Kidney disease in adults with Prader-Willi syndrome: international cohort study and systematic literature review

Background Prader-Willi syndrome (PWS) is a rare, complex, genetic disorder characterized by hyperphagia, hypotonia, delayed psychomotor development, low muscle mass and hypothalamic dysfunction. Adults with PWS often have obesity, hypertension and type 2 diabetes mellitus (DM2), known risk factors for cardiovascular disease (CVD) and chronic kidney disease (CKD). Early symptoms of CVD and CKD may be masked by intellectual disability and inability to express physical complaints. Furthermore, kidney diseases are often asymptomatic. Therefore, renal and cardiovascular disease might be missed in patients with PWS. Microalbuminuria is an early sign of microvascular damage in the kidneys and other vascular beds. Therefore, we screened our adult PWS cohort for the presence of elevated urinary albumin and (micro)albuminuria. Methods We retrospectively collected anthropometric measurements, blood pressure, medical history, medication use, urine dipstick and biochemical measurements form electronic patient files. In addition, we performed a systematic literature review on kidney disease in PWS. Results We included 162 adults with genetically confirmed PWS (56% male, median age 28 years), of whom 44 (27%) had DM2. None had known CVD. All subjects had normal estimated glomerular filtration rate (eGFR) according to non-PWS reference intervals. Elevated urinary albumin or (micro)albuminuria was present in 28 (18%); 19 out of 75 (25%) had an increased urinary albumin-to-creatinine ratio (UACR) and 10 out of 57 (18%) had an increased urinary protein-to-creatinine ratio. Elevated urinary albumin was present at a young age (median age 26 (IQR 24-32) years) and was associated with an significantly higher BMI and LDL-cholesterol levels and higher prevalence of DM2, hypertension and dyslipidemia than those with normal UACR (p=0.027, p=0.019, p<0.001, p<0.001, p=0.011 and respectively). Conclusion Upon screening, one in every five adults with PWS had increased urinary albumin or (micro)albuminuria, early signs of microvascular disease. All had normal eGFR, according to non-PWS reference intervals, and none had a formal diagnosis of CVD. As muscle mass is low in PWS, creatinine levels and eGFR may be spuriously normal. Urinalysis in this patient group can be used as a screening tool for microvascular (kidney) disease. We propose an algorithm for the detection and management of microvascular disease in adults with PWS.

Background: Prader-Willi syndrome (PWS) is a rare, complex, genetic disorder characterized by hyperphagia, hypotonia, delayed psychomotor development, low muscle mass and hypothalamic dysfunction.Adults with PWS often have obesity, hypertension and type 2 diabetes mellitus (DM2), known risk factors for cardiovascular disease (CVD) and chronic kidney disease (CKD).Early symptoms of CVD and CKD may be masked by intellectual disability and inability to express physical complaints.Furthermore, kidney diseases are often asymptomatic.Therefore, renal and cardiovascular disease might be missed in patients with PWS.Microalbuminuria is an early sign of microvascular damage in the kidneys
With improved healthcare for patients with PWS, life expectancy has increased.As patients with PWS become older, age-related diseases like CVD and CKD become increasingly relevant.Early detection and treatment are crucial to prevent complications, loss of quality of life and early mortality (24,25).
As symptoms are often atypical, non-specific or even absent (26-29), the detection of CKD can be challenging.In patients with PWS, the detection of CVD and CKD might be even more complicated due to the presence of intellectual disability (mild in the majority of patients) and reduced ability to express physical complaints (30,31).Furthermore, patients with PWS have a high pain threshold that might mask the sudden thoracic pain or headache that may indicate cardiac or cerebrovascular events.The lack of fever and pain in people with PWS can also mask other risk factors for CKD, including chronic urinary tract infections (UTIs) (1,8,32,33) that may further increase the risk of CKD (34).Although never investigated, CKD and CVD might be more difficult to detect due to the combination of physical, behavioral, and neurocognitive potentially increasing morbidity and mortality in adults with PWS.
CKD can be diagnosed based on a decreased (estimated) glomerular filtration rate (eGFR) and the presence of (micro) albuminuria (35).(Micro)albuminuria can be defined as an increased urinary albumin-to-creatinine ratio (UACR) or urinary protein-to-creatinine ratio (UPCR) and is associated with an increased mortality independent of the eGFR (36,37).Preferably, the UACR should be assessed in first morning urine samples as it corresponds more closely with a 24-hour urine collection, which is the golden standard (35,38,39).Microalbuminuria can indicate microvascular injury and is therefore a risk factor for CKD and CVD (40)(41)(42).In patients with diabetes mellitus, microalbuminuria is predictive of early mortality (43).Besides (micro)albuminuria, an increase in the exertion of low molecular weight (LMW) proteins such as cystatine-C, ubiquitin, retinol binding protein or alpha1microglobulin, might also indicate kidney disease (44)(45)(46).
Little is known about CKD and CVD in the general population of adults with PWS, as these co-morbidities have only been studied in specific 'subtypes' of patients with PWS.In hospitalized patients with PWS, the prevalence of CKD was estimated to be 6.5% (47).In PWS patients with DM, microalbuminuria was found in 6.9-55.6% and proteinuria in 3.4-11.1% of (48)(49)(50), which is similar to the prevalence in the general population with DM (51).Apart from these studies in PWS 'subpopulations' of hospitalized and/or diabetic PWS patients, microalbuminuria and proteinuria have not been systematically assessed in the broader adult PWS population.This lack of information is relevant, as both microalbuminuria and proteinuria are associated with an increased risk of CKD and cardiovascular morbidity and mortality (40 -42, 52-56).Also, to our knowledge, systematic screening for kidney disease (by both urine and serum analysis) has never been performed in adults with PWS.As a result, there are no guidelines for screening and treatment of CKD in adults with PWS.
To fill this knowledge gap, we have systematically screened urine and serum samples for signs of CKD in a cohort of adults with PWS.We also assessed risk factor for CKD and performed a systematic literature review on kidney and urinary tract disease in PWS.Based on our cohort study and systematic review, we provide practical recommendations for the screening and treatment of (micro)albuminuria and CKD in patients with PWS.

Method
This international cross-sectional study was performed at the Center for Adults with Complex Rare Genetic Syndromes (CRGS) at the Erasmus University Medical Center, Rotterdam, the Netherlands, the Rare Disease Center of reference 'Prader-Willi Syndrome and obesity with eating disorders', in Paris, France and the Division of Auxology in Piancavallo, Italy.Only adults with genetically confirmed PWS who visited the outpatient clinic of one of these reference center between January 2020 and December 2022 were included.Urinalysis was part of a routine systematic health screening in all included PWS patients.Ethical approval and/or individual informed consent was obtained by the Medical Ethics Committee of participating centers according to local rules and regulations.
As previously described (20), the systematic screening generally consists of a structured interview, an extensive physical examination, a medical questionnaire (depending on local guidelines), a review of medical records and biochemical measurements.The systematic screening took place during the first or follow-up visit to the outpatient clinic.If measurements could not be done during the first visit (due to behavioral or logistic issues), the next available measurement was used for the statistical analysis, provided that the time interval between first and follow up visit was less than 12 months.
Height and weight measurements were collected from medical records.When multiple measurements were available, we included the measurement the was closest (in time) to the biochemical and urinalysis.A Body Mass Index (BMI, in kg/m 2 ) between 18.5 and 24.9 kg/m 2 was considered lean, between 25.0 and 29.9 kg/m 2 overweight, 30.0 or more kg/m 2 obese and a BMI of 40.0 or more grade III obesity according to the World Health Organization Criteria (57).
Resting blood pressure was measured in all patients during their routine visit to the outpatient clinic.In the Dutch cohort, blood pressure was measured using a Mindray sphygmomanometer with trueBP ™ technology.If the blood pressure was above 140/90 mmHg (58), the measurement was repeated.If the blood pressure was still elevated, a 30-minute automated blood pressure measurement was carried out.The most recent blood pressure was included in the analysis.As patients with PWS in the Dutch reference center were not evaluated yearly for substance abuse (smoking, alcohol), most recent data were included.However, in some cases, this was more than twelve months prior to blood and urine sample collection.

Biochemical analysis
Blood and midstream urine samples were collected for general medical screening.As the outpatient clinic was often in the afternoon and fasting can be problematic in PWS due to hyperphagia, non-fasting blood samples were used.We evaluated kidney function (urea, creatinine, eGFR using Chronic Kidney Disease Epidemiology Collaboration (CKD-epi) calculation) and lipid profiles ((non-fasting) low density lipoprotein (LDL)cholesterol, high density (HDL)-cholesterol, triglycerides, total cholesterol), glucose metabolism (fasting or random glucose, hemoglobin A1c (HbA1c)) in blood.In random spot urine samples, we measured microalbumin, urinary albumin-tocreatinine ratio (UACR), total protein and/or urinary protein-tocreatinine ratio (UPCR).As the outpatient clinic was in the afternoon in some centers, first morning urine samples were not available.If microalbuminuria was found, the patients were referred to the general practitioner to confirm microalbuminuria in a second urine sample.Second samples were not available for this study.A urine dipstick was used to evaluate the presence of glucose, protein, leukocytes, and nitrite.If urine and blood samples could not be collected on the same day, the blood samples with the shortest time to urinalysis were used.If urine and blood samples were collected more than 12 months apart, the blood sample was excluded from analysis.

Cut-off levels
DM2 was defined as a fasting glucose of 7.0 mmol/L or higher (if available) or a repeated non-fasting glucose above 11.1 mmol/L.As fasting glucose were only available in a few patients, non-fasting glucoses were used in further analysis.Impaired glucose tolerance was tested using an oral glucose tolerance test and diagnosed if measured serum glucoses was between 7.8 to 11.1 mmol/L after two hours according to the American Diabetes Association classification (59).HbA1c was not used as a diagnostic criterium for DM2, but was used to monitor glycemic control.Furthermore, patients with a previous diagnosis of DM2 were considered as having DM2.Hypercholesterolemia was defined as a non-fasting LDLcholesterol above 4.1 mmol/L and dyslipidemia was defined as a non-fasting LDL-cholesterol above 4.1 mmol/L or a (non-fasting or fasting) triglyceride above 2.0 mmol/L (60).The CKD-epi formula was used to calculate the estimated glomerular filtration rate (eGFR) values (61).An eGFR >90 mL/min was defined as normal kidney function.Microalbumin was considered elevated when >0.02 g/L (measured in one random spot urine sample).Microalbuminuria was defined as an UACR between 3 and 30 mg/mmol creatinine and proteinuria as an UACR >30 mg/mmol creatinine according to the Kidney Disease: Improving Global Outcomes (KDIGO) guideline (35).Proteinuria was defined as an UPCR of >20 mg/mmol creatinine (0.2 mg/mg) and nephrotic proteinuria as an UPCR of >300 mg/mmol creatinine (3.0 mg/mg) (62).Urinary total protein levels >0.15 g/L in a24 hour urine collection were considered elevated.An albumin excretion rate of 30-300mg/24 hours was considered increased and >300 mg/24 hours was considered severely increased or nephrotic proteinuria (35).When leukocytes and/or nitrite was positive in urine dipstick analysis, patients were referred to their general practitioner for urine culture and, if positive, (antibiotic) treatment.

Systematic literature review
We performed a literature search in June 2020 (updated in March 2022) using Embase, Medline, the Web of Science Core Collection, Cochrane Central Register of Controlled Trials and Google Scholar (Table S1).Studies were assessed for eligibility using the PRISMA 2020 (63) guideline and were included if they reported on both 1) (the prevalence of) kidney disease, urological abnormalities, treatment for kidney disease or markers for kidney function and 2) PWS, Prader-Willi-like syndrome (PWLS) with an abnormality in 15q11.2-13or one of the genes on the PWS critical region.We also scanned the references of included articles.Studies reporting on cases of kidney or urological cancer, combinations of PWS with other genetic syndromes, PWLS with a deletion extending beyond the PWS critical region and articles only describing the expression of a gene in the kidney were excluded.Meeting reports, workshop summaries, reviews, conference abstracts, guidelines, articles without full-text availability and articles that were not available in English were also excluded.
Screening based on title and abstract was performed independently by three reviewers (either NN and KP or DA and KP).Full-text screening was performed independently by DA and KP.Disagreements were resolved in a meeting with a third reviewer (LG).Data was extracted by DA and NN.

Data analysis
Data was analyzed using IBM SPSS version 28.0.Continuous variables are presented as median [interquartile range (IQR)], dichotomous variables as number and percentage of patients n (%).To investigate the association between microalbuminuria, macroalbuminuria and the presence of risk factors (e.g.smoking, alcohol usage, DM2, blood pressure/hypertension and hypercholesterolemia), a Chi-squared test was used for dichotomous variables and a Mann-Whitney U test for continuous variables.A p-value of <0.05 was considered statistically significant.To adjust for BMI as possible confounder, a logistic regression analysis was performed.

French cohort n=43
Total n=162 Age at urine sample, years

Urinalysis Microalbuminuria and proteinuria
Twenty-eight out of 160 patients (18%) had elevated urine albumin, microalbuminuria or proteinuria (Table 2).Two patients were excluded from this analysis as data on (micro)albuminuria, proteinuria, UACR and UPCR were missing.Of the two males and two females with impaired renal function according to alternative reference values for eGFR for adults with PWS, only one female had microalbuminuria (Case 8, Table 3) (20).
Urine albumin was elevated in 24 of 157 subjects (15%).The UACR was calculated in 75 patients, as urine creatinine was not systematically assessed in all patients, and was abnormal in 19 of 75 patients (25%), of whom 15 had microalbuminuria and four had macroalbuminuria.UPCR was available for 57 patients, of whom ten (18%) had proteinuria and one nephrotic proteinuria (i.e.460 mg/mmol).
We did not find any association between elevated urine microalbumin or (micro)albuminuria and age, gender, genotype, GH treatment, alcohol consumption or smoking (Table 2).Three of 28 patients with normal BMI, three of 37 who were overweight and 22 of 95 with obesity had elevated urine microalbumin or (micro) albuminuria.Patients with elevated urine microalbumin or (micro) albuminuria had significantly higher BMI (38.7 3.In the Dutch cohort, six out of nine (67%) patients with elevated UACR or UPCR also had leukocyturia and in two subjects, nitrite was also positive.Patients with both leukocyturia and nitrite in the urine sample were advised to have their urine checked for urinary tract infection at the general practitioner.Two patients had transient microalbuminuria or proteinuria (Table 3), one of them had leukocytosis at dipstick analysis.Neither had hypertension.Two other Dutch patients had microalbuminuria/proteinuria that persisted after repeating the urine sample.In one case (case 6) a 24-hour urine sample was collected that showed an elevated total protein of 0.56g/24h).A 30-minute blood pressure measurement at home was carried out, which showed systolic and diastolic hypertension.During analysis of hypertension, psychosocial stress turned out to be a strong contributing factor.This patient was referred to our CKD stage (35) G?A2 a G?A2 a G1A1 G1A2 G1A2 G1A3 G1A2 G1A2 G1A2 neuropsychologist for psychosocial support and stress management.
In addition, an angiotensin-converting enzyme (ACE) inhibitor was prescribed.

Leukocyturia in the Dutch cohort
Urine dipstick in the Dutch cohort was positive for leukocytes in 18 of 55 (33%) patients and nitrite in two of 54 (4%).In males, leukocyturia was seen in four of 33 patients (12%, and there was 1+ protein in two, 2+ protein in one and amount unknown in one patient).In females, 14 out of 22 (64%) had leukocyturia with 1+ protein in two, 2+ protein in four, 3+ protein in one, 4+ protein in two and amount unknown in five patients.As in the non-PWS population, leukocyturia was more prevalent in females than in males (p<0.001).

Glucosuria in the Dutch cohort
Urinary glucose was present in three out of 51 (6%) patients, all of whom had DM2 and were already on a treatment regime.Nonfasting (random) blood glucose levels of these patients ranged from 11.9 to 17.4 mmol/L and HbA1c from 62 to 70 mmol/mol, indicating suboptimal glucose management.None of them had proteinuria.

Systematic literature review
Forty-nine publications were included in the literature search (Figure 1).Multiple case reports have been published on patients with PWS and kidney disease.We found eight case reports describing proteinuria (64-71).There were three case reports in which albuminuria was reported in patients with PWS and DM2 (72-74).One article described a patient with stage 3 CKD (75).Five cases were found describing acute kidney injury (76)(77)(78)(79)(80) due to several different causes: pneumoperitoneum with abdominal compartment syndrome (76), secondary rhabdomyolysis caused by hyperpyrexia (77), nephrotoxic medication (78), hyperosmolar hyperglycemic state (79) and respiratory failure due to obesity hypoventilation syndrome (80).We found two case reports describing adults with PWS who needed hemodialysis due to end stage kidney disease, one patient from intrinsic nephrotic disease, the other case from uncontrolled diabetes mellitus (81,82).Congenital renal or urological abnormalities, such as a horseshoe kidney, bilateral non-communicating paraurethral meatus, and absence of one kidney were also reported (83)(84)(85).Multiple UTI were reported in one case report (86) and a congenital disorder (megacystic-microcolon-intestinal hypoperistalsis syndrome) in another (87).Two articles reported on potential underlying genetic components: i) the NDN gene which is on the PWS region and ii) NIPA2, a gene not imprinted in patients with PWS that influences magnesium transportation, and that possibly reflects an association between kidney homeostasis or disease and PWS (88, 89).Three studies reported on the use of specific medications, such as ACEinhibitors and (loop)diuretics, that could influence kidney function (90-92).One article mentioned the use of these drugs, without reporting on kidney function (90).In two other articles about these drugs, kidney function was normal (91,92).

Kidney disease
We found 16 studies reporting on kidney disease in patients with PWS (Table 4).Sinnema et al. (17) reported a prevalence of urinary tract problems or kidney disorders in 6% (six of 102 adults) with a confirmed diagnosis of PWS.In these patients, blood and urine analysis were performed based on a clinical suspicion of urine tract or kidney problems.Systematic screening of urine or blood samples was not performed in these patients.
In patients with PWS and diabetes, three studies reported microalbuminuria (7-56%) or proteinuria (3-11%) (48)(49)(50).In 480 patients with PWS, Pemmasani et al. (47) reported a prevalence of CKD of 6.5% (31 patients), that increased with age.Renal insufficiency was reported by Höybye et al. (93) in one out of ten patients with PWS who were treated with GHt.Koizumi et al. studied body composition in seven patients with PWS, one of whom was diagnosed with proteinuria and was taking an ACE-inhibitor (94).Van Nieuwpoort et al. reported a median [IQR] serum creatinine in 15 patients of 69.0 [10.0] µmol/L with no significant difference between men and women.According to Manzardo et al. (96) renal dysfunction was more prevalent in patients with PWS who had been diagnosed with thromboembolism than those without (6 out of 34 (17%) vs 14 out of 1013 (1.4%), p<0.001).
Multiple studies have investigated causes of death in PWS.Cohen et al. (97) described autopsy results in three children with PWS in whom urinalysis was negative for proteinuria.The kidneys had normal weight but showed glomerular enlargement and dilatation of capillaries as well as increased cellularity compared to controls, suggesting hyperfiltration.Nagai et al. (98) compared 20 patients with PWS with (age range 0.7 to 15 years) and without (never received, age range 9 months to 34 years) GHt.In the group that did not receive GHt, two patients died from renal and cardiac failure, versus none in the GHt group.Two other studies investigating causes of death in PWS both found that 2% died with or because of CKD (24, 99).

Urological disorders, diseases and congenital anomalies of kidney and urinary tract
Results of studies reporting urological disorders in PWS are shown in Table 5. Torrado et al. (104) reported that five out of 180 (3%) patients with PWS had Congenital Anomalies of Kidney and Urinary Tract (CAKUT), which was significantly higher than in the general population (0.1-0.5%) (p<0.05).Malformations found included renal hypoplasia, ureteral duplication, bifid renal pelvis, vesicoureteral reflux, pelvicalyceal dilatation and ureteral valves (104).No association between genotype and reno-ureteral malformations was found (p=1.00).Sinnema et al. (101) found CAKUT (bilateral duplication of kidney and ureter system) in one of 12 patients and Pacilli et al. ( 105) found a horseshoe kidney in one of 33 patients.
Chao et al. (34) reported that urodynamic tests were abnormal in 17 out of 34 patients with PWS, with abnormal flowmetry in 12 out of 34 and elevated residual bladder volume in 10 out of 20.In patients with lower urinary tract symptoms (LUTS), peak flow rate, voided volume, and bladder capacity were significantly decreased compared to patients without LUTS.

Discussion
Our study in 162 adults with PWS showed that one in every five (young) adults had early signs of microvascular disease, measured by elevated urine microalbumin or (micro)albuminuria, while eGFR was normal.However, as muscle mass is low in PWS (6), creatinine levels and eGFR may be falsely 'normal'.Our findings suggest that pre-symptomatic kidney injury may be missed when only blood measurements are performed.Therefore, in this patient group, urinalysis is essential for timely screening of microvascular (kidney) disease.We propose an algorithm for the detection and management of microvascular disease in adults with PWS.
Twenty-eight out of 160 (18%) patients had elevated urine albumin, (micro) albuminuria or proteinuria; eleven of whom were diagnosed with hypertension and sixteen with DM2.No age difference was found between patients with normal and elevated UACR.
In our literature review, we found an overall prevalence of microalbuminuria of 7-56% and macroalbuminuria of 3-11% in patients with PWS and diabetes (48)(49)(50).We did not find any studies on microalbuminuria or proteinuria in patients with PWS without DM2.Our study is the first to analyze the presence of (micro)albuminuria in patients with PWS with and without DM2.In the general population, around 25% of patients with longstanding (>10 years) DM2 have microalbuminuria and 5-20% proteinuria (51).Before antihypertensive treatment, patients with essential hypertension (in the general population) developed proteinuria in 35-65% and renal insufficiency in 33%, depending on severity of hypertension, DM2, medication, and age (106).In our cohort, 16 of 44 (36%) of patients with DM2 had (micro) albuminuria.Notably, the prevalence of albuminuria in our PWS patients without DM2 was similar to that in non-PWS patients with long-term DM2.The albuminuria in our PWS patients suggests microvascular damage is already present at a young age (23-33 years).
In this study, as in the normal population, albuminuria was associated with hypertension, DM2 and obesity.Obesity is Urinalysis was negative for protein in two PWS patients and not done in one.One patient died of aspiration pneumonia, one of massive pulmonary embolism and the last after infectious complications.Combined weight of the kidneys in all patients were not greater than expected.
In the patients with PWS, smooth capsular surfaces, widened cortices and absence of scarring was observed on the kidneys.The mean area of Bowman's capsule and glomerular tuft were increased in all three PWS patients compared to the controls.Glomerular enlargement was seen, as well as mild dilatation of capillaries and increased cellularity (mainly mesangial origin).No changes consisted with diabetic nephropathy were seen in all patients.
PWS not genetically confirmed.associated with chronic kidney disease and glomerulosclerosis.A possible mechanism is hyperfiltration, which leads to albuminuria, regardless of the presence or absence of diabetes (107)(108)(109)(110). Furthermore, obesity has been associated with increased progression of CKD (111).
In the general population, microalbuminuria is an indicator for microvascular injury, and a risk factor for both CKD and CVD (40- 42,53,54).Although none of our patients was formally diagnosed with CVD, the presence of microalbuminuria might indicate latent CVD.Young adults with PWS often have cardiovascular risk factors including obesity (60%), DM2 (27%), hypertension (18%), smoking (8%) and hypercholesterolemia (14%) or dyslipidemia (10%).Timely and adequate follow-up of microalbuminuria and treatment of these risk factors might decrease cardiovascular and renal comorbidities and complications.

Initial visit
During the initial visit, the physician should ask about previous UTIs and use of cigarettes and/or alcohol.In case of multiple UTIs, one should consider a kidney ultrasound to detect malformations, assess the post-void residual urine volume and referral to a urologist.In case of substance use, we advise to encourage cessation of smoking or alcohol usage.Furthermore, weight and blood pressure should be measured yearly.In case of hypertension, one should consider analysis to secondary causes such as stress, left ventricle hypertrophy or renal causes and manage both obesity and hypertension according to general guidelines.

Monitoring of eGFR
In our study, we showed that median creatinine levels were at the lower limit of the reference interval of the general (non-PWS) population in both male and female adults with PWS.This finding could be related to low muscle mass, which is decreased by 25 to 37% in patients with PWS (6), resulting in an overestimation of kidney function when measured by eGFR.The eGFR (calculated using the CKD-EPI) might therefore not be the optimal test to estimate kidney function in adults with PWS.We previously proposed alternative reference values of eGFR for adults with PWS (>98 ml/min/1.73m 2 in males and >93 ml/min/1.73m 2 in females) (20) to avoid underdiagnosis of CKD.Another method of assessing kidney injury is by measuring the excretion of LMW Angiotensin-converting enzyme (ACE), body mass index (BMI), cardiovascular disease (CVD), chronic heart failure (CHF), chronic kidney disease (CKD), computed tomography (CT), confidence interval (CI), deletion (del), diabetes mellitus (DM), dual-energy X-ray absorptiometry (DXA), female (F), glomerular filtration rate (GFR), growth hormone treatment (GHt), imprinting center defect (ICD), interquartile range (IQR), males (M), maternal uniparental disomy (mUPD), not available (NA), Prader-Willi syndrome (PWS), standard deviation (SD), urinary albumin-to-creatinine ratio (UACR), visceral adipose tissue (VAT).
van  proteins, such as cystatine-C, retinol binding protein and alpha1microglobulin.LMW proteins are filtered and reabsorbed by the kidneys.In kidney injury and (diabetic) proteinuria, an increased excretion of these proteins might be found in urine samples (44)(45)(46).Therefore, a more accurate methods of assessing kidney function in patients with PWS might include cystatin-C clearance adjusted for BMI (112,113) or 24-hour urine creatinine measurement.However, multiple factors such as thyroid dysfunction and corticosteroid use might influence cystatin-C levels (114,115).We therefore recommend yearly evaluation of creatinine and eGFR according to PWS-specific cut-off values and yearly urinalysis in all adults with PWS (see Figure 2) (20).

Laboratory measurements
We advise yearly evaluation of kidney function, glucose levels and lipids.If an impaired renal function is detected, additional investigation using a 24-hour urine and kidney ultrasound should be performed to identify the cause.In addition, a nephrologist could be consulted.A new diagnosis of diabetes or hypercholesterolemia should be managed according to non-PWS guidelines.

Detection of UTIs
Urgency symptoms (sudden need to pass urine with or without urine incontinence) and UTIs are common in adults with PWS, according to our literature review and previous study in our cohort (26-60% and 22-32%, respectively) (100,102,119).UTIs might be prevalent in adults with PWS related to obesity (120).Health care providers should be aware of the increased pain threshold and inability to mount a fever (due to disturbed temperature regulation) in patients with PWS (8,32).This combination of features might lead to an atypical presentation of UTI (e.g.only changes in behavior), causing underdiagnosis of UTIs in this vulnerable patient population.If leukocytes and/or nitrite are positive, a urine culture should be performed and the patient should be treated with antibiotics if indicated.

Urinalysis for the detection of micro-or macroalbuminuria
We recommend performing urinalysis (midstream urinalysis using urine dipstick and measurement of UACR and UPCR) yearly in all adults with PWS.First morning urine samples are preferred over random urine samples to decrease the influence of orthostatic proteinuria (35).In those diagnosed with hypertension and DM2, we advise stringent control of blood pressure and glucose levels.In addition, we recommend performing urinalysis yearly, as per guidelines for DM2 and hypertension in the general population (116,117).
In case of abnormalities, a second sample urinalysis should be performed.When microalbuminuria or proteinuria persists, 24hour urine should be collected to calculate the 24 hour creatinineand protein excretion.However, in people with of intellectual disability, collecting 24 hour urine might be challenging.In those with normal eGFR but elevated UACR or UPCR without a UTI, an renal ultrasound should be performed.
When proteinuria is confirmed, patients should be screened for underlying diseases and any risk factors should be treated.In three LUTS patients in whom videourodynamic studies were performed, all showed detrusor sphincter dyssynergia.
van Cardiovascular status should be optimized and cardiovascular risk should be assessed.Referral to (or consultation with) a nephrologist should be considered.Those with progressive proteinuria should be referred to a nephrologist.In adults with PWS with albuminuria and hypertension, an ACE inhibitor or angiotensin-II receptor antagonist should be the first line of treatment due to their renoprotective effects (121).If obesity is present, a weight reducing regime should be started.

Strengths and limitations
As all studies, our study has strengths and limitations.This study is, to our knowledge, the first cohort study to systematically assess kidney function and urinalysis in adults with PWS.However, there are several limitations.Blood and urine samples were not always collected on the same day as the blood samples [and in eight Dutch and all French patients, blood results were not included in the statistical analysis due to the prolonged time interval between blood and urine samples (>12 months)].As random spot urine samples were collected instead of first morning urine samples, the high prevalence of (micro)albuminuria might partly be caused by sample contamination, orthostatic proteinuria or physical exertion.Furthermore, microalbuminuria was not confirmed in a second sample as repeated samples were only available in four patients.In addition, as muscle mass might be low in patients with PWS, leading to a low serum creatinine, defining microalbuminuria by the UACR or UPCR might have led to an overestimation (6).Smoking and alcohol usage were not assessed yearly in some patients, and data could therefore have been out of date.Patients with leukocyturia or nitrites were first referred to their general practitioner for urine culture and UTI treatment if positive for infection, after which urinalysis was repeated.Therefore, the results of urine cultures were not available at our center.Urinalysis was not available in the entire cohort, which might have led to selection bias and a 24-hour urinalysis was only available in one patient.Additionally, patients were not routinely screened for CVD (e.g. by an ECG or cardiac ultrasound).Furthermore, this cohort consists of relatively young adults without cardiovascular events, though risk factors for CVD are present, which justifies urinalysis even in young patients with PWS.Prospective studies in a larger cohort are needed to overcome these limitations.

Conclusion
Upon screening, one in every five adults with PWS had elevated urine albumin or (micro)albuminuria which was already present at young age.All had normal eGFR according to non-PWS reference intervals.However, as muscle mass is low in PWS, normal creatinine levels and eGFR may overestimate kidney function in people with this syndrome.As kidney function may be overestimated when only measured by serum creatinine, renal problems may be missed when urinalysis is omitted.Routine screening for microalbuminuria may allow early intervention to avoid CVD and CKD.Health care providers should be aware of the increased risk for CKD and CVD in adults with PWS and should optimize treatment to reduce these risk factors.To prevent longterm complications and impaired quality of life, we provide an Recommendations for screening and the management for (micro)albuminuria and its risk factors in adults with PWS.angiotensin-converting enzyme inhibitor (ACE-I), angiotensin-II receptor antagonist (ATII) electrocardiogram (ECG), estimated glomerular filtration rate (eGFR), Left ventricular hypertrophy (LVH), minutes (min), urinary tract infection (UTI), urinary albumin-to-creatinine ratio (UACR), urinary protein-to-creatinine ratio (UPCR).a. Diagnosed by repeated measurements or 30-minute measurement.b. preferably in a first morning urine sample.c.Consider using adjusted reference values for adults with PWS (e.g.>98 ml/min/1.73m2 in males and >93 ml/min/1.73m2 in females) (20).d.Obesity in adults with PWS should be aggressively management with for example a hypocaloric diet, structured regular exercise and restriction of access to food.e.For example the 2018 ESC/ESH guidelines for the management of arterial hypertension (116).f.For example the 2019 ESC guidelines on diabetes (117) Frontiers in Endocrinology frontiersin.orgalgorithm for the screening and management of micro-albuminuria and proteinuria in adults with PWS (Figure 2).

TABLE 1
Baseline characteristics of included adults with PWS.

TABLE 2
vs 32.4 kg/m 2 , Comparison of adults with PWS and normal versus (micro)albuminuria a .Data presented for n=160 as in two Dutch patients, only urine dipstick was available.b del vs mUPD.c not available because of small numbers.d After adjusting for BMI, p=0.002.e p=0.013 after adjusting for BMI.f After adjusting for BMI, p=0.032.Bold values mean a statistically significant difference.

TABLE 4
Results of studies reporting on PWS and kidney diseases with more than one patients.

TABLE 5
Results of studies reporting on PWS and urological diseases, symptoms or congenital anomalies with more than one patients.