessment for registration to a graft waiting list, they may not be part of routine follow-up care. Implementing them would be resource-intensive, particularly if administered directly by a psychologist, and may not be well-received by all patients. And while the affordability, practicability, costeffectiveness and acceptability may all be higher for a different type of operationalization and/or of measurement method, eg, self-reporting via questionnaire, those methods would also likely be less accurate. Consequently, in addition to the APEASE-criteria, each proposed assessment method's sensitivity and specificity must also be considered.
Similarly, while dual-energy x-ray (DEXA) scans are the gold standard method to measure body fat, their cost per use (including technician time) makes them unfeasible for frequent application. For routine clinical use or home monitoring, various simple, costeffective and reasonably accurate methods of calculating body composition are available. Bioelectrical impedance analysis (BIA), for example, is very quick, inexpensive and acceptably precise, generally making it a more feasible option. 18

Evaluation of Evidence-Based Practice
To foster the uptake of routine assessment of risk factors for posttransplant weight and body fat gain in post-transplant care, changes in clinical practice should be tailored to the local setting. A careful analysis of that context (ie, values and beliefs, treatment guidelines, reimbursement policies) may help the interventionists engage stakeholders and decision makers from a very early stage. 19 The assessment of body trunk fat, physical activity, depression and well-being, and eating habits requires patient participation and may even involve considerable effort on their part; therefore, their acceptance and engagement are keys to enhancing their adoption of this intervention. As implementation outcomes, patients' and health professionals' acceptance and adoption of the risk assessment may also function as quality indicators for the evaluation of evidence-based practice.

Future Areas for Investigation
The aim of the literature review and expert opinion was to produce a selection of potential factors behind increases in overall weight (BMI) or body fat in the first year following kidney transplantation. It is not yet clear which of those identified have the strongest potential to predict weight gain. Therefore, the next major step will be to reduce the potential risk factors to the most relevant.