Journal of Clinical Medicine Article Prevalence of Vertebral Fractures and Their Prognostic Significance in the Survival in Patients with Chronic Kidney Disease Stages 3-5 Not on Dialysis Cristina Castro-Alonso 1 , Luis D’Marco 2 , Jaume Pomes 3 , Monserrat Del Amo Conill 3 , Ana Isabel García-Diez 3 , Pablo Molina 1 , María Jesús Puchades 2 , José Manuel Valdivielso 4 , Verónica Escudero 1 , Jordi Bover 5 , Juan Navarro-González 6 , Begoña Ribas 7 , Luis Manuel Pallardo 1 and José Luis Gorriz 2, * 1 Department of Nephrology, Hospital Universitario Dr. Peset, FISABIO, Universidad de Valencia, 46017 Valencia, Spain; cristina.c.med@gmail.com (C.C.-A.); molina_pab@gva.es (P.M.); veesque@gmail.com (V.E.); pallardo_lmi@gva.es (L.M.P.) 2 Department of Nephrology, Hospital Clínico Universitario, INCLIVA, Universidad de Valencia, 46010 Valencia, Spain; luisgerardodg@hotmail.com (L.D.M.); chuspuchades@gmail.com (M.J.P.) 3 Department of Radiology, Hospital Clínic, 08036 Barcelona, Spain; JPOMES@clinic.cat (J.P.); MDELAMO@clinic.cat (M.D.A.C.); AIGARCIA@clinic.cat (A.I.G.-D.) 4 Vascular and Renal Translational Research Group, IRBLleida, RedinRen-ISCIII, 25198 Lleida, Spain; valdivielso@medicina.udl.cat 5 Department of Nephrology, Fundació Puigvert, IIB Sant Pau, RedinRen, 08025 Barcelona, Spain; jbover@fundacio-puigvert.es 6 Research Unit and Nephrology Service, University Hospital Nuestra Señora de Candelaria, Santa Cruz de Tenerife and Instituto de Tecnologías Biomédicas, Universidad de La Laguna, 38010 Tenerife, GEENDIAB REDINREN, Spain; jnavgon@gobiernodecanarias.org 7 Department of Nephrology, Hospital La Paz, 28046 Madrid, Spain; bego_rivas@hotmail.com * Correspondence: jlgorriz@senefro.org  Received: 22 April 2020; Accepted: 19 May 2020; Published: 25 May 2020  Abstract: Background: The prevalence of vertebral fractures (VF) and their association with clinical risk factors and outcomes are poorly documented in chronic kidney disease (CKD) cohorts. The aim of the study was to evaluate the prevalence of VF in patients with non-dialysis dependent CKD (NDD-CKD), their value in predicting mortality and its correlation with parameters of bone mineral metabolism and vascular calcification. Materials and Methods: 612 NDD 3-5 stage CKD patients participating in the OSERCE-2 study, a prospective, multicenter, cohort study, were prospectively evaluated and categorized into two groups according to presence or absence of VF at enrollment. VF were assessed with lateral radiographs and Genant semi-quantitative method was applied. Three radiologists specialized in musculoskeletal radiology performed consensual reading of individual images obtained using a Raim DICOM Viewer and a Canon EOS 350 camera to measure with Java Image software in those who had traditional acetate X-ray. Factors related to VF were assessed by logistic regression analysis. Association between VF and death over a 3-year follow-up was assessed by Kaplan-Meier survival curves and Cox-proportional hazard models. Results: VF were detected in 110 patients (18%). Serum phosphate levels (OR 0.719, 95% CI 0.532 to 0.972, p = 0.032), ankle-brachial index < 0.9 (OR 1.694, 95% CI 1.056-2.717, p = 0.029) and treatment with bisphosphonates (OR 5.636, 95% CI 1.876-16.930, p = 0.002) were independently related to the presence of VF. After a median follow-up of 35 months (IQR: 17-37 months), 62 patients (10%) died. The causes of death were cardiovascular (n = 21, 34%) and infectious (n = 11, 18%). In the crude analysis, fractured patients group had poorer survival (log-rank test, p = 0.02). After multivariate adjustment for age, MDRD, albumin, diabetes mellitus, comorbidity, Adragao Score > 3 and serum phosphate, the presence of VF (HR 1.983, 95% CI 1.009-3.898, p = 0.047) were an independent predictor of all-cause mortality. Conclusions: In our study 18% of patients with NDD-CKD have VF. Factors associated with VF J. Clin. Med. 2020, 9, 1604; doi:10.3390/jcm9051604 www.mdpi.com/journal/jcm J. Clin. Med. 2020, 9, 1604 2 of 11 were age, low serum phosphate levels and peripheral vascular disease. The presence of VF was an independent risk factor for mortality in stages 3-5 NDD-CKD patients. Clinical trials are needed to confirm whether this relationship is causal and reversible with treatment for osteoporosis. Keywords: chronic kidney disease; fractures; vascular calcification 1. Introduction Fractures are an important and debilitating condition, mainly in older people and particularly in women. Although osteoporosis is the main risk factor for fractures in the general population, renal osteodystrophy and “uremic” osteoporosis associated with a reduced glomerular filtration rate (GFR) have also been implicated [1]. Importantly, in patients with chronic kidney disease (CKD), mineral and bone disorder (CKD-MBD) related to abnormalities of calcium and phosphate metabolism and bone formation/turnover dysregulation (renal osteodystrophy) are also associated with an increased risk. Thus, CKD patients are at increased risk of hip and vertebral fractures (VF) [2,3]. However, hip fracture is the most common type among end stage renal disease (ESRD) patients; although other studies have found a high incidence of this type of fracture among patients with progressive CKD [4–6]. The epidemiological data varies among stages of CKD with diverse and even controversial results regarding incidence, prevalence and risk factors, at least partially due to different screening policies [7,8]. Nevertheless, patients on dialysis have worse prognosis and they also experience fractures at a younger age with higher mortality rates following a fracture compared to those patients without CKD [9]. To date, a few studies have examined the risk of fracture in patients with non-dialysis dependent (NDD) CKD [5,10,11]. Whereas hip fractures are rarely undiagnosed, many VFs are asymptomatic or cause mild pain, being not diagnosed until the fracture is associated with severe pain [12]. The prevalence of asymptomatic VFs among CKD patients is estimated between 2-28% [7]. The association of VFs with clinical risk factors and worse outcomes are also poorly documented in CKD cohorts. Age, sex and the presence of diabetes have been described as common risk factors for new fractures [3,13,14]. Besides this, the role of bone and mineral disorders, the metabolism of 25-hydroxyvitamin D and PTH levels have been of recent interest and still scarcely explored in this population [14–17]. In the recent literature the impact of vascular disease on bone pathology has been emphasized. Recent data from general population and CKD patients has shown an inverse relationship between cardiovascular morbidity and mortality, vascular calcification and bone mineral density the pathogenic mechanisms and factors involved remain to be clarified. Additionally, it is increasingly acknowledged the narrow relationship between bone remodeling and vascular calcification (the so-called bone-vascular cross-talk) both in CKD patients and the general population [18,19]. Moreover, few studies have demonstrated an association between bone fractures and an increased morbidity and mortality among dialysis patients [20–22]. However, these possible associations remain to be further explored in NDD-CKD patients. In this context, this study aims to explore the prevalence of VF in patients with stages 3-5 NDD-CKD, the possible impact on survival, and investigate the relationship between parameters of renal disease, bone-mineral disorder and vascular calcification with the presence of VF. For this purpose, a post-hoc analysis of the OSERCE 2 study data was performed [23]. 2. Material and Methods This study of mineral and bone disorders in CKD describes the results of a post-hoc analysis of the observational prospective Estudio sobre las alteraciones del metabolismo OSeo-mineral en la Enfermedad Renal Cronica en España (OSERCE) 2 study carried out from May 2009 to May 2012 [23]. Thirty-nine hospitals participated in this 3-year follow-up multicenter investigation aimed to assess the prevalence of vascular calcification, its correlation with parameters of bone mineral metabolism, J. Clin. Med. 2020, 9, 1604 3 of 11 J. Clin.its and Med. 2020,on effect 9, mortality, x FOR PEERhospitalization, REVIEW and progression of renal failure after a follow-up period 3 of of 11 3 years in patients with NDD-CKD. OSERCE 2 study received the approval of the Institutional Review of 3 years Board in patients of Hospital with NDD-CKD. Universitario Dr. Peset OSERCE (Valencia)2andstudy receivedsigned all patients the approval of theconsent the informed Institutional forms for their inclusion, consistent with the principles of the Declaration of Helsinki. The study informed Review Board of Hospital Universitario Dr. Peset (Valencia) and all patients signed the included consent 742 CKDforms for with patients theiragedinclusion, ≥18 yearsconsistent with thestages in NDD-CKD principles of the Declaration 3–5 (estimated GFR (eGFR) of Helsinki. < 60 mL/min The study per 1.73included m2 ). The742 CKD patients exclusion with acute criteria were aged renal ≥18 years failure, inlife NDD-CKD expectancy stages 3–5 (estimated <12 months, and hospitalGFR (eGFR) < 60 mL/min per 1.73 m 2 ). The exclusion criteria were acute renal failure, admission during the month before inclusion. Researchers recruited consecutively the first 20 patients life expectancy <12 months, in and hospital admission during the month before inclusion. Researchers recruited each center. consecutively Lateral fullthespine first 20 patients X-ray in each center. was available in 612 patients representing the final sample of this sub Lateral full spine X-ray analysis. A semiquantitative approximationwas available in 612 patientswas of Genant representing the final sample used to determine VF (Figureof this sub 1) [24]. analysis. A of Assessment semiquantitative the images wereapproximation of Genant by independently performed was used three to determine experts VF (Figure in musculoskeletal 1) [24]. radiology Assessment of the images were independently performed by three experts without access to clinical data of the patients. Consensus was sought in case of inconsistency in the in musculoskeletal radiologyorwithout readings access tocaused doubt, generally clinicalbydata of theanomalies. transition patients. Thus,Consensus in the was sought 355 (58%) in case patients withof inconsistency digital images,inmeasurement the readings was or doubt, generally quantified usingcaused a RaimbyDICOMtransition anomalies. Viewer imaging Thus, systemin the and355 in (58%) patients with digital images, measurement was quantified using the others 257 (42%) with traditional acetate X-ray was convert to digital image using a Canon EOS a Raim DICOM Viewer imaging 50 camera system and in the and measure others with Java257 (42%) Image with traditional software acetate X-ray wasAconvert (http://imagej.nih.gov/ij/). VF wastodescribed digital image as a using a Canon EOS 350 camera and measure with Java Image software (http://imagej.nih.gov/ij/). change in the shape and size of the vertebral body, with a reduction in height comparing with expected A VF was described as a change in the shape and size of the vertebral body, with vertebral height or that of the adjacent vertebra, considering fracture criteria a height loss of ≥20%. a reduction in height comparing The Genantwith method expected allowsvertebral height orthe one to determine that of theofadjacent severity VF, where vertebra, considering mild deformity fracture is classified criteria a height loss of ≥20%. The Genant method allows one to determine as a 20–25% height loss (anterior, middle, and/or posterior relative to adjacent vertebrae); moderatethe severity of VF, where mild deformity is classified as a 20–25% height loss (anterior, middle, a 25–40% height loss, and severe > 40% height loss) and differentiating VF from other, nonfractureand/or posterior relative to adjacent vertebrae); deformities (Figure 1) moderate [24]. a 25–40% height loss, and severe > 40% height loss) and differentiating VF from other, nonfracture deformities (Figure 1) [24]. Figure 1. (A). Lateral thoracic radiograph showing (green lines) vertebral fractures in T-8 and T-10 Figure 1. (A). Lateral thoracic radiograph showing (green lines) vertebral fractures in T-8 and T-10 with >20% of change in shape, size and height of vertebral body. (B). Normal vertebrae (L2) without with >20% of change in shape, size and height of vertebral body. (B). Normal vertebrae (L2) without pathological changes (green lines). T, thoracic vertebrae; L, lumbar vertebra. pathological changes (green lines). T, thoracic vertebrae; L, lumbar vertebra. Baseline visit included routine clinical anamnesis, details on current medication, ankle-brachial Baseline visit included routine clinical anamnesis, details on current medication, ankle-brachial index (ABI) measurement, centralized laboratory analysis (serum calcium, phosphate, intact PTH index (ABI) measurement, centralized laboratory analysis (serum calcium, phosphate, intact PTH (chemoluminiscence; Immulite 2000), calcidiol and calcitriol levels (radioimmunoassay), high sensitivity (chemoluminiscence; Immulite 2000), calcidiol and calcitriol levels (radioimmunoassay), high C-reactive protein (hsCRP), and creatinine, among other routine parameters). Abdomen, pelvis and hands sensitivity C-reactive protein (hsCRP), and creatinine, among other routine parameters). Abdomen, X rays were performed to calculate the Adragao and Kauppila scores for vascular calcification [25,26]. pelvis and hands X rays were performed to calculate the Adragao and Kauppila scores for vascular The progression of kidney function was monitored with creatinine levels and eGFR determined at baseline calcification [25,26]. The progression of kidney function was monitored with creatinine levels and and at months 12, 24, and 36, respectively. The eGFR was calculated using the Modification of Diet in eGFR determined at baseline and at months 12, 24, and 36, respectively. The eGFR was calculated Renal Disease (MDRD) formula [27]. Death event and its cause were recorded as main outcome. using the Modification of Diet in Renal Disease (MDRD) formula [27]. Death event and its cause were recorded as main outcome. J. Clin. Med. 2020, 9, 1604 4 of 11 3. Statistical Analysis This sub-analysis included all patients with available full spine X-ray in the OSERCE 2 study. Therefore, sample size depended on the previous recruitment of OSERCE 2 and was determined by the availability of full spine X-ray to obtain the primary variable [23]. The results of the continuous variables are expressed as the mean ± SD or median (interquartile range [IQR]) as appropriate. Given that Kauppila and Adragao scores did not present a normal distribution, both vascular calcification scores were grouped together in dichotomous variables according to the presence or absence of prominent vascular calcifications (AS > 3 or KS > 6 as previously reported) [23]. The univariate analysis was undertaken using the Student t-test, Mann–Whitney U, or chi-squared test as appropriate. Factors independently related to presence of VF were assessed by logistic regression analysis, including those variables that were significant in a univariate analysis. Overall survival was estimated using the Kaplan-Meier method, and the log-rank test was used to compare the survival curve. The univariate and multivariate association between VF and death was assessed by Cox-proportional hazard models. For the multivariate model building, we first included the presence of VF and those variables related to VF and to mortality according to the current literature (age, smoking and phosphate levels), and therefore these four variables were included in all the final models. Then we generated different models by entering the other predictors significantly associated with death in the univariate analysis (comorbidity, diabetes mellitus, overweight, diastolic blood pressure, eGFR, and serum 25 (OH) Vitamin D, 1,25 (OH) vitamin D, albumin and hemoglobin levels). Analyses were performed using SPSS statistical package version 15 (IBM, Armonk, NY, USA), and a p-value < 0.05 was considered significant. 4. Results Demographic and clinical characteristics are shown in Table 1. VFs were observed in 18% (n = 110) of the patients, with a total of 206 VFs, being the eleventh and twelfth thoracic vertebra (T11-12) and the first lumbar vertebra (L1) the most affected sites, corresponding to 62% of total VF (Table 2). Thus the median of VFs was 1 (1-4). Patients with VF were significantly older and all of them presented arterial hypertension (Table 1). Prevalence of VF was analyzed according to age and categorized in 4 ranges: VF was present in 24.5 % of patients aged >76 years, while prevalence was 14.7% in those aged <60 years, 17.5% in 60-68 years, and 15.3% in the group aged 69-76 years. This difference did not reach statistical significance (ANOVA, p = 0.091). The distribution of VF according CKD stages was, 46 (41.8%) patients in stage 3, 50 (47%) patients in stage 4 and 14 (12.7%) patients in stage 5 NDD CKD. Regarding the severity of the VF (22), patients were discriminate in terms of number of fractures as mild in 73 (46.8%) patients; moderate in 70 (44.9%) patients and 28 (17.9%) with severe compromise. Among biochemical variables, only low serum levels of phosphate and high levels of hsCRP were significant in patients with VF, who in turn tended to achieve in less proportion the normal ranges than patients without VF. Vascular calcification assessed by Adragao Score (hands only) tended to be more prevalent in patients with VF, although this difference did not reach significance. Treatment with bisphosphonates was more prevalent in the VF group. In the univariate analysis, the group with VF was significantly older and presented lower serum phosphate levels, higher levels of hsCRP (p = 0.059), higher prevalence of ABI < 0.9, vascular calcification according Adragao (hands only) (p = 0.051) and lower compliance with K/DOQI recommendations. Binary logistic regression indicated that phosphate levels, ABI < 0.9, and treatment with bisphosphonates were independent factors associated with the presence of VF (Table 3). After a median follow-up of 35 months (IQR: 17-37 months), sixty-two patients (10%) died, being the cardiovascular disease (n = 21; 34%) and infectious (n = 11; 18%) the leading causes of death Table 4). In the crude analysis, the group of fractured patients had poorer survival (log-rank test, p = 0.02; Figure 2). The VF severity was not associated with higher mortality (log-rank test, p = 0.392). After multivariate adjustment for age, eGFR, serum albumin, diabetes mellitus, comorbidity, Adragao J. Clin. Med. 2020, 9, 1604 5 of 11 Score > 3 and serum phosphate, the presence of VF (HR 1.983, 95% CI 1.009-3.898, p = 0.047) was and independent predictor of all-cause mortality (Table 5). Table 1. Demographic and clinical characteristics of OSERCE 2 patients included. Vertebral Fractures No Vertebral Fractures Characteristics Overall Sample (n = 612) p-Value (n = 110, 18%) (n = 502, 82%) Age (y) 66.1 ± 12.7 68.0 ± 11.9 65.6 ± 12.9 0.043 Sex (male, %) 65.0% 60.0% 66.1% 0.133 BMI ≥ 30 (obese) 34.6% 36.4% 34.3% 0.722 BMI (kg/m2 ) 28.8 ± 5.1 28.7 ± 5.2 28.8 ± 5.1 0.974 Smoking 12.7% 9.4% 13.5% 0.425 Diabetes mellitus (%) 37.1% 33.6% 37.9% 0.232 Arterial hypertension (%) 94.4% 100.0% 93.6% 0.006 Mean bilateral ABI 1.0 ± 0.2 0.97 ± 0.2 1.0 ± 0.2 0.059 eGFR (MDRD mL/min/1.73 m2 ) 27.3 ± 11.6 27.5 ± 11.1 27.2 ± 11.7 0.823 Serum calcium (mg/dL) 9.8 ± 0.8 9.7 ± 0.8 9.6 ± 0.8 0.788 Serum phosphate (mg/dL) 3.5 ± 0.8 3.3 ± 0.8 3.5 ± 0.9 0.028 Serum iPTH(pg/mL) a 61 (20-123) 57 (17–107) 62 (80-128) 0.437 K/DOQI Ca-target (%) 33.3% 30.2% 34.0% 0.449 K/DOQI P-target (%) 76.9% 78.3% 76.3% 0.708 K/DOQI PTH-target (%) 26.4% 24.5% 26.8% 0.628 All 3 K/DOQI crit. (%) 7.0% 2.8% 7.9% 0.065 Serum bicarbonate (mmol/L) 24.7 ± 3.5 24.4 ± 3.8 24.8 ± 3.4 0.440 C-reactive protein (mg/L) a 2.0 (2.0–6.7) 3.8 (2.0-7.7) 2.0 (2.0-6.6) 0.012 Serum albumin (g/dL) 4.0 ± 0.5 3.9 ± 0.8 4.0 ± 0.5 0.243 Serum calcidiol (ng/mL) 20.5 ± 8.6 19.6 ± 7.6 20.7 ± 8.8 0.246 Serum calcitriol (pg/mL) 39.1 ± 10.4 39.7 ± 9.0 39.0 ± 10.6 0.498 Ca-based phosphate binder (%) 18.2% 14.7% 19.0% 0.180 Bisphosphonates (%) 3.0% 7.3% 2.0% 0.003 Adragao Score ≥ 3 (%) 30.9% 29.6% 31.2% 0.752 Adragao (hands only) >1 (%) 24.5% 31.4% 23.0% 0.075 Kauppila score > 6 (%) 31.4% 26.0% 32.6% 0.199 a Skewed values are presented as median with inter-quartile range. Abbreviations: BMI, body mass index; ABI, ankle-brachial index; eGFR, estimated glomerular filtration rate; iPTH, intact parathyroid hormone; K/DOQI crit., complies with criterion/a. Bold numbers highlight significant p-values or those close to significance. If not indicated otherwise, results are presented as mean ± SD, or number (percent). Table 2. Location of vertebral fractures. Vertebrae Vertebral Fractures (n,%) T4 0 (0.0%) T5 2 (1.0%) T6 3 (1.5%) T7 7 (3.5%) T8 8 (4.0%) T9 9 (4.5%) T10 11 (5.5%) T11 34 (17.0%) T12 47 (23.0%) L1 45 (22.0%) L2 17 (8.0%) L3 9 (4.0%) L4 5 (2.0%) L5 9 (4.0%) Total vertebral fractures T: 121 L: 85 (n = 206, 100%) After a median follow-up of 35 months (IQR: 17‒37 months), sixty-two patients (10%) died, being the cardiovascular disease (n = 21; 34%) and infectious (n = 11; 18%) the leading causes of death Table 4). In the crude analysis, the group of fractured patients had poorer survival (log-rank test, p = 0.02; Figure 2). The VF severity was not associated with higher mortality (log-rank test, p = 0.392). After J. Clin. Med. 2020, 9, 1604 6 of 11 multivariate adjustment for age, eGFR, serum albumin, diabetes mellitus, comorbidity, Adragao Score > 3 and serum phosphate, the presence of VF (HR 1.983, 95% CI 1.009‒3.898, p = 0.047) was and independent Tablepredictor of all-cause 3. Multivariate analysismortality for factors(Table 5). with the presence of vertebral fractures. associated Factor HR 95% CI p-Value Table 4. Death events and causes in patients with or without vertebral fracture. Lower phosphate levels Events 0.719 0.532−0.972 Fracture n = 110 No Fracture n = 502 0.032 p-Value in blood Total Ankle-brachial index 21 (19%) 41 (8%) 0.05 1.694 1.056−2.717 0.029 <0.9 disease Cardiovascular 7 (33%) 14 (34%) 0.38 Treatment Infectious with disease 3 8 5.636 1.876−16.930 0.002 bisphosphonates Malignancy 1 6 Sudden death The model included 2 to vertebral fracture in1 the univariate analysis: age, the variables found significantly related C-reactive protein, use of anticoagulants or antihypertensive Other 1 drugs, phosphoremia, 7 ankle-brachial index < 0.9, Adragao Score (hands only) > 1, compliance of the three K/DOQI criteria, and treatment with bisphosphonates. Unknown HR, hazard ratio; CI, confidence interval. 7 5 There were 62 deaths (10%). Table 4. Death events and causes in patients with or without vertebral fracture. Table 5. Multivariate Cox-proportional hazard model for risk factors of mortality in CKD patients not Events Fracture n = 110 No Fracture n = 502 p-Value on dialysis. Total 21 (19%) 41 (8%) 0.05 Factor Cardiovascular disease 7 (33%) HR 14 (34%) 95% CI 0.38 p-Value Infectious diseaseAge 3 1.074 8 1.035‒1.114 <0.001 Malignancy 1 6 Adragao Score ≥3 2.487 1.345‒3.898 0.004 Sudden death 2 1 OtherPhosphate levels 1 1.699 7 1.175‒2.488 0.005 Presence Unknownof vertebral fracture 7 1.983 5 1.009‒3.898 0.047 Multivariate analysis after adjusting There by age, estimated were glomerular filtration rate, serum albumin, 62 deaths (10%). presence of diabetes, Adragao Score >3 and phosphate levels. Figure 2. Figure Kaplan Meier 2. Kaplan Meier survival survival depending depending on on the the presence presence of of vertebral vertebral fractures. fractures. Table 5. Multivariate Cox-proportional hazard model for risk factors of mortality in CKD patients not on dialysis. Factor HR 95% CI p-Value Age 1.074 1.035-1.114 <0.001 Adragao Score ≥3 2.487 1.345-3.898 0.004 Phosphate levels 1.699 1.175-2.488 0.005 Presence of vertebral 1.983 1.009-3.898 0.047 fracture Multivariate analysis after adjusting by age, estimated glomerular filtration rate, serum albumin, presence of diabetes, Adragao Score >3 and phosphate levels. 5. Discussion This sub-analysis of the largest prospective study that explored the clinical outcomes and implications of vascular calcification in NDD CKD patients in stages 3-5 reveals in 18% the presence J. Clin. Med. 2020, 9, 1604 7 of 11 of VF. Few studies in NDD CKD patients have focused on the prevalence of VF. We found that the prevalence of VF was similar to that reported in previous studies (19.4-17%) [2,3]. However, these studies have not described clinical outcomes related to VFs in NDD CKD patients. In our findings, the factors associated to VF included age, peripheral vascular disease and, most strikingly, low serum phosphate levels. These results provide the first evidence that a decrease in blood flow of the lower extremities (measured as a decrease in the ankle/arm index) is associated with an increased risk of fractures in CKD patients; Laroche et al. [28], observed in 17 men with peripheral occlusive arterial disease, lower bone mineral density values and more VF than matched controls. Moreover, our findings are agreed with results from a community-based study where a decrease in the ankle/arm index was associated with an increase in the annual rate of bone loss [29]. Our observation is of interest, because recently there are several publications that suggesting a link between vascular disease and bone abnormalities [30,31], supporting the hypothesis that generalized atherosclerosis also affects bone circulation with reduced intraosseous circulation that may contribute to the development of osteoporosis [32,33], suggesting a link between peripheral artery disease and decreased bone mineral density. Correspondingly with our results, previous data suggested that serum biomarkers of bone metabolism are not clinically useful to assess the risk of fractures in patients with CKD [10,34]. Other investigation proved that pro-inflammatory cytokines (Il-1β, Il-6, IL-8 and TNF-α) accelerated bone loss and correlated well with some markers of bone turnover (PTH and osteocalcin) [35]. On the contrary, a novel finding of our study is the inverse correlation between serum phosphate levels with the presence of VF. The association of reduced serum phosphate levels with VF does not seem to be mediated by low PTH (not different in VF vs no-VF patients) or by differences in active serum vitamin D levels since serum calcitriol concentration were similar in VF vs no-VF. In this regard, low protein intake may contribute to a reduction in serum phosphate, poor nutrition, low grade inflammation and deficient muscle strength, all these factors are associated with VF. Certainly, this aggravated in older patients. It should be taken into consideration that in CKD patients most reports show that both high and low phosphate concentration are associated with increased mortality. In addition, our study confirms that elevation of serum phosphate is associated with vascular calcification and an aggravation of cardiovascular disease which contributes to mortality. This is not new; it provides certainty about our phosphate data. The observation of an association between low phosphate levels and VF is not incompatible with an association between high phosphate concentrations and mortality. Accepting these facts may help us to understand the importance of phosphate in kidney disease. Moreover, in kidney post transplants population, the association between some degrees of hypophosphatemia and fractures has been reported [36]. Remarkably, our results suggested that peripheral vascular calcification assessed by Adragao Score (hands) is related to VF. These findings are in agreement with previous studies in hemodialysis patients, which showed how the presence of vascular calcification in medium-sized and muscular arteries was significantly associated with VF [37]. Additionally, BMD seems to be associated with the progression of aortic calcifications in the general population [38–40]. These findings suggest that dysregulation in bone metabolism contributes to calcium deposits in arteries and may contribute to the pathogenesis of cardiovascular disease in CKD patients. However, details on the pathogenic mechanisms are beyond the purpose of this analysis and remain to be clarified. We found that VF are closely related to mortality in NDD-CKD stages 3-5. Adjusted multivariate analysis revealed that the presence of VF nearly doubled mortality risk compared to CKD no-VF patients. The relationship between hip fractures with increased mortality and hospitalization in patients on dialysis was previously found [20], but the association of well- documented VF and mortality in NDD-CKD patients is a novel finding that warrants further research. In line with previous studies, hyperphosphatemia and vascular calcification were confirmed as independent factors for mortality [41,42]. In this regard, the fibroblast growth factor-23 (FGF23) may be involved in the pathogenesis of atherosclerosis and myocardial hypertrophy [43–45]. However, a few studies have been conducted to evaluate the effect of hypophosphatemia on cardiac function [46]. Thus, has been postulated that hypophosphatemia-induced cardiomyopathy and/or cardiac arrythmias and the J. Clin. Med. 2020, 9, 1604 8 of 11 mechanisms involved are a depletion of adenosine triphosphate (ATP) in myocardial cells and decreased 2,3-diphosphoglycerate (2,3-DPG) in erythrocytes [47,48]. Interestingly, this kind of cardiomyopathy can be potentially reversible. The OSERCE 2 study, and subsequently this sub-analysis, had among its strengths the centralization of samples analysis and the radiologic reading, combined with a long period of follow-up. Moreover, the observational design does not allow the determination of outcomes related to prevention or treatment and other factors (physical activity, nutritional status, alcohol consumption). The approach of semiquantitative radiologic entails a limitation due to observer dependence. However, we minimized this factor by means of centralized and blind reading of X-rays undertaken by three independent expert radiologists. To exclude any selection bias, we compared baseline characteristics between patients excluded and included in this post-hoc analysis, not showing differences between groups. In conclusion, our study revealed that VF are frequent and may be underdiagnosed in NDD-CKD patients. Factors associated with VF were age, low levels of phosphate and peripheral vascular disease. The presence of VF was an independent risk factor for mortality in these patients. Clinical trials are needed to confirm whether this relationship is causal and reversible with treatment for osteoporosis. Author Contributions: C.C.-A., V.E., B.R. have contributed to revision of the characteristics of the patients of the study, conducted the research and investigation process and editing the manuscript; writing original draft preparation and editing the manuscript, L.D.M.; reading and digital processing of the radiologic images, J.P., M.D.A.C. and A.I.G.-D.; statistical analysis and editing the manuscript, P.M.; reviewing and editing, M.J.P., J.N.-G., L.M.P.; laboratory validation analysis, J.M.V.; design of the study, conceptualization, reviewing and editing, J.B.; design of the study, conceptualization, methodology, reviewing and editing the manuscript, and leadership responsibility for the research activity planning and execution, J.L.G. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Acknowledgments: The following investigators participated in the study: R. Alcázar (Hospital Infanta Leonor), A. Almoguera (Hospital Reina Sofia), A. Antolín (Hospital Francesc de Borja), J. Ampuero (Hospital Gregorio Marañón), G. Barril (Hospital de La Princesa), A. Blasco (Hospital Miguel Servet), J. Bover (Fundació Puigvert), S. Cigarran (Hospital de Burela), M.D. Del Pino (Hospital Torrecardenas), M. Dotori (Hospital Universitario de Málaga), V. Escudero (Hospital U. Dr. Peset), J.E. Fernández-Nájera (Hospital Francesc de Borja), M. Galicia (Hospital Vall d’Hebron), C. García-Canton (Hospital Insular de Gran Canarias), G. García Erauskin (Hospital de Cruces), S. García-Vinuesa (Hospital Gregorio Marañón), J. Gascó (Hospital Son Llàtzer), J.M. Gil-Cunquero (Complejo Hospitalario de Jaén), T. González Álvarez (Hospital Bellvitge), J.L. Górriz (Hospital Clínico Universitario), E. Gruss (Fundación Hospital Alcorcón), M.A. Guerrero (Hospital Virgen del Rocío), J. Herrero (Hospital Clínico), M.P. Marco (Hospital Arnau de Vilanova), G. Martin-Reyes (Hospital Universitario de Málaga), A.L. Martin de Francisco (Hospital Marqués de Valdecilla), A. Martínez Castelao (Hospital Bellvitge), I. Martínez Fernández (Hospital de Galdácano), E. Morales Ruiz (Hospital Doce de Octubre), P. Molina (Hospital U. Dr. Peset), R. Mouzo (Hospital del Bierzo), M.A. Munar (Hospital Son Dureta), J. Navarro-González (H. Virgen de la Candelaria), J. Nieto (Hospital General de Ciudad Real), E. Novoa (Complejo Hospitalario de Orense), A. Osuna (H. Virgen de las Nieves), L. Orte (Hospital Ramón y Cajal), J.R. Ortiz-Vigon (Hospital Basurto), M.C. Páez (Hospital Universitario Virgen Macarena), A. Peris (Hospital Francesc de Borja), J. Pérez-Pérez (Hospital Miguel Servet), C. Piñera (Hospital Marqués de Valdecilla), M. Puerta (Hospital Infanta Leonor), B. Rivas (Hospital Universitario La Paz), R. Ruiz- Calero (Hospital Infanta Cristina), P. Sánchez-Pérez (Hospital Francesc de Borja), A. Soldevilla (Hospital La Fe), C. Solozabal (H. Virgen del Camino), S. Soriano (Hospital Infanta Sofia), N. Vega (H.U. Gran Canarias Dr. Negrin), and P. Vidau (Hospital General de Asturias). Conflicts of Interest: C.C.-A., L.D.M., J.P., D.M.A.C., A.I.G.-D., P.M., M.J.P., J.M.V., B.R., L.M.P., and V.E. declare no conflicts of interest. J.F.N.G. has served as a consultant and speaker from Abbvie, Amgen, AstraZeneca, Boehringer Ingelheim, Esteve, Genzyme, Lilly, Novartis, Servier, Shire and Vifor Fresenius Medical Care, Renal Pharma. J.B. declares advisory and/or lecture fees from Amgen, Abbvie, Sanofi-Genzyme, Shire, Vifor-Fresenius-Pharma and Sanifit. J.L.G. has served as consultant for Boëhringer-Ingelheim, Mundipharma, AstraZeneca, Novonordisk and has received speaker honoraria from Boëhringer-Ingelheim, Mundipharma, AstraZeneca, Novonordisk, Novartis and Eli Lilly. J. Clin. Med. 2020, 9, 1604 9 of 11 References 1. Malmgren, L.; McGuigan, F.; Christensson, A.; Akesson, K.E. Reduced kidney function is associated with BMD, bone loss and markers of mineral homeostasis in older women: A 10-year longitudinal study. Osteoporos. Int. 2017, 28, 3463–3473. [CrossRef] [PubMed] 2. Nickolas, T.L.; Cremers, S.; Zhang, A.; Thomas, V.; Stein, E.; Cohen, A.; Chauncey, R.; Nikkel, L.; Yin, M.T.; Liu, X.S.; et al. Discriminants of prevalent fractures in chronic kidney disease. J. Am. Soc. Nephrol. 2011, 22, 1560–1572. [CrossRef] [PubMed] 3. Jamal, S.A.; West, S.L.; Miller, P.D. Fracture risk assessment in patients with chronic kidney disease. Osteoporos. Int. 2012, 23, 1191–1198. [CrossRef] [PubMed] 4. Ensrud, K.E.; Lui, L.Y.; Taylor, B.C.; Ishani, A.; Shlipak, M.G.; Stone, K.L.; Cauley, J.A.; Jamal, S.A.; Antoniucci, D.M.; Cummings, S.R.; et al. Renal function and risk of hip and vertebral fractures in older women. Arch. Intern. Med. 2007, 167, 133–139. [CrossRef] 5. Nickolas, T.L.; McMahon, D.J.; Shane, E. Relationship between moderate to severe kidney disease and hip fracture in the United States. J. Am. Soc. Nephrol. 2006, 17, 3223–3232. [CrossRef] 6. Moe, S.M.; Nickolas, T.L. Fractures in patients with CKD: Time for action. Clin. J. Am. Soc. Nephrol. 2016, 11, 1929–1931. [CrossRef] 7. Bucur, R.C.; Panjwani, D.D.; Turner, L.; Rader, T.; West, S.L.; Jamal, S.A. Low bone mineral density and fractures in stages 3-5 CKD: An updated systematic review and meta-analysis. Osteoporos. Int. 2015, 26, 449–458. [CrossRef] 8. Goldenstein, P.T.; Jamal, S.A.; Moysés, R.M. Fractures in chronic kidney disease: Pursuing the best screening and management. Curr. Opin. Nephrol. Hypertens. 2015, 24, 317–323. [CrossRef] 9. Eknoyan, G. Cardiovascular mortality and morbidity in dialysis patients. Miner. Electrolyte Metab. 1999, 25, 100–104. [CrossRef] 10. Chen, H.; Lips, P.; Vervloet, M.G.; van Schoor, N.M.; de Jongh, R.T. Association of renal function with bone mineral density and fracture risk in the Longitudinal Aging Study Amsterdam. Osteoporos. Int. 2018, 29, 2129–2138. [CrossRef] 11. Kwon, Y.E.; Choi, H.Y.; Kim, S.; Ryu, D.R.; Oh, H.J.; ESRD Registry Committee of the Korean Society of Nephrology. Fracture risk in chronic kidney disease: A Korean population-based cohort study. Kidney Res. Clin. Pract. 2019, 38, 220–228. [CrossRef] [PubMed] 12. Fusaro, M.; D’Angelo, A.; Scalzo, G.; Gallieni, M.; Giannini, S.; Guglielmi, G. Vertebral fractures in dialysis: Endocrinological disruption of the bone-kidney axis. J. Endocrinol. Invest. 2010, 33, 347–352. [CrossRef] [PubMed] 13. Jadoul, M.; Albert, J.M.; Akiba, T.; Akizawa, T.; Arab, L.; Bragg-Gresham, J.L.; Mason, N.; Prutz, K.G.; Young, E.W.; Pisoni, R.L. Incidence and risk factors for hip or other bone fractures among hemodialysis patients in the Dialysis Outcomes and Practice Patterns Study. Kidney Int. 2006, 70, 1358–1366. [CrossRef] [PubMed] 14. Mishima, T.; Motoyama, K.; Imanishi, Y.; Hamamoto, K.; Nagata, Y.; Yamada, S.; Kuriyama, N.; Watanabe, Y.; Emoto, M.; InabaM. Decreased cortical thickness, as estimated by a newly developed ultrasound device, as a risk for vertebral fracture in type 2 diabetes mellitus patients with eGFR of less than 60 mL/min/1.73 m2 . Osteoporos. Int. 2015, 26, 229–236. [CrossRef] 15. West, S.L.; Jamal, S.A. Determination of bone architecture and strength in men and women with stage 5 chronic kidney disease. Semin. Dial. 2012, 25, 397–402. [CrossRef] 16. Yamamoto, S.; Fukagawa, M. Uremic toxicity and bone in CKD. J. Nephrol. 2017, 30, 623–627. [CrossRef] 17. Molina, P.; Carrero, J.J.; Bover, J.; Chauveau, P.; Mazzaferro, S.; Torres, P.U. European Renal Nutrition (ERN) and Chronic Kidney Disease-Mineral and Bone Disorder (CKD-MBD) Working Groups of the European Renal Association-European Dialysis Transplant Association (ERA-EDTA). Vitamin D, a modulator of musculoskeletal health in chronic kidney disease. J. Cachexia. Sarcopenia Muscle. 2017, 8, 686–701. 18. Covic, A.; Vervloet, M.; Massy, Z.A.; Torres, P.U.; Goldsmith, D.; Brandenburg, V.; Mazzaferro, S.; Evenepoel, P.; Bover, J.; Apetrii, M.; et al. Bone and mineral disorders in chronic kidney disease: Implications for cardiovascular health and ageing in the general population. Lancet Diabetes Endocrinol. 2018, 6, 319–331. [CrossRef] 19. London, G.M. Bone-vascular cross-talk. J. Nephrol. 2012, 25, 619–625. [CrossRef] J. Clin. Med. 2020, 9, 1604 10 of 11 20. Tentori, F.; McCullough, K.; Kilpatrick, R.D.; Bradbury, B.D.; Robinson, B.M.; Kerr, P.G.; Pisoni, R.L. High rates of death and hospitalization follow bone fracture among hemodialysis patients. Kidney Int. 2014, 85, 166–173. [CrossRef] 21. Collins, A.J.; Foley, R.N.; Gilbertson, D.T.; Chen, S.C. The state of chronic kidney disease, ESRD, and morbidity and mortality in the first year of dialysis. Clin. J. Am. Soc. Nephrol. 2009, 4 (Suppl. 1), S5–S11. [CrossRef] 22. Fusaro, M.; Gallieni, M.; Jamal, S.A. Fractures in chronic kidney disease: Neglected, common, and associated with sickness and death. Kidney Int. 2014, 85, 20–22. [CrossRef] [PubMed] 23. Górriz, J.L.; Molina, P.; Cerverón, M.J.; Vila, R.; Bover, J.; Nieto, J.; Barril, G.; Martínez-Castelao, A.; Fernández, E.; Escudero, V.; et al. Vascular calcification in patients with nondialysis CKD over 3 years. Clin. J. Am. Soc. Nephrol. 2015, 10, 654–666. [CrossRef] [PubMed] 24. Genant, H.K.; Wu, C.Y.; van Kuijk, C.; Nevitt, M.C. Vertebral fracture assessment using a semiquantitative technique. J. Bone Miner. Res. 1993, 8, 1137–1148. [CrossRef] [PubMed] 25. Adragao, T.; Pires, A.; Lucas, C.; Birne, R.; Magalhaes, L.; Gonçalves, M.; Negrao, A.P. A simple vascular calcification score predicts cardiovascular risk in haemodialysis patients. Nephrol. Dial. Transplant. 2004, 19, 1480–1488. [CrossRef] [PubMed] 26. Kauppila, L.I.; Polak, J.F.; Cupples, L.A.; Hannan, M.T.; Kiel, D.P.; Wilson, P.W. New indices to classify location, severity and progression of calcific lesions in the abdominal aorta: A 25-year follow-up study. Atherosclerosis 1997, 132, 245–250. [CrossRef] 27. Levey, A.S.; Coresh, J.; Greene, T.; Stevens, L.A.; Zhang, Y.L.; Hendriksen, S.; Kusek, J.W.; Van Lente, F.; Chronic Kidney Disease Epidemiology Collaboration. Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann. Intern. Med. 2006, 145, 247–254. [CrossRef] 28. Laroche, M.; Puech, J.L.; Pouillès, J.M.; Arlet, J.; Boccalon, H.; Puel, P.; Mazières, B.; Arlet, P.; Ribot, C. Lower limb arteriopathy and male osteoporosis. Rev. Rhum. Mal. Osteoartic. 1992, 59, 95–101. 29. Vogt, M.T.; Cauley, J.A.; Kuller, L.H.; Nevitt, M.C. Bone mineral density and blood flow to the lower extremities: The study of osteoporotic fractures. J. Bone Miner. Res. 1997, 12, 283–289. [CrossRef] 30. London, G.M. Soft bone-hard arteries: A link? Kidney Blood Press. Res. 2011, 34, 203–208. [CrossRef] 31. Thompson, B.; Towler, D.A. Arterial calcification and bone physiology: Role of the bone-vascular axis. Nat. Rev. Endocrinol. 2012, 8, 529–543. [CrossRef] [PubMed] 32. Bridgeman, G.; Brookes, M. Blood supply to the human femoral diaphysis in youth and senescence. J. Anat. 1996, 188, 611–621. [PubMed] 33. Laroche, M. Intraosseous circulation from physiology to disease. Joint Bone Spine. 2002, 69, 262–269. [CrossRef] 34. Prasad, B.; Ferguson, T.; Tangri, N.; Ng, C.Y.; Nickolas, T.L. Association of bone mineral density with fractures across the spectrum of chronic kidney disease: The Regina CKD-MBD Study. Can. J. Kidney Heal. Dis. 2009, 6, 205435811987053. [CrossRef] [PubMed] 35. Al-Daghri, N.M.; Yakout, S.; Al-Shehri, E.; Al-Fawaz, H.; Aljohani, N.; Al-Saleh, Y. Inflammatory and bone turnover markers in relation to PTH and vitamin D status among saudi postmenopausal women with and without osteoporosis. Int. J. Clin. Exp. Med. 2014, 7, 2812–2819. [PubMed] 36. Aleksova, J.; Wong, P.; Mulley, W.R.; Choy, K.W.; McLachlan, R.; Ebeling, P.R.; Kerr, P.G.; Milat, F. Serum phosphorus levels and fracture following renal transplantation. Clin. Endocrinol. 2017, 87, 141–148. [CrossRef] 37. Rodríguez-García, M.; Gómez-Alonso, C.; Naves-Díaz, M.; Diaz-Lopez, J.B.; Diaz-Corte, C.; Cannata-Andía, J.B.; Asturias Study Group. Vascular calcifications, vertebral fractures and mortality in haemodialysis patients. Nephrol. Dial. Transplant 2009, 24, 239–246. [CrossRef] 38. Naves, M.; Rodríguez-García, M.; Díaz-López, J.B.; Gómez-Alonso, C.; Cannata-Andía, J.B. Progression of vascular calcifications is associated with greater bone loss and increased bone fractures. Osteoporos. Int. 2008, 19, 1161–1166. [CrossRef] 39. Schulz, E.; Arfai, K.; Liu, X.; Sayre, J.; Gilsanz, V. Aortic calcification and the risk of osteoporosis and fractures. J. Clin. Endocrinol. Metab. 2004, 89, 4246–4253. [CrossRef] 40. Piscopo, G.; Morrone, L. Association between vertebral fractures and vascular calcifications. G. Ital. Nefrol. 2017, 34. J. Clin. Med. 2020, 9, 1604 11 of 11 41. Sigrist, M.K.; Taal, M.W.; Bungay, P.; McIntyre, C.W. Progressive vascular calcification over 2 years is associated with arterial stiffening and increased mortality in patients with stages 4 and 5 chronic kidney disease. Clin. J. Am. Soc. Nephrol. 2007, 2, 1241–1248. [CrossRef] [PubMed] 42. Watanabe, R.L.; Lemos, M.M.; Manfredi, S.R.; Draibe, S.A.; Canziani, M.E. Impact of cardiovascular calcification in nondialyzed patients after 24 months of follow-up. Clin. J. Am. Soc. Nephrol. 2010, 5, 189–194. [CrossRef] [PubMed] 43. Vogt, I.; Haffner, D.; Leifheit-Nestler, M. FGF23 and phosphate-cardiovascular toxins in CKD. Toxins (Basel) 2019, 11. [CrossRef] [PubMed] 44. Gutiérrez, O.M.; Wolf, M.; Taylor, E.N. Fibroblast growth factor 23, cardiovascular disease risk factors, and phosphorus intake in the health professionals follow-up study. Clin. J. Am. Soc. Nephrol. 2011, 6, 2871–2878. [CrossRef] [PubMed] 45. Khan, A.M.; Chirinos, J.A.; Litt, H.; Yang, W.; Rosas, S.E. FGF-23 and the progression of coronary arterial calcification in patients new to dialysis. Clin. J. Am. Soc. Nephrol. 2012, 7, 2017–2022. [CrossRef] [PubMed] 46. Ariyoshi, N.; Nogi, M.; Ando, A.; Watanabe, H.; Umekawa, S. Hypophosphatemia-induced Cardiomyopathy. Am. J. Med. Sci. 2016, 352, 317–323. [CrossRef] 47. Lichtman, M.A.; Miller, D.R.; Cohen, J.; Waterhouse, C. Reduced red cell glycolysis, 2,3-diphosphoglycerate and adenosine triphosphate concentration, and increased hemoglobin-oxygen affinity caused by hypophosphatemia. Ann. Intern. Med. 1971, 74, 562–568. [CrossRef] 48. Rozentryt, P.; Nowak, J.; Niedziela, J.; Hudzik, B.; Doehner, W.; Jankowska, E.A.; von Haehling, S.; Partyka, R.; Kawecka, E.; Myrda, K.; et al. Serum phosphorus level is related to degree of clinical response to up-titration of heart failure pharmacotherapy. Int. J. Cardiol. 2014, 177, 248–254. [CrossRef] © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).