What Is Known
- Osteopenia is a frequent finding in pediatric Crohn disease.
- Prevalence of osteopenia in pediatric Crohn disease varies amongst studies.
- There lacks evidence-based guidelines as to which patients should be referred for dual-energy X-ray absorptiometry screening.
What Is New
- We identified clinical and biochemical features associated with osteopenia in pediatric Crohn disease including body mass index z-score, weight z-score, and Pediatric Crohn disease activity index.
- Osteopenia was not associated with disease location, phenotype, extent, or estimated duration of disease before diagnosis.
- Multivariate logistic regression model identified body mass index z-score as a simple clinical item to guide referrals for dual-energy X-ray absorptiometry screening.
Bone mass is accrued during the first 2 decades of life, with gradual attrition thereafter (1,2). Hence, conditions that impair or delay bone development during childhood may have profound impact throughout life (3). Studies in children with inflammatory bone diseases (IBD) have found a high prevalence of osteopenia at diagnosis, ranging at 11% to 50% for mild and 12% to 35% moderate-severe osteopenia, mainly in Crohn disease (CD) (4–8).
Early detection of osteopenia allows appropriate management during childhood when increasing bone mass is still optimal. As with growth failure, the main contributors of osteopenia in IBD are inflammatory cytokines with both an osteoclastic effect and inhibition of bone osteoblastic effect (9). Hence osteopenia is an indication for more intensive early treatment as recommended by the ECCO-ESPGHAN pediatric CD guidelines (10). Only one-third of children with CD have, however, clinically significant osteopenia at diagnosis, and thus a clinical screening tool is required to select those who should be referred to dual-energy X-ray absorptiometry (DXA) scans.
Adult data have identified several risk factors for low bone mineral density (BMD) in CD, including age, recent corticosteroid use, low body mass index (BMI), high disease activity index, perianal disease, and malabsorption syndrome (11). There are, however, very limited data identifying risk factors of osteopenia in pediatric CD (12).
We thus aimed to identify various clinical, biochemical, and endoscopic features of CD associated with osteopenia in an aim to provide a screening tool for referring children to DXA at diagnosis.
METHODS
This was a retrospective cohort of children (5–18 years of age) diagnosed with CD between 2013 and 2017 at the pediatric IBD centers of Shaare Zedek Medical Center. Children with other bone diseases or conditions unrelated to CD, which could impact bone density were excluded. Clinical, anthropological, biochemical, endoscopic, and radiological data were recorded at diagnosis from patient medical records. Disease activity score was recorded using the Pediatric Crohn disease activity index (PCDAI) (13).
During the study period, all patients were referred to whole body dual-energy x-ray absorptiometry (DXA) scans (Hologic Discovery ADR, software version 3.3.0.1) at diagnosis as part of clinical practice, with patient bone density recorded as an age-adjusted z-score. BMD results were obtained for whole-body, whole body less head, lumbar spine, left hip, and femoral neck. As per routine practice, and based on extensive data (14–20), BMD z-scores were corrected for patient height-for-age z-score (HAZ). The calculation was performed on an online calculator provided by The Children's Hospital of Philadelphia Research Institute (https://zscore.research.chop.edu/bmdCalculator.php) based on revised reference curves in children from Zemel et al (19). On the basis of guidelines of the International Society for Clinical Densitometry, we defined osteopenia as age-, sex-, and HAZ-adjusted BMD z-score <−2, based on whole body less head DXA measurement (4), with borderline osteopenia defined as z-score between −1 and −29 (9).
Statistical Analysis
Data are presented as mean ± standard deviation, or medians (interquartile range [IQR]), as appropriate for the distribution normality. First, univariate analyses were performed to associate baseline variables with clinically significant osteopenia (ie, BMD z-score < −2) using chi-square or Fisher exact test for categorical variables and Student t-test or Wilcoxon rank sum test for continuous variables. Variables that were found to be statistically associated with the presence of osteopenia, as well as clinically important variables identified from prior studies (9,12), were entered into a multivariate logistic regression model. Area under the receiver operating characteristics curve (AUROC) was determined for the evaluation of the optimal cut-off of identified variables to predict osteopenia. Primary analysis grouped normal BMD and borderline osteopenia patients together, compared with patients with osteopenia. Secondary analysis was performed on all 3 groups independently. Statistical analyses were performed using SPSS (IBM SPSS Statistics for Windows, Version 21.0. Armonk, NY) with P <0.05 taken as the significance threshold. This study was approved by the local ethics committee.
RESULTS
One hundred sixteen CD patients were included in the analysis (Table 1). Normal BMD, borderline osteopenia, and osteopenia were observed in 27 (23%), 36 (31%), and 53 (46%) children, respectively.
TABLE 1 -
Baseline characteristics of included children
|
Total (n = 116) |
Normal and borderline osteopenia∗ (n = 63; 54%) |
Osteopenia∗ (n = 53; 46%) |
P value |
| Male gender |
67 (58%) |
33 (52%) |
34 (64%) |
0.20 |
| Age at diagnosis, years |
13 ± 3.1 (range 5–18) |
12.7 ± 3.2 |
13.5 ± 3 |
0.13 |
| BMI, kg/m2
|
16.7 ± 2.6 |
17.4 ± 2.8 |
15.9 ± 2.1 |
<0.001
|
| BMI, z-score |
−0.9 (IQR −2.1 to 0) |
−0.7 (−1.7 to 0.1) |
−1.8 (−2.4 to 0.9) |
<0.001
|
| Height-for-age z-score |
−0.7 (IQR −1.5 to 0.3) |
−0.7 (−1.5 to 0.24) |
−1.3 (−2.1 to 0) |
0.22 |
| Weight-for-age z-score |
−1 (IQR −2.1 to −0.7) |
−0.9 (−1.8 to 0) |
−2 (−2.6 to 0.9) |
0.001
|
| PCDAI |
25 (IQR 17.5–37.5) |
25 (10–37.5) |
32.5 (20–47.5) |
0.009
|
| Disease location |
| L1 |
40 (34%) |
19 (30%) |
21 (40%) |
0.75 |
| L2 |
15 (13%) |
9 (14%) |
6 (11%) |
|
| L3 |
58 (50%) |
33 (52%) |
25 (47%) |
|
| L4 |
41 (35%) |
26 (41%) |
15 (28%) |
0.33 |
| Stricturing or penetrating disease |
14 (12%) |
4 (6%) |
10 (19%) |
0.13 |
| Perianal disease†
|
27 (23%) |
14 (22%) |
13 (24%) |
0.77 |
| Growth delay†
|
47 (40%) |
21 (33%) |
26 (49%) |
0.13 |
| SES-CD |
8 (IQR 3–14) |
8 (3–12) |
9 (4–15) |
0.21 |
| Disease extent on MRE/CTE, cm |
12.5 (IQR 0–21.8) |
10 (1.5–20.3) |
15 (0–30) |
0.22 |
| Blood tests (medians) |
| Hemoglobin, g/dL (n = 115) |
11.4 ± 1.7 |
11.5 ± 1.8 |
11.3 ± 1.6 |
0.42 |
| Platelets ×10‡ (n = 111) |
426 ± 117 |
420 ± 116 |
431 ± 119 |
0.60 |
| Albumin, g/dL (n = 112) |
3.6 ± 0.6 |
3.7 ± 0.5 |
3.5 ± 0.7 |
0.06 |
| CRP, mg/L (n = 113) |
9.8 (IQR 3–32) |
12 (3.3–37) |
16.6 (3.1–48.1) |
0.69 |
| ESR, mm/hour (n = 104) |
38 (IQR 20–55) |
36.5 (20.3–55) |
39 (19–53.8) |
0.93 |
| Vitamin D, pg/mL (n = 60) |
31.9 (IQR 21.1–50) |
25.2 (20.1–38.7) |
23.3 (17.4–33.1) |
0.19 |
| Calcium corrected for albumin‡
|
9.7 ± 2.2 |
9.9 ± 2.9 |
9.4 ± 1.2 |
0.34 |
| Phosphorus, mg/dL |
4.3 ± 0.8 |
4.1 ± 1 |
4.5 ± 0.6 |
0.1 |
| ASCA positivity (n = 40) |
14/20 (70%) |
7/11 (64%) |
7/9 (78%) |
0.64 |
Numbers presented as mean ± SD or median (IQR) as appropriate. ASCA = anti-Saccharomyces cerevisiae antibody; BMI = body mass index; CRP = C-reactive protein; CTE = CT enterography; ESR = erythrocyte sedimentation rate; MRE = magnetic resonance enterography; PCDAI = Pediatric Crohns Disease Activity Index; SES-CD = simple endoscopic score-CD; wPCDAI = weighted PCDAI; Disease location: L1 = ileal, L2 = colonic, L3 = ileocolonic, L4 = upper GI.
∗Osteopenia defined as age-, gender-, and HAZ-adjusted BMD z-score <-2, borderline osteopenia defined as adjusted z-score between −1 and −2.
†Growth delay and perianal disease as identified by physician-allocated PCDAI scoring.
‡Corrected calcium: Ca (mg/dL) + 0.8 × [4 – albumin (g/dL)].
Association of Clinical Factors With Osteopenia
Univariate analysis was performed comparing children with osteopenia (z-score ≤−2) with the others. Osteopenia was associated with low BMI z-score, weight z-score, and PCDAI (Table 1 and Fig. 1). Multivariate logistic regression analysis identified an association between children with osteopenia and BMI z-score (OR 2.08 [95% CI 1.45–2.96]). The AU ROC curve of BMI z-score to reflect osteopenia was fair 0.73 (95% CI 0.64–0.83; P < 0.001). As BMI z-score should be used merely as a screening tool to rule out children not requiring DXA scan, high sensitivity and NPV are desired. Different BMI z-scores were assessed for reliability and BMI z-score <−0.5 was found to be the most appropriate cut-off with 87% sensitivity, 50% specificity, NPV 82%, and PPV 59%. Using this cut-off, 77 out of 116 (66%) were classified correctly, of whom 31/63 (49%) of patients without osteopenia would be referred for DXA screening, and 7/53 (13%) of patients with osteopenia being missed.
;)
FIGURE 1: Body mass index in patients with osteopenia versus normal and borderline osteopenia.
Using a threshold of 0.3 in the multivariable logistic regression model, for whole body less head, we calculated a sensitivity of 94% and a specificity of 40%, PPV of 57%, and NPV of 89%. This model was able to classify 65 out of 116 (56%) patients correctly.
There was no association between presence of osteopenia and patient age, HAZ, disease extent, disease location, disease phenotype, perianal disease, extra-intestinal manifestations, or serum inflammatory markers at diagnosis (Table 1).
Mean BMI also predicted the severity of osteopenia between those patients with borderline osteopenia versus those with osteopenia as defined (18.6 ± 2.3 no osteopenia vs 16.4 ± 2.3 borderline vs 15.9 ± 2.1 osteopenia; P < 0.001) (Table 2). Correlation between BMI z-score and DXA z-score was fair (r = 0.502; Fig. 2).
TABLE 2 -
Clinical features associated with bone mineral density in separate categories of normal bone mineral density, borderline osteopenia, and osteopenia
|
Total (n = 116) |
Normal BMD (n = 27) |
Borderline osteopenia (n = 36) |
Osteopenia (n = 53) |
P value |
| PCDAI |
30 (17.5–40) |
25 (16–37.2) |
23.7 (17.3–39.1) |
33.7 (18–41.4) |
0.027 |
| BMI z-score |
−0.9 (−2.1 to 0) |
−0.1 (−0.9 to 0.4) |
−1.1 (−2.2–0.3) |
−1.8 (−2.4 to −0.9) |
<0.001 |
| BMI z-score <0.5 |
105 (91%) |
21 (78%) |
32 (89%) |
52 (98%) |
<0.001 |
| Weight z-score |
−1.0 (−2.1 to 0.7) |
−0.6 (−1.7 to 0.1) |
−1.1 (−1.8 to −0.2) |
−2.0 (−2.6 to −0.9) |
0.001 |
Numbers presented as median (IQR) as appropriate. BMD = bone mineral density; BMI = body mass index; PCDAI = Pediatric Crohn's disease activity index. Normal BMD—BMD z-score ≥−1; borderline osteopenia— BMD z-score <−1 and ≥−2; osteopenia—BMD z-score <−2.
FIGURE 2: Scatterplot body mass index z-score versus bone mineral density z-score.
DISCUSSION
Osteopenia is a challenging clinical scenario in CD (11,21,22), especially during childhood and young adolescence during which most bone mass accrual occurs (23,24). In the long-term, osteopenia may be associated with increased risk of fractures (3,25–27) and loss of height because of vertebral crush fractures (8). Early detection of osteopenia could prompt more intensive anti-inflammatory treatment and tight monitoring of disease outcome (28).
DXA testing of all patients is inefficient and a simple screening tool is needed to identify those patients at risk of osteopenia. Our analysis identified BMI z-score, weight z-score, and PCDAI at diagnosis as risk factors for low BMD but following multivariate analysis only BMI maintained statistical significance. No correlation was identified between BMD and disease location, phenotype, extent, or estimated duration of disease before diagnosis.
Our results agree in part with previous studies, which identified several risk factors associated with low BMD in pediatric CD, including higher PCDAI, exposure to nasogastric tube feeds, total parenteral nutrition, growth delay (9) and hypoalbuminemia (12). The association with BMI z-score was, however, not previously reported. Similar to our findings, prior data in adults demonstrate a correlation between osteopenia in CD and BMI (9,11,29) and weight z-scores at diagnosis (21,22). Additionally, the lack of association of osteopenia with disease location and duration was previously identified (12). Some adult studies identified associations not seen in our study including C-reactive protein (CRP) and perianal disease (11). Age at diagnosis was associated with low BMD in some studies (21,29), but not others (11,22), so too for gender (11,21,22,29); however, none of these parameters were associated with osteopenia in our study.
A potential explanation of these findings may lie in the appreciation that most of the identified associations relate to markers of disease severity. Perianal disease has been identified as an independent marker of disease severity (30–33), and the need for higher anti-TNF drug levels in perianal CD than in luminal CD (34–36). BMI is a proxy of inflammation-driven malnutrition, which is independently associated with more severe disease (37). Additionally, CRP (38) and hypoalbuminemia (39) directly reflects inflammatory burden. Hence the association of these features with osteopenia may reflect disease severity per se, with the associations identified in ours, and previous studies serving as proxy markers of disease severity. The differences in specific proxy markers identified between adult and pediatric studies may reflect differences in the phenotypic features, weighting, or relevance of these markers of disease severity in these different populations (40).
The Pediatric IBD Porto Group of the European Society of Pediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) (41) recommend performing DXA only in children deemed to be at higher risk for osteopenia. The only specific example given of higher risk patients are those on long-term corticosteroid therapy. Indeed, in our cohort of newly diagnosed CD, osteopenia was identified in close to half of patients amongst whom none were exposed to long-term corticosteroid therapy.
It is noteworthy that the incidence of osteopenia in our cohort is somewhat higher than previously reported data (4–8). Although the practice at our center during the period of this study was to routinely refer all newly diagnosed CD patients for DXA screening, the retrospective nature of the study did not ensure referral of all patients. As such, some more mild patients were not referred, hence providing some referral bias.
Our findings identified clinical features at diagnosis associated with osteopenia and would be well utilized when considering, which patients need be referred for DXA screening. In our cohort, children diagnosed with CD with BMI z-score >−0.5, osteopenia could be excluded with 82% NPV.
Our study has several limitations. Firstly, as a retrospective study, not all data were available, including some patients who did not have DXA calculated with same protocol. Similarly, growth delay was recorded based on physician allocation in the PCDAI score because of lack of ability to recalculate height velocity in each patient. Symptom duration before diagnosis and pubertal stage was not recorded sufficiently to incorporate into the analysis. Although pubertal stage is an important parameter, correction of BMD to height somewhat alleviates this limitation. Additionally, despite the finding that vitamin D was a significant parameter in the pediatric population, because of the absence of vitamin D levels in a large number of patients, this could not be included in our model. Furthermore, while to the best of our knowledge, this is the largest known study of this type in the pediatric population, the sample size is still somewhat limited, and our findings must be validated on an external cohort.
CONCLUSIONS
We highlight associations between various clinical features at diagnosis of CD and osteopenia. These findings may facilitate decision-making regarding appropriateness of DXA screening in children diagnosed with CD.
REFERENCES
1. Davies JH, Evans BA, Gregory JW. Bone mass acquisition in healthy children.
Arch Dis Child 2005; 90:373–378.
2. Kemper H. Skeletal development during childhood and adolescence and the effects of physical activity.
Pediatr Exerc Sci 2000; 12:198–216.
3. Chevalley T, Bonjour JP, van Rietbergen B, et al. Fractures during childhood and adolescence in healthy boys: relation with bone mass, microstructure, and strength.
J Clin Endocrinol Metab 2011; 96:3134–3142.
4. Levine A, Mishna L, Ballin A, et al. Use of quantitative ultrasound to assess osteopenia in children with Crohn disease.
J Pediatr Gastroenterol Nutr 2002; 35:169–172.
5. Boot AM, Bouquet J, Krenning EP, et al. Bone mineral density and nutritional status in children with chronic inflammatory bowel disease.
Gut 1998; 42:188–194.
6. Gokhale R, Favus MJ, Karrison T, et al. Bone mineral density assessment in children with inflammatory bowel disease.
Gastroenterology 1998; 114:902–911.
7. Herzog D, Bishop N, Glorieux F, et al. Interpretation of bone mineral density values in pediatric Crohn's disease.
Inflamm Bowel Dis 1998; 4:261–267.
8. Semeao EJ, Stallings VA, Peck SN, et al. Vertebral compression fractures in pediatric patients with Crohn's disease.
Gastroenterology 1997; 112:1710–1713.
9. Pappa H, Thayu M, Sylvester F, et al. Skeletal health of children and adolescents with inflammatory bowel disease.
J Pediatr Gastroenterol Nutr 2011; 53:11–25.
10. Ruemmele FM, Veres G, Kolho KL, et al. European Crohn's and Colitis Organisation, European Society of Pediatric Gastroenterology, Hepatology and Nutrition. Consensus guidelines of ECCO/ESPGHAN on the medical management of pediatric Crohn's disease.
J Crohns Colitis 2014; 8:1179–1207.
11. Schule S, Rossel JB, Frey D, et al. Swiss IBD cohort study. Prediction of low bone mineral density in patients with inflammatory bowel diseases.
United European Gastroenterol J 2016; 4:669–676.
12. Semeao EJ, Jawad AF, Stouffer NO, et al. Risk factors for low bone mineral density in children and young adults with Crohn's disease.
J Pediatr 1999; 135:593–600.
13. Turner D, Griffiths AM, Walters TD, et al. Mathematical weighting of the pediatric Crohn's disease activity index (PCDAI) and comparison with its other short versions.
Inflamm Bowel Dis 2012; 18:55–62.
14. Clark EM, Ness AR, Bishop NJ, et al. Association between bone mass and fractures in children: a prospective cohort study.
J Bone Miner Res 2006; 21:1489–1495.
15. Clark EM, Tobias JH, Ness AR. Association between bone density and fractures in children: a systematic review and meta-analysis.
Pediatrics 2006; 117:e291–e297.
16. Goulding A, Cannan R, Williams SM, et al. Bone mineral density in girls with forearm fractures.
J Bone Miner Res 1998; 13:143–148.
17. Flynn J, Foley S, Jones G. Can BMD assessed by DXA at age 8 predict fracture risk in boys and girls during puberty?: an eight-year prospective study.
J Bone Miner Res 2007; 22:1463–1467.
18. Kalkwarf HJ, Laor T, Bean JA. Fracture risk in children with a forearm injury is associated with volumetric bone density and cortical area (by peripheral QCT) and areal bone density (by DXA).
Osteoporos Int 2011; 22:607–616.
19. Zemel BS, Kalkwarf HJ, Gilsanz V, et al. Revised reference curves for bone mineral content and areal bone mineral density according to age and sex for black and non-black children: results of the bone mineral density in childhood study.
J Clin Endocrinol Metab 2011; 96:3160–3169.
20. Zemel BS, Leonard MB, Kelly A, et al. Height adjustment in assessing dual energy x-ray absorptiometry measurements of bone mass and density in children.
J Clin Endocrinol Metab 2010; 95:1265–1273.
21. Guz-Mark A, Rinawi F, Egotubov O, et al. Pediatric-onset inflammatory bowel disease poses risk for low bone mineral density at early adulthood.
Dig Liver Dis 2017; 49:639–642.
22. Laakso S, Valta H, Verkasalo M, et al. Compromised peak bone mass in patients with inflammatory bowel disease--a prospective study.
J Pediatr 2014; 164:1436.el–1443.el.
23. Ghersin I, Khteeb N, Katz LH, et al. Trends in the epidemiology of inflammatory bowel disease among Jewish Israeli adolescents: a population-based study.
Aliment Pharmacol Ther 2019; 49:556–563.
24. Sykora J, Pomahacova R, Kreslova M, et al. Current global trends in the incidence of pediatric-onset inflammatory bowel disease.
World J Gastroenterol 2018; 24:2741–2763.
25. Bernstein CN, Blanchard JF, Leslie W, et al. The incidence of fracture among patients with inflammatory bowel disease. A population-based cohort study.
Ann Intern Med 2000; 133:795–799.
26. Card T, West J, Hubbard R, et al. Hip fractures in patients with inflammatory bowel disease and their relationship to corticosteroid use: a population based cohort study.
Gut 2004; 53:251–255.
27. Berger C, Langsetmo L, Joseph L, et al. Association between change in BMD and fragility fracture in women and men.
J Bone Miner Res 2009; 24:361–370.
28. Pappa HM, Gordon CM, Saslowsky TM, et al. Vitamin D status in children and young adults with inflammatory bowel disease.
Pediatrics 2006; 118:1950–1961.
29. Wada Y, Hisamatsu T, Naganuma M, et al. Risk factors for decreased bone mineral density in inflammatory bowel disease: a cross-sectional study.
Clin Nutr 2015; 34:1202–1209.
30. Siegel CA, Horton H, Siegel LS, et al. A validated web-based tool to display individualised Crohn's disease predicted outcomes based on clinical, serologic and genetic variables.
Aliment Pharmacol Ther 2016; 43:262–271.
31. Dias CC, Pereira Rodrigues P, Fernandes S, et al. The risk of disabling, surgery and reoperation in Crohn's disease - a decision tree-based approach to prognosis.
PLoS One 2017; 12:e0172165.
32. Yao JY, Jiang Y, Ke J, et al. Development of a prognostic model for one-year surgery risk in Crohn's disease patients: a retrospective study.
World J Gastroenterol 2020; 26:524–534.
33. Dias CC, Rodrigues PP, Coelho R, et al. on behalf GEDII. Development and validation of risk matrices for Crohn's disease outcomes in patients who underwent early therapeutic interventions.
J Crohns Colitis 2017; 11:445–453.
34. Davidov Y, Ungar B, Bar-Yoseph H, et al. Association of induction infliximab levels with clinical response in perianal Crohn's disease.
J Crohns Colitis 2017; 11:549–555.
35. Razvi M, Lazarev M. Optimization of biologic therapy in Crohn's disease.
Expert Opin Biol Ther 2018; 18:263–272.
36. El-Matary W, Walters TD, Huynh HQ, et al. Higher postinduction infliximab serum trough levels are associated with healing of fistulizing perianal Crohn's disease in children.
Inflamm Bowel Dis 2019; 25:150–155.
37. Jansen I, Prager M, Valentini L, et al. Inflammation-driven malnutrition: a new screening tool predicts outcome in Crohn's disease.
Br J Nutr 2016; 116:1061–1067.
38. Solem CA, Loftus EV Jr, Tremaine WJ, et al. Correlation of C-reactive protein with clinical, endoscopic, histologic, and radiographic activity in inflammatory bowel disease.
Inflamm Bowel Dis 2005; 11:707–712.
39. Qin G, Tu J, Liu L, et al. Serum albumin and C-reactive protein/albumin ratio are useful biomarkers of Crohn's disease activity.
Med Sci Monit 2016; 22:4393–4400.
40. Levine A, Griffiths A, Markowitz J, et al. Pediatric modification of the Montreal classification for inflammatory bowel disease: the Paris classification.
Inflamm Bowel Dis 2011; 17:1314–1321.
41. Miele E, Shamir R, Aloi M, et al. Nutrition in pediatric inflammatory bowel disease: a position paper on behalf of the Porto Inflammatory Bowel Disease Group of the European Society of Pediatric Gastroenterology, Hepatology and Nutrition.
J Pediatr Gastroenterol Nutr 2018; 66:687–708.
