Table 4.

Summary of studies using DXA scanning in children with CKD. Some notable studies that have used DXA to study BMD in children with CKD (Pubmed search strategy: All English language papers from 2000 to 2018).LS, lumbar spine; TH, total hi; WB/TB, whole body/total body; WBLH/TBLH, whole body less head/total body less head; Tx, transplantation; BMD, bone mineral density; iPTH, intact parathyroid hormone; P, phosphate; HD, haemodialysis; PD, peritoneal dialysis; Ca, ,calcium; iCa, ionised calcium; ALP, alkaline phosphatase; GC, glucocorticoid; 25OHD, 25-hydroxy vitamin D; rhGH, recombinant human growth hormone; PINP, serum type I procollagen intact amino-terminal propeptide

Authors, year	Population (n)	Age of population (years)	DXA	Key findings	Biochemical correlations	Limitations	Comments
Dialysis (haemodialysis and peritoneal dialysis) and chronic kidney disease
Pluskiewicz et al., 2002 [123]	HD and PD (30: 11 HD, 19 PD)	9–23	LS, TB	Low spine and TB BMD Z-scores (− 1.47 and − 1.53). These also correlated with each other (p < 0.0001) and with dialysis vintage (p < 0.05).	No correlation found between BMD and Ca, iPTH and P. iCa correlated with low spine BMD.	No longitudinal data Small number of participants
Pluskiewicz et al., 2003 [89]	CKD5, HD and PD (40: 15 CKD5, 9 HD, 16 PD)	7–19	LS, TB	Low spine and TB BMD Z-scores in all CKD Population Dialysis vintage correlates with low TB-BMD in dialysis population (p < 0.05)	High iPTH correlated with low TB-BMD Z-scores in pre-dialysis patients (p < 0.05).	A small number of participants	The study compared DXA with QUS also, with QUS parameters lower in CKD population
Bakr, 2004 [124]	CKD5 and HD (65: 21 CKD5, 44 HD)	3–16	LS	61.9% of pre-dialysis children had low LS BMD and 59.1% of HD patients.	LS Z-scores of the osteopenic children negatively correlated with P (p = 0.004), iPTH (p = 0.03), and ALP (p = 0.02). There was a positive correlation between LS Z-scores and 25OHD	The biochemical analysis only is done in children with low Z-scores
Pluskiewicz et al., 2005 [125]	HD and PD (18: 9 HD, 9 PD)	8–21	LS, TB	Longitudinal data over 2 years showed TB Z-score was lower at the end of the study compared to baseline (p < 0.05). Spine BMD was lower at the end of the study compared to baseline (p < 0.01) in participants without GC use, and those with (p < 0.05).	iPTH, Ca, iCa and P did not correlate with skeletal measures.	A small number of participants Comparisons are done in 2 groups; GC use and no GC use	The study compared DXA with QUS. Significant population overlaps with the 2 aforementioned studies, as this study provides the longitudinal follow-up.
Andrade et al., 2007 [126]	HD and PD (20: 6 HD, 14 PD)	4–17	LS	25% had LS Z-scores < − 2 SD, but these improved when adjusted for height. 60% of children has the low-bone turnover disease.	No correlations found between bone turnover and Ca, P, PTH or ALP	A small number of participants No comparison of DXA BMD with biochemical findings Limited mineralisation reporting on bone biopsies	LS BMD Z-scores improved when adjusted for height. BMD did not correlate with high or low bone turnover
Chronic kidney disease
van der Sluis et al., 2000 [127]	CKD3-5 (33)	3–12	LS, TB	LS BMD increased with rhGH; ∆SDS 0.72/year (p < 0.01) No change was seen with TB BMD	ALP increased in the growth hormone group significantly (p < 0.05)	The study aimed at comparing GH use vs no GH use over 2 years	The study compared 18 children with CKD receiving rhGH vs 15 who did not over 2 years.
van Dyck et al., 2001 [128]	CKD 4-5 (10)	2–8	LS, TB	LS and TB BMD Z-scores increased after 1 year of rhGH treatment (p < 0.01 and p < 0.05)	After 1 year of rhGH, there was a significant rise in ALP from 308 μ/L (124 ± 621) to 720 μ/L (226 ± 1067)	The study aimed at comparing BMD before and after 1 year of rhGH treatment. Small cohort
Waller et al., 2007 [90]	CKD3-5 (64)	4–16	LS	The mean Z-score for BMD was normal (Z-score = 0.0 (95% CI − 0.29 to 0.28)). Only 8% of the patients had a BMD Z-score of less than − 2.0.	BMD Z-score did not correlate with any biochemical markers.	Only 2 participants had significantly raised PTH (> 200 pg/mL).	Strict PTH and MBD control in this population shows that maintenance of normal calcium, phosphate and PTH concentrations allows for normal LS BMD and good growth.
Swolin-Eide et al., 2007 [129]	CKD2-5 (16)	4–18	TB, TH, LS	TB and TH BMD increased on average after 1 year (p < 0.01). LS Z-scores did not change significantly.	There was a correlation between iPTH and LS BMD. PINP correlated with TB (p < 0.05), LS (p < 0.01) and TH BMD (p = 0.05).	A small number of participants No healthy controls	All biochemical markers were within the normal range. Strict MBD control in this cohort may be the reason that only 44% had BMD Z-scores below zero and 38% for LS BMD. Also, the severity of CKD must be factored in; median GFR was 46 (12–74) mL/min/1.73 m2.
Swolin-Eide et al., 2009 [130]	CKD1-5 (15)	4–15	TB, TH, LS	Only 5 patients had TB Z-scores below 0 at start of the study. On average, LS, TB and TH BMD increased over the study period of 3 years.	Most patients had raised PTH levels (median 95, 23–407 ng/L).	A small number of participants Wide range of GFR, with earlier stages of CKD, included.	Median glomerular filtration rate of 48(8–94 mL/min/1.73 m2) may explain why the BMD Z-scores were good at baseline.
Griffin et al., 2012* [84]	CKD 4-5 (88)	5–21	LS, TBLH	Adjusting for lower height Z-scores in CKD population results in increased BMD Z-scores in LS and TB DXA scans.	LS BMD & TB BMC Z-scores not associated with iPTH or P levels.	No comparison to fracture events Biochemical comparison did not include calcium. Cross-sectional data only No comparison to bone histomorphometry	pQCT showed lower tibial cortical density in CKD, but higher trabecular Z-scores.
Post-renal transplant
Tsampalieros et al., 2014* [91]	Post-renal Tx (56)	5–21	LS, TBLH	Children under 13 years had a significant reduction in LS BMD over 12 months (− 0.65, − 1.16 to − 0.09), p = 0.006). Greater GC exposure correlated with greater LS and TBLH Z-score reduction. TBLH Z-scores were significantly lower in Tx recipients than controls (p = 0.02).	iPTH reduction correlated with greater LS Z-score BMD reduction.	No comparison to bone histomorphometry No comparison to fracture events	The Pearson correlations between tibia pQCT trabecular volumetric BMD and DXA LS BMD Z-scores were 0.45 (p < 0.01) and 0.36 (p = 0.02) at baseline and 12 months.

*Authors also used peripheral quantitative CT (pQCT) as a comparator