SEARCH

SEARCH BY CITATION

Keywords:

  • RETT SYNDROME;
  • MECP2;
  • BONE DENSITOMETRY;
  • CLINICAL/PEDIATRICS;
  • LOW-ENERGY FRACTURES

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Low bone mass is reported in growth-retarded patients harboring mutations in the X-linked methyl-CpG-binding protein 2 (MECP2) gene causing Rett syndrome (RTT). We present the first study addressing both bone mineral density (BMD) and bone size in RTT. Our object was to determine whether patients with RTT do have low BMD when correcting for smaller bones by examination with dual-energy X-ray absorptiometry (DXA). We compared areal BMD (aBMDspine and aBMDtotal hip) and volumetric bone mineral apparent density (vBMADspine and vBMADneck) in 61 patients and 122 matched healthy controls. Further, spine and hip aBMD and vBMAD of patients were associated with clinical risk factors of low BMD, low-energy fractures, MECP2 mutation groups, and X chromosome inactivation (XCI). Patients with RTT had reduced bone size on the order of 10% and showed lower values of spine and hip aBMD and vBMAD (p < .001) adjusted for age, pubertal status, and body mass index (BMI). aBMDspine, vBMADspine, and aBMDtotal hip were associated with low-energy fractures (p < .05). Walking was significantly associated to aBMDtotal hip and vBMADneck adjusted for age and body mass index (BMI). Further, vBMADneck was significantly associated to a diagnosis of epilepsy, antiepileptic treatment, and MECP2 mutation group, but none of the associations with vBMADneck remained clinically significant in a multiple adjusted model including age and BMI. Neither aBMDspine, vBMADspine, nor aBMDtotal hip were significantly associated with epilepsy, antiepileptic treatment, MECP2 mutation group, XCI, or vitamin D status. Low bone mass and small bones are evident in RTT, indicating an apparent low-bone-formation phenotype. © 2011 American Society for Bone and Mineral Research


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Rett syndrome (RTT) is a severe neurodevelopmental disorder with mental retardation and a broad spectrum of symptoms including diminished motor skills and locomotion, epileptic seizures, and movement disorders.1–3 For the majority of patients with RTT, the cause is a de novo mutation in the X-linked gene methyl-CpG-binding protein 2 (MECP2) located at Xq28.4 Bone modeling/remodeling also seems affected because several studies using different methods have shown low bone mass from an early age in RTT,5–13 but only three studies included a few patients over 30 years of age.9, 11, 12 Low bone mass has been reported in patients with RTT and a previous fracture,11 but since most patients are growth retarded as part of the syndrome,14 assessing bone mass poses a special problem. Dual-energy X-ray absorptiometry (DXA) is the “gold standard” for evaluating bone mineral density (BMD), but the areal measures are highly dependent on bone size15 because small bones result in lower areal bone mineral density (aBMD).16 Of six studies investigating aBMD in RTT,5, 7, 8, 11–13 only four accounted for body size.5, 8, 11, 13 Volumetric estimates of BMD have been reported previously in only one study.13

Our aim was to clarify whether Danish patients with RTT have a low BMD even when correcting for smaller bones and to investigate the effect of age by comparing with matched controls. Furthermore, we wanted to relate our findings of aBMD and volumetric bone mineral apparent density (vBMAD) with low-energy fracture occurrence,17 MECP2 mutation group, X chromosome inactivation (XCI), and known risk factors for low bone mass such as mobility, a diagnosis of epilepsy, treatment with antiepileptic drugs (AEDs), and vitamin D status.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Patients and controls

We included 61 female patients with RTT verified by MECP2 mutations and 122 healthy controls matched according to sex, age, and pubertal/menopausal status, as described previously.17 Pubertal status was categorized into three groups—a prepubertal group of Tanner stages 1 and 2, a pubertal group of Tanner stages 3 through 5, and a postpubertal group of persons older than 20 years.

Informed consent was received from all participants, including parental or guardian permission obtained for cognitively impaired individuals, according to the guidelines of the local ethics committee. The study was approved by the Danish Data Protection Agency and the local ethics committee and was carried out according to the Helsinki Declaration.

Height and weight

Height for 37 patients was measured using a stadiometer (Model 222, Seca, Hamburg, Germany). Length for the remaining 24 patients was tape measured in the supine position from the vertex of the head to the heal with the ankle positioned in 90 degrees of fixation. Height for 122 control individuals was measured using a stadiometer (Harpenden, Holtain Limited, Crymych, UK). Weight for patients was measured with an electronic wheelchair scale (Model M50010, ADE, Hamburg, Germany) and for control individuals with an electronic scale (Model 707, Seca Delta, Hamburg, Germany). Body mass index (BMI) was calculated as weight divided by height2 or length2.

To illustrate changes in weight, height/length, and BMI, patients and control individuals were divided in groups by age intervals reflecting prepuberty (6 to 9 years), puberty (10 to 15 years), postpuberty/early adulthood (16 to 29 years), and late adulthood (30 to 61 years) according to known development of spinal and femoral bone mass in healthy children and adolescents.18–20

DXA parameters

Bone area (BA), bone mineral content (BMC), and aBMD of the lumbar spine (L1–L4) and total hip were analyzed by dual-energy X-ray absorptiometry (DXA; Hologic Discovery A, Hologic, Inc., Waltham, MA, USA) equipped with Hologic software (Version 12.4:3) with references from 3 years of age. The measured coefficient of variation (CV) at the lumbar spine was 0.4% for BA, 0.5% for BMC, and 0.3% for aBMD on 404 repeated phantom scans. To optimize the number of useful scans, only the L2–L4 area was applied for areal DXA parameters.

Estimates of vBMADspine of L1–L4 and vBMADneck of the femur were calculated from DXA data according to the Kröger model assuming a cylindrical shape of the vertebra and of the femoral neck21: vBMADspine = aBMDspineL1-L4 × 4/(π × width) and vBMADneck = aBMDneck × 4/(π × width).

Width was calculated from the given BA and the height of the scanned area of the L1–L4 box or the neck box. Height was calculated by the number of scan lines of the scanned area of the L1–L4 or neck box multiplied with a constant (0.1008 cm/line). Further, the ratios of the height and width of the L1–L4 box area for patients and control individuals were calculated to examine vertebral size and possible symmetry.

We also calculated the expected analogue aBMD of the L1–L4 spine, aBMDanalogue, of control individuals assuming the same size of vertebrae of control individuals and patients. This correction relied on equal ratios of widthpatients/widthcontrols and heightpatients/heightcontrols, of the L1–L4 box area, by which we considered it reasonable to assume that the third unknown dimension of the ratio depthpatients/depthcontrols of the L1–L4 box area was equal to the two known ratios of height and width. Since the DXA scanner already provided BMD values adjusted for area (height and width), we corrected for the third dimension of depth, thereby multiplying spinal aBMD L1–L4 of control individuals by the ratio of depthpatients/depthcontrols. The aBMDanalogue of patients equals the values of aBMD L1–L4.

Fractures and risk factors

Data regarding low-energy fractures and risk factors for fractures/low bone mass (ability to walk, a diagnosis of epilepsy, and treatment with an AED) were described previously.17

Classification of mutations and XCI analysis

The MECP2 mutations of patients had been determined previously and were classified according to assumed functionality of methyl CpG-binding protein 2 (MeCP2)—mutations leading to early truncation before the nuclear localization signal (NLS), mutations leading to truncation after the NLS, C-terminal mutations, and missense mutations. C-terminal and late truncating mutations were analyzed as one group because they both occur after the NLS. We described the specific mutations previously.17

XCI studies were performed in 61 patient samples of DNA from peripheral blood using the highly polymorphic X-linked androgen receptor (AR) locus.22 The XCI pattern was calculated and defined as skewed when 90% or more of the cells preferentially inactivated the same X chromosome. The rationale for XCI studies of peripheral blood was that a potential skewed pattern might be relevant in relation to a bone phenotype, considering a common mesodermal origin of the tissues.

Blood samples

Fasting blood samples were analyzed as part of a general screening17 and to determine 25-hydroxyvitamin D (Immunoassay, DiaSorin S.p.A, Sallugia, Vercelli, Italy) status in patients and control individuals. A value for 25-hydroxyvitamin D above 50 nmol/L was accepted as normal.

Statistical analysis

To minimize the risk of a type II error, a case-control ratio of 1:2 was chosen. According to initial power calculations with an α of 0.05, a power of 85% and SD of 0.140 (average of patients and controls), only 5 patients and 10 control individuals were needed to show significant differences in aBMD.5, 23

Weight, height/length, and BMI were compared between patients and control individuals in a linear regression model adjusted for age groups and the interaction between patient-control status and age groups. Missing variables on all DXA parameters were treated in pairs owing to the matched design. If a DXA measure was not interpretable in a patient, control individuals of the patient also were excluded from further analysis regarding this particular measure.

Overall DXA data in patients and control individuals have been presented as medians and ranges, and data were compared between patients and control individuals nonparametrically by Mann-Whitney U tests because they were not normally distributed.

Furthermore, differences between patients and control individuals in areal and volumetric densities related to age were analyzed in linear regression models adjusted for actual age, pubertal status, and BMI. Further, interaction between patient-control status and the other covariates, as well as nonlinear association of age, were tested. For the adjusted analysis, goodness of fit was evaluated regarding a linear relationship between areal and volumetric densities and covariates, a normal distribution of residuals, and homogeneity of variances.

Differences in vBMADspine, aBMDspine, vBMADneck, and aBMDtotal hip in patients with or without low-energy fractures were analyzed as patient-control differences for patients with a low-energy fracture compared with the patient-control differences for patients without a low-energy fracture owing to higher age in the fracture group.17 Mann-Whitney U tests were performed because the data were not normally distributed.

Associations between vBMADspine, aBMDspine, vBMADneck, and aBMDtotal hip each and ability to walk, a diagnosis of epilepsy, treatment with an AED, MECP2 mutation groups, XCI, and vitamin D in patients were analyzed in linear regression models adjusted for actual age, BMI, and a nonlinear association of age. Further, goodness of fit was tested as described earlier. If significant, the association was tested in a multiple regression model with a strict significance level of .025 to prove clinically important associations.

Apart from this, two-sited levels of p< .05 were considered significant. Calculations were performed with commercial software (SPSS, Version 17.0, SPSS, Inc., Chicago, IL, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Table 1 shows median weight, height/length, and BMI among patients and control individuals. A match of height and weight was not possible owing to growth retardation in RTT patients. Overall, RTT patients had significantly lower age group-adjusted weight (p < .001), height/length (p < .001), and BMI (p < .001) than controls. Further, interaction between age and patient-control status showed that the difference increased with age regarding all three parameters. The ratio of median total heightpatients to heightcontrols was 0.9 for all age groups. Thirty-four percent (21 of 61) of patients and 35% (43 of 122) of control individuals were in Tanner stages 1 to 2, and 15% (9 of 61) of patients and 14% (17 of 122) of control individuals had reached Tanner stages 3 through 5.

Table 1. Clinical and Anthropometric Data of Patients and Control Individuals
 PatientsControls
  1. Note: Anthropometric numbers are reported as median [min–max] for weight, height/length, and BMI.

6 to 9 yearsn = 16n = 32
 Weight (kg)22.0 [14.9–42.8]26.0 [19.0–43.2]
 Height/length (cm)119.3 [102.0–134.8]132.7 [115.7–145.6]
 BMI (kg/m2)15.2 [11.5–23.6]15.6 [12.1–21.4]
10 to 15 yearsn = 11n = 22
 Weight (kg)30.4 [19.3–53.0]48.5 [30.5–74.6]
 Height/length (cm)141.5 [127.8–151.5]160.4 [136.0–176.0]
 BMI (kg/m2)15.9 [11.6–23.1]18.5 [14.7–28.9]
16 to 29 yearsn = 13n = 26
 Weight (kg)42.6 [21.8–56.4]59.2 [48.5–114.6]
 Height/length (cm)143.2 [128.0–159.5]167.1 [158.4–184.4]
 BMI (kg/m2)22.0 [12.3–26.6]21.2 [18.4–33.7]
30 to 61 yearsn = 21n = 42
 Weight (kg)41.6 [26.8–58.7]66.3 [52.0–96.9]
 Height/length (cm)145.4 [131.0–157.5]167.5 [155.3–179.4]
 BMI (kg/m2)18.8 [14.9–32.5]23.5 [17.0–33.5]

Owing to difficulties in positioning, involuntary movements, and metal implants, DXA data were obtained from the spine in 37 of 61 (L1–L4) and in 42 of 61 (L2–L4) patients and from the hip and femoral neck in 48 of 61 patients and compared with matched control individuals (Table 2). Table 2 shows that RTT patients overall had significantly lower values on all DXA parameters measured on spine and hip compared with control individuals. The differences were less pronounced when comparing volumetric bone mineral apparent densities, vBMADspine and vBMADneck. Both the height and width of L1–L4 vertebrae of patients were 90% of the height and width of the L1–L4 vertebrae of control individuals (Table 2), indicating that it is reasonable to assume that the depth of vertebrae also was reduced to 90%. Therefore, aBMDanalogue of control individuals was calculated by multiplying L1–L4 aBMD by the ratio 0.9. Using this correction for comparison, the aBMDanalogue values of patients still were significantly lower than those of controls (Table 2).

Table 2. DXA Parameters on Patients and Control Individuals
DXA region and parameterPatientsControlsDXA ratiopatients/controlsp Value
  1. Note: Numbers are reported as median [min-max] for all DXA parameters. Patient-control DXA parameters were compared by Mann-Whitney U tests.

Spine L2–L4n = 42n = 84  
 BAspine (cm2)32.58 [19.54–47.95]42.83 [23.52–58.89]0.76<.001
 BMCspine (g)22.68 [7.80–56.73]38.35 [11.09–72.48]0.59<.001
 aBMDspine (g/cm2)0.637 [0.389–1.183]0.905 [0.472–1.399]0.70<.001
Spine L1–L4n = 37n = 74  
 vBMADspine (g/cm3)0.228 [0.145–0.335]0.258 [0.166–0.401]0.88.013
 Box height (cm)11.4 [7.9–14.2]13.0 [8.9–15.0]0.88.001
 Box width (cm)3.6 [3.0–4.5]4.0 [3.3–5.1]0.90<.001
 aBMDanalogue (g/cm2)0.610 [0.370–1.080]0.777 [0.400–1.240]0.79.016
Hipn = 48n = 96  
 BAtotal hip (cm2)26.06 [8.88–34.33]32.88 [17.91–47.10]0.79<.001
 BMCtotal hip (g)14.04 [3.81–26.75]28.26 [9.86–47.90]0.50<.001
 aBMDtotal hip (g/cm2)0.541 [0.302–0.840]0.834 [0.550–1.331]0.65<.001
 vBMADneck (g/cm3)0.243 [0.118–0.414]0.299 [0.216–0.434]0.81<.001

Table 3 shows the mean difference for each of the areal and volumetric BMD values between patients and control individuals from the models adjusted for age, pubertal status, and BMI. No significant interaction was found between patient-control status and covariates. In general, RTT patients had significantly lower areal and volumetric BMD values through all ages. Figure 1 illustrates how vBMADneck values maintained the same level in both patients and control individuals regardless of age.

Table 3. Mean Difference for vBMADspine, aBMDspine, aBMDanalogue, vBMADneck, and aBMDtotal hip Between Patients and Control Individuals From Linear Regression Models Adjusted for Actual Age, Tanner Group, and BMI
Bone mineral densitynpatients|ncontrolsβ coefficients (95% confidence interval)p Value
vBMADspine (g/cm3)36|72−0.022 (−0.011,−0.033)<.001
aBMDspine (g/cm2)41|82−0.148 (−0.111,−0.185)<.001
aBMDanalogue (g/cm2)36|72−0.072 (−0.038,−0.106)<.001
vBMADneck (g/cm3)46|92−0.049 (−0.032,−0.067)<.001
aBMDtotal hip (g/cm2)46|92−0.266 (−0.231,−0.301)<.001
thumbnail image

Figure 1. Mean vBMADneck for patients and control individuals by age group.

Download figure to PowerPoint

A total of 12 of 61 patients had sustained low-energy fracture.17 It was not possible to obtain DXA scans on all 12 fracture and all 49 nonfracture patients, but among the patients with DXA scans, those with low-energy fractures had significantly lower aBMDspine, aBMDtotal hip, and vBMADspine values than patients without low-energy fractures (Table 4).

Table 4. Differences in vBMADspine, aBMDspine, vBMADneck, or aBMDtotal hip in Patients With and Without Low-Energy Fractures
 nfracture|nnonfracture patientsPatient-control difference in fracture patientsPatient-control difference in nonfracture patientsp Value
  1. Note: Numbers of patient-control differences in fracture and nonfracture patients are reported as median [min-max] and compared by Mann-Whitney U tests.

vBMADspine6|310.068 [0.000–0.134]0.019 [−0,582–0.143].021
aBMDspine8|340.309 [0.152–0.521]0.130 [−1,178–0.658].003
vBMADneck8|400.060 [−0.039–0.147]0.059 [−0.083–0.247].934
aBMDtotal hip8|400.345 [0.263–0.530]0.259 [0.021–0.664].022

Table 5 summarizes the mean vBMADspine, aBMDspine, vBMADneck, and aBMDtotal hip values in patients regarding ability to walk, a diagnosis of epilepsy, treatment with an AED, MECP2 mutation group, XCI, and vitamin D attained in linear regression models, with each risk factor adjusted for age and BMI. As shown, only the mean vBMADneck and aBMDtotal hip values varied significantly in patients based on ability to walk. Further, vBMADneck was significantly associated with having a diagnosis of epilepsy, being treated with an AED, or MECP2 mutation groups. Therefore, these significant associations were included in a multiple regression model with vBMADneck as outcome and age and BMI as covariates. Thus BMI was the only significant (p = .005) covariate left at the significance level of .025.

Table 5. Mean (95% Confidence Interval) of vBMADspine, aBMDspine, vBMADneck, and aBMDtotal hip in RTT Patients Based on Different Risk Factors From Linear Regression Models Adjusted for Age and BMI
Risk factor (numbers, n)vBMADspine (g/cm3)aBMDspine (g/cm2)vBMADneck (g/cm3)aBMDtotal hip (g/cm2)ntotal (%)
  • a

    p < .05.

  • b

    p < .01.

  • c

    p < .02.

  • d

    Large deletions affecting the NLS were included in the group of mutations early truncated for NLS.

  • e

    The AR locus was not informative in five patients: vBMADspine: 3, aBMDspine: 3, vBMADneck: 4, and aBMDtotal hip: 4.

  • f

    No significant difference between winter (October 16–March 31) or summer (April 1–October 15) blood sampling was found among the proportion of patients with low or normal vitamin D values (p = NS, by Pearson chi-square).

  • g

    Vitamin D analysis failed in two patients.

Ability to walk (n1|n2|n3)17|10|1019|11|1224|14|10a24|14|10b 
 1: Independently0.229 (0.215–0.243)0.694 (0.648–0.741)0.262 (0.242–0.283)0.592 (0.559–0.626)28 (46)
 2: With light help0.231 (0.231–0.249)0.685 (0.626–0.744)0.225 (0.198–0.253)0.528 (0.483–0.572)17 (28)
 3: With massive help/not0.223 (0.204–0.241)0.662 (0.603–0.722)0.225 (0.195–0.255)0.488 (0.436–0.539)16 (26)
Diagnosis of epilepsy (nyes|nno)28|933|936|12b36|12 
 Yes0.229 (0.218–0.240)0.690 (0.656–0.724)0.235 (0.219–0.251)0.541 (0.512–0.571)47 (77)
 No0.226 (0.206–0.245)0.655 (0.589–0.721)0.288 (0.260–0.316)0.583 (0.530–0.636)14 (23)
Current AED (nyes|nno)15|2220|2224|24c24|24 
 Yes0.221 (0.206–0.235)0.678 (0.633–0.723)0.230 (0.208–0.251)0.540 (0.503–0.579)33 (54)
 No0.233 (0.221–0.245)0.687 (0.645–0.730)0.267 (0.245–0.288)0.562 (0.525–0.601)28 (46)
Mutation group (n1|n2|n3)12|9|1615|10|1718|12|18c18|12|18 
 1: Early truncated for NLSd0.226 (0.210–0.248)0.665 (0.617–0.714)0.236 (0.212–0.259)0.536 (0.495–0.578)22 (36)
 2: Late truncated for NLS and0.245 (0.227–0.263)0.747 (0.682–0.800)0.285 (0.257–0.313)0.595 (0.545–0.645)17 (28)
  C-terminal deletions     
 3: Missense0.220 (0.206–0.234)0.664 (0.618–0.710)0.237 (0.213–0.260)0.538 (0.496–0.580)22 (36)
XCI (nrandom|nskewed)e27|732|736|836|8 
 Random0.227 (0.216–0.238)0.682 (0.649–0.715)0.247 (0.230–0.264)0.544 (0.516–0.573)45 (80)
 Skewed0.236 (0.215–0.257)0.705 [0.633–0.777)0.269 (0.232–0.306)0.611 (0.549–0.673)11 (20)
25-Hydroxyvitamin D (nlow|nnormal)fg15|2017|2323|2323|23 
 Low: ≤50 mmol/L0.228 (0.212–0.244)0.680 (0.626–0.734)0.236 (0.214–0.258)0.562 (0.522–0.602)27 (46)
 Normal: >50 mmol/L0.230 (0.217–0.244)0.690 (0.646–0.736)0.256 (0.233–0.278)0.543 (0.503–0.583)32 (54)

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

This is the first study to describe a generally reduced bone size on the order of 10% in patients with RTT compared with matched control individuals. This growth retardation applied for body height and both height and width of the lumbar spine (L1–L4) among all age groups and suggests that lumbar vertebrae are symmetrically reduced in RTT. Our finding of small bones in RTT patients is in agreement with the results of an earlier study of histomorphometry on bone biopsies in RTT24 showing reduced cancellous bone volume along with a low bone-formation rate. In support of our observation, a study on a MECP2 null mouse model showed growth plate abnormalities and reduced cortical and trabecular bone volume in the mice increasing over time.25

Several DXA studies in RTT5, 7, 8, 11, 12 have shown low absolute values of BMC (g) and/or aBMD (g/cm2) for age, which we also could reproduce in our study. The problem is that size-unadjusted absolute DXA values of aBMD (g/cm2) may lead to interpretation of a relatively lower bone density among RTT patients than is actually the case. Our study is the first DXA study demonstrating significantly lower values of size-adjusted parameters of BMD of the hip (vBMADneck) and spine (vBMADspine). One-third of patients were older than 30 years, which is unique compared with other RTT DXA studies.11, 12 Our study indicates that abnormal bone mineral accrual for bone size in RTT is pronounced from a young age and continues throughout life.

The constant level of vBMADneck by age in our study also have been shown previously.26, 27 According to British reference centile curves27 of femoral neck vBMAD in 6- to 18-year-olds, the control individuals in our study have values equal to the 50th centile curve, and RTT patients are well below the 5th centile curve.27 Even though the influence of height may not be fully corrected, especially on spine vBMAD measures in children,26, 28 measures of spine and neck vBMAD seem to be less confounded by height than aBMD values.15, 21, 26

DXA measures on the hips of growing children can be difficult to perform owing to variability in skeletal development and problems with reproducibility.16, 29 Despite a larger uncertainty, patients had significantly lower values on all hip DXA parameters, including size-adjusted values, indicating that the proximal femurs of patients contain less mineral than would be expected for age.

We have reported previously a higher occurrence of low-energy fractures in RTT patients.17 This is the first study to show significant associations of lower aBMDspine, aBMDtotal hip, and vBMADspine among RTT patients with low-energy fractures compared with those without low-energy fractures. Comparison of vBMADneck did not reach significance likely owing to the rather equal values of vBMADneck in the whole group of RTT patients. Fracture risk seems to be related to both the size and density of bone,30 both of which were reduced in our RTT patients. Another study has reported lower bone mass and density in fracture patients with RTT, but the fractures were not described in detail or verified.11 Yet a third study13 did not find any association. Prospective studies of the relation between bone mass and fracture risk need to be performed to fully understand the implication of low bone mass in RTT.

Mobility status and femoral bone mass in RTT patients seem to be associated because we found significantly higher values of aBMDtotal hip and vBMADneck of independently walking patients compared with less mobile patients. Unexpectedly, the significance of vBMADneck disappeared in the multiple adjusted model. In previous studies of different bone mass measures in RTT patients at bone sites other than the legs, some did not find associations with ambulatory status,6, 10, 11 whereas other did.7, 12, 13 A recent study reported a trend toward decreasing femoral neck aBMD with decreasing ambulatory level.13 Therefore, the association to some degree may be site-specific. Reduced weight bearing in RTT patients may be more detrimental for bone mass accruing in the proximal femur than in the lumbar spine, and possibly, our observation reflects lack of site-specific loading of the hip (eg, reduced ambulatory abilities). According to the theory of the mechanostat, development of bone mass, geometry, and strength is highly dependent on mechanical signals from local muscle contractions giving rise to bone strain (tissue deformation).31

The types of MECP2 mutations of our patients equal the distribution of MECP2 mutation types in other studies in RTT.13, 17 MECP2 is a transcriptional regulatory gene that undergoes XCI with cells expressing either the mutant MECP2 allele or the wild-type MECP2 allele.32 No significant associations between mutation groups and vBMADspine, aBMDspine, vBMADneck, and aBMDtotal hip were evident from our data, as in other studies.11, 12 Only one study reported a relative association of certain mutation types with levels of aBMD of the spine and femoral neck, but the numbers were small.13 In general, clear phenotype-genotype correlations have been difficult to find in RTT33, 34 but seem to emerge when specific mutations in larger groups are investigated.35, 36 Whether an apparently low-bone-formation phenotype in RTT is directly caused by MECP2 effects in bone cells is still unclear. In support of a possible tissue-specific regulation of bone modeling by MECP2, mice carrying an MECP2 deletion targeting only brain tissue developed neurologic and autonomic symptoms similar to Mecp2 null mice but apparently not other tissue-specific phenotypes.37

To our knowledge, the relation between bone mass and XCI pattern in RTT patients has not been analyzed before. Most of our patients had random XCIs in peripheral blood, as described previously,38–40 and skewed XCI was not associated with high or low values of DXA parameters in our study, acknowledging small numbers. Nor has a general association between skewed XCI in blood and neurologic symptoms in RTT been shown.39, 40

A diagnosis of epilepsy, treatment with an AED, and vitamin D levels did not explain the differences in vBMADspine, aBMDspine, vBMADneck, and aBMDtotal hip in patients in our study. Inconsistency in studies with RTT patients6, 7, 9, 11, 13 indicates that an epilepsy diagnosis and/or treatment with an AED in general may be important for individual bone health without being a major determinant of bone mass in RTT. The varying findings in RTT regarding relation to vitamin D7, 8, 10, 12 may reflect use of AED treatment, different approaches to calcium and vitamin D supplementation, and selection of patients rather than being a specific factor responsible for low bone mass in RTT.

In conclusion, RTT patients seem to accrue less bone than expected compared with healthy control individuals. This results in low bone mass, low bone density, and small bones, which seem to be associated with low-energy fracture occurrence. No association with specific MECP2 mutation groups or XCIs was shown, but MECP2 still may have a general role in regulating bone growth and bone mass in RTT, as reflected in overall reduced bone size. Further, our study indicates that mobility status appears to be important for bone modeling and maintenance of bone mass in RTT.

The pathogenesis behind an apparent low-bone-formation phenotype in RTT awaits further studies, and the impact on fracture risk needs to be studied prospectively. Improved knowledge of bone metabolism is important to assist directions for prevention and treatment in order to improve bone health in RTT.

Disclosures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

All the authors state that they have no conflicts of interest. Our study has complied with a prespecified analysis.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

We sincerely thank the Danish patients with Rett syndrome and their families and caregivers as well as all control persons and their families for participation in the project. Thanks to the Danish Association of Rett Syndrome for supporting the project. We thank Pernille Strøm and Mette Lisa Jørgensen at the Kennedy Center for handling of project logistic. We also acknowledge the work of Anne Mette Rasmussen at the Endocrinological Department, Hvidovre Hospital, for controlling the flow of examinations at Hvidovre Hospital for patients and control individuals. Further, we thank statistician Janne Petersen at the Clinical Research Center, Hvidovre Hospital, for valuable support in statistical analyses. This research was funded by Bevica Fonden, Forskningsfonden vedrørende medfødte sygdomme under Vanførefonden, Danske Banks Fond, Elsass Fonden, Fonden til Lægevidenskabens Fremme. None of these funds had influence on study design; collection, analysis, and interpretation of data; writing of the manuscript; or decision to submit to this journal.

Authors' roles: All authors have contributed to the design of the study. GR is responsible for conducting the study; data collection, analysis, and interpretation; and drafting of the manuscript. KF and HA have contributed to collection and interpretation of data. KR has performed the molecular genetics studies. KR, JBN, KBN, and JEBJ have been involved in analysis and interpretation of the data and the drafting of the manuscript. All authors have critically revised and approved the final version of the manuscript. GR takes responsibility for the integrity of the data analysis.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References