A Cross-Sectional and Longitudinal Analysis of Trabecular Bone Score in Adults With Turner Syndrome

A Cross-Sectional and Longitudinal Analysis of Trabecular Bone Score in Adults With Turner Syndrome Abstract Context Turner syndrome (TS) is associated with short stature, gonadal failure, and fractures. Spinal trabecular bone score (TBS) is a novel bone imaging modality that has not been evaluated in TS. Objective To evaluate TBS in TS and its association with bone mineral density (BMD), prevalent fracture, and risk factors. Design and Setting Longitudinal study of TS from a single tertiary hospital between 2006 and 2017. Patients or Other Participants Fifty-eight subjects with TS aged 20 to 49 years who underwent dual-energy X-ray absorptiometry (DXA). Main Outcome Measures TBS, DXA parameters, and prevalent fractures were investigated. Results Normal, partially degraded, and degraded TBSs were observed in 39 (67%), 15 (26%), and four (7%) subjects, respectively. High rates of prescribed estrogen replacement therapy (ERT) with stable TBS and BMD were observed during follow-up. TBS was positively correlated with spine and femoral neck (FN) BMD and Z-scores (all P < 0.05) and negatively correlated with age (−0.004 per year; P = 0.014) and delay in ERT initiation in women with primary amenorrhea (−0.010 per year; P < 0.001). Fractures were present in 17 (31%) subjects. Low TBS had a significantly higher area under the receiver operator curve for predicting prevalent fracture than low bone mass at either the spine or FN (P < 0.05). Subjects with no history of fracture were more likely to have a normal TBS (P = 0.023). Conclusions BMD and TBS can be preserved with early initiation and continued use of ERT. TBS may provide additional fracture risk prediction to standard DXA parameters in TS and needs to be validated in larger prospective studies. Turner syndrome (TS) affects one in 2000 live-born girls and results from complete or partial X chromosome monosomy (1). Key features of TS include gonadal failure and short stature, whereas osteoporosis and an increased risk of fractures are associated complications. Osteoporosis has historically been associated with TS since initial reports of skeletal hypomineralization on radiographs of patients with TS (2). Subsequent studies noted an increased risk of fractures in TS, the relative risk ranging from 1.16 to 2.16 and most commonly affecting the upper limb, as well as low bone mineral density (BMD) as measured by dual-energy X-ray absorptiometry (DXA) (3–5). Gonadal failure is a major contributor to fracture risk (6), as fracture risk in patients with TS and adequate estrogen replacement therapy (ERT) is similar to that in control populations (7). Exogenous growth hormone therapy (GHT) is now standard treatment in TS for short stature and may also have beneficial effects on skeletal health (8, 9). The assessment of osteoporosis and fracture risk in women with TS is challenging. Bone density is routinely quantified by DXA, which provides a two-dimensional measure of areal BMD; however, it does not assess bone microarchitecture and can underestimate BMD in individuals with short stature and smaller bone size (10). Studies that adjusted for height or bone size have demonstrated a reduction in BMD in some TS cohorts (6, 11–13) but not others (8, 14, 15). As such, BMD alone may not be an accurate predictor of skeletal fragility and fracture risk in TS, particularly when short stature is prevalent. Although bone density is highly correlated with bone strength, it explains only 60% to 80% of the variance in bone (16). Information on other determinants of bone strength, such as bone geometry and microarchitecture, is not provided on routine BMD measurements. Peripheral quantitative computed tomography (pQCT) and high-resolution pQCT are noninvasive measures of bone microarchitecture that are not widely available for clinical use. The few TS studies using these modalities have shown cortical deficits in the forearm (12, 17–19), whereas compromised trabecular microarchitecture at the radius and tibia was identified in a study using high-resolution pQCT (20). Spinal trabecular bone score (TBS) has recently emerged as an indirect measure of trabecular microarchitecture at the spine (21). This novel software calculates a gray-level textural index derived from routine lumbar spine DXA images and does not require additional radiation exposure. A low TBS is associated with major osteoporotic fractures (defined as vertebral, hip, humeral, and wrist) in postmenopausal women, with similar findings in older men and other conditions associated with secondary osteoporosis (21). Given that estrogen deficiency is associated with fractures at sites rich in trabecular bone and a correlation between TBS and height has not been reported in the literature (22), TBS is an attractive bone imaging tool in TS populations. To date, no studies have investigated TBS in women with TS. Thus, in this cross-sectional and longitudinal study of an Australian adult TS cohort, we aimed to (1) assess spinal TBS and its relationship with clinical and densitometric parameters, (2) investigate longitudinal changes and clinical predictors of TBS and bone density, and (3) explore the capacity of TBS to predict fractures. Subjects and methods Patients This cross-sectional and longitudinal study involved women aged 20 to 50 years who had TS and underwent one or more DXA scans between the years 2006 and 2017 at Monash Medical Centre, Clayton, Victoria, Australia. Participants were identified from the Monash Medical Centre Adult TS Clinic, in addition to a comprehensive search of our institution’s Bone Density Department Database for all women documented as having hypogonadism and aged 20 to 50 years. Subsequent review of medical records confirmed 60 women with a history of TS. Two women taking osteoporosis-specific drugs were excluded, and none of the women had a history of oral glucocorticoid therapy. The study was approved by the Monash Health Human Research Ethics Committee (reference number 07062A). Informed consent was not required because this was a retrospective observational study, and DXA scans were performed as part of routine clinical practice. Clinical data Data regarding karyotype, menstrual history, medication use, and comorbidities were verified from hospital medical records and the Adult TS Clinic database. Data on age at ERT initiation and current use of ERT at the time of DXA scanning were collected, as well as any history of GHT use. Karyotype reports were available for 50 women and were coded as either “45XO” or “other X chromosome abnormalities.” A history of all documented fractures was established from medical records or by radiographically proven fractures from the hospital radiology database. Age at the time of fracture and the skeletal site were recorded. Measurement of bone density and TBS DXA scans were conducted on a single Lunar Prodigy device (GE Healthcare, Piscataway, NJ) at the Monash Medical Centre. For each woman, baseline and serial DXA scans during the study period were included. All women with TS had BMD (g/cm2) measured at the lumbar spine (L1-L4) and femoral neck (FN), and BMD Z-scores at each skeletal site were calculated. Low bone mass was defined as a Z-score ≤−2.0 as per the International Society for Clinical Densitometry guidelines for young premenopausal women (23). TBS measurements were obtained from DXA spinal images using TBS iNsight Software (version 3.0.2.0; Medimaps, Geneva, Switzerland). On the day of the DXA scan, each participant’s height and weight were measured with a wall-mounted stadiometer and electronic scale, respectively. The coefficients of variation for TBS and BMD using a lumbar spine phantom measured daily on the GE Lunar Prodigy were 0.94% and 0.92%, respectively. Statistical analysis The software program IBM SPSS Statistics v23 (IBM Corp, Armonk, NY) was used to conduct all statistical analyses, except when Stata v14.2 statistical software (StataCorp, College Station, TX) was used to test for significant differences between area under the receiver operator curve (AUROC) using pairwise comparison. A P value <0.05 was considered statistically significant. Baseline data Data from the baseline DXA scan of 58 women with TS were expressed as frequencies and median (range) values because continuous variables were not normally distributed. The relationship between baseline TBS, clinical variables, and densitometric variables was examined using univariate and age-adjusted multivariate linear regression analyses. Longitudinal data Women who had serial DXA scans were included in a longitudinal analysis. Longitudinal changes in bone parameters (TBS, spine BMD, spine Z-score, FN BMD, and FN Z-score) were assessed using linear mixed model analysis, where time was the number of years since the baseline DXA scan. All models were analyzed as random intercepts with fixed or random effects, and different candidate covariance models were considered, with the smallest Akaike information criterion used to choose the most appropriate model. The age- and body mass index (BMI)‒adjusted mean change per year (95% CI) for each bone parameter was calculated. The relationships between bone parameters and clinical predictor variables (age, karyotype, current ERT use, age at ERT initiation, and history of GHT) in the longitudinal data set were also assessed using linear mixed model analysis with adjustments for age and BMI. The model assessing the relationship between age and TBS was also further adjusted for potential confounders, including bone density parameters, karyotype, current ERT use, age at ERT initiation, history of GHT, smoking history, and secondary osteoporosis risk factors. The estimated change in bone outcome variable (95% CI) per unit change in clinical predictor variable is presented. Fracture analysis All fractures regardless of precipitating trauma or skeletal site were analyzed. Baseline clinical and densitometric data between fracture and nonfracture groups were compared using the Mann-Whitney U test for continuous variables and χ2 or Fisher’s exact test for categorical variables where appropriate. Because there was a significant difference in height between fracture and nonfracture groups, logistic regression was used to assess height-adjusted differences in bone parameters between groups. AUROC analysis was applied to determine the diagnostic value of age, TBS, spine Z-score, FN Z-score, and composite models to predict all fractures. TBS, spine Z-score, and FN Z-score were analyzed as dichotomous variables using established thresholds for abnormal values: TBS <1.35; spine Z-score ≤−2.0; and FN Z-score ≤−2.0. AUROC value and 95% CI are presented, and differences between AUROCs was tested using pairwise comparison. Results Baseline data Fifty-eight women with TS (median age, 28.5 years; range, 20 to 49 years) were included in this study. Their baseline characteristics, spinal TBS, and DXA-derived areal BMD and Z-scores of the spine and FN are described in Table 1. Table 1. Relationship Between Baseline TBS and Anthropometric, Clinical, and Bone Parameters Using Linear Regression in Women With TS Baseline Parameters n All Unadjusted Age Adjusted R2a B P Value R2a B P Value TBS 58 1.40 (1.10–1.69) — — — — — — Age, y 58 28.5 (20.0–49.0) 0.097 −0.004 0.017 — — — Ethnicity, Asian/white 58 6 (10.3)/52 (89.7) 0.003 0.020 0.708 0.098 0.008 0.882 Height, cm 58 148.0 (134.5–170.0) 0.008 −0.001 0.500 0.132 −0.003 0.142 Weight, kg 58 60.2 (29.4–121.8) 0.029 0.001 0.204 0.117 0.001 0.273 BMI, kg/m2 58 25.5 (13.8–49.5) 0.059 0.004 0.065 0.168 0.004 0.035 Karyotype, 45XO 50 15 (30.0) — 0.005 0.896 — 0.001 0.978 Age at TS diagnosis, y 58 11 (0–46) 0.048 −0.002 0.099 0.120 −0.002 0.241 Primary amenorrhea 58 47 (81.0) 0.003 −0.017 0.689 0.107 −0.031 0.435 Age at ERT initiation,b y 44 15 (11–30) 0.024 −0.002 0.263 0.190 −0.012 0.023 Current ERTc 51 44 (86.3) — 0.110 0.029 — 0.067 0.192 Prior GHT 58 35 (60.3) — 0.036 0.274 — 0.017 0.611 Current or ex-smoker 55 8 (14.5) — −0.011 0.812 — 0.027 0.579 Secondary osteoporosisd 51 9 (17.6) — 0.026 0.587 — 0.003 0.955 Hearing impairment 41 41 (80.4) — 0.011 0.768 — 0.020 0.589 Spine BMD, g/cm2 58 1.081 (0.681–1.411) 0.420 0.466 <0.001 0.500 0.457 <0.001 Spine Z-score 58 −1.0 (−3.9 to 1.9) 0.363 0.056 <0.001 0.459 0.056 <0.001 FN BMD, g/cm2 58 0.859 (0.566–1.123) 0.293 0.514 <0.001 0.345 0.478 <0.001 FN Z-score 58 −0.9 (−3.1 to 1.2) 0.199 0.056 <0.001 0.304 0.057 <0.001 Baseline Parameters n All Unadjusted Age Adjusted R2a B P Value R2a B P Value TBS 58 1.40 (1.10–1.69) — — — — — — Age, y 58 28.5 (20.0–49.0) 0.097 −0.004 0.017 — — — Ethnicity, Asian/white 58 6 (10.3)/52 (89.7) 0.003 0.020 0.708 0.098 0.008 0.882 Height, cm 58 148.0 (134.5–170.0) 0.008 −0.001 0.500 0.132 −0.003 0.142 Weight, kg 58 60.2 (29.4–121.8) 0.029 0.001 0.204 0.117 0.001 0.273 BMI, kg/m2 58 25.5 (13.8–49.5) 0.059 0.004 0.065 0.168 0.004 0.035 Karyotype, 45XO 50 15 (30.0) — 0.005 0.896 — 0.001 0.978 Age at TS diagnosis, y 58 11 (0–46) 0.048 −0.002 0.099 0.120 −0.002 0.241 Primary amenorrhea 58 47 (81.0) 0.003 −0.017 0.689 0.107 −0.031 0.435 Age at ERT initiation,b y 44 15 (11–30) 0.024 −0.002 0.263 0.190 −0.012 0.023 Current ERTc 51 44 (86.3) — 0.110 0.029 — 0.067 0.192 Prior GHT 58 35 (60.3) — 0.036 0.274 — 0.017 0.611 Current or ex-smoker 55 8 (14.5) — −0.011 0.812 — 0.027 0.579 Secondary osteoporosisd 51 9 (17.6) — 0.026 0.587 — 0.003 0.955 Hearing impairment 41 41 (80.4) — 0.011 0.768 — 0.020 0.589 Spine BMD, g/cm2 58 1.081 (0.681–1.411) 0.420 0.466 <0.001 0.500 0.457 <0.001 Spine Z-score 58 −1.0 (−3.9 to 1.9) 0.363 0.056 <0.001 0.459 0.056 <0.001 FN BMD, g/cm2 58 0.859 (0.566–1.123) 0.293 0.514 <0.001 0.345 0.478 <0.001 FN Z-score 58 −0.9 (−3.1 to 1.2) 0.199 0.056 <0.001 0.304 0.057 <0.001 Data are expressed as median (range) or n (%). Boldface indicates P values that are statistically significant. a R2 reported only for continuous variables. b In those with primary amenorrhea. c At time of DXA scan in those with amenorrhea (two patients with normal menses not included). d Including celiac disease (n = 3), inflammatory bowel disease (n = 2), hyperthyroidism (n = 2), primary hyperparathyroidism (n = 1), and type 1 diabetes mellitus (n = 1). View Large Table 1. Relationship Between Baseline TBS and Anthropometric, Clinical, and Bone Parameters Using Linear Regression in Women With TS Baseline Parameters n All Unadjusted Age Adjusted R2a B P Value R2a B P Value TBS 58 1.40 (1.10–1.69) — — — — — — Age, y 58 28.5 (20.0–49.0) 0.097 −0.004 0.017 — — — Ethnicity, Asian/white 58 6 (10.3)/52 (89.7) 0.003 0.020 0.708 0.098 0.008 0.882 Height, cm 58 148.0 (134.5–170.0) 0.008 −0.001 0.500 0.132 −0.003 0.142 Weight, kg 58 60.2 (29.4–121.8) 0.029 0.001 0.204 0.117 0.001 0.273 BMI, kg/m2 58 25.5 (13.8–49.5) 0.059 0.004 0.065 0.168 0.004 0.035 Karyotype, 45XO 50 15 (30.0) — 0.005 0.896 — 0.001 0.978 Age at TS diagnosis, y 58 11 (0–46) 0.048 −0.002 0.099 0.120 −0.002 0.241 Primary amenorrhea 58 47 (81.0) 0.003 −0.017 0.689 0.107 −0.031 0.435 Age at ERT initiation,b y 44 15 (11–30) 0.024 −0.002 0.263 0.190 −0.012 0.023 Current ERTc 51 44 (86.3) — 0.110 0.029 — 0.067 0.192 Prior GHT 58 35 (60.3) — 0.036 0.274 — 0.017 0.611 Current or ex-smoker 55 8 (14.5) — −0.011 0.812 — 0.027 0.579 Secondary osteoporosisd 51 9 (17.6) — 0.026 0.587 — 0.003 0.955 Hearing impairment 41 41 (80.4) — 0.011 0.768 — 0.020 0.589 Spine BMD, g/cm2 58 1.081 (0.681–1.411) 0.420 0.466 <0.001 0.500 0.457 <0.001 Spine Z-score 58 −1.0 (−3.9 to 1.9) 0.363 0.056 <0.001 0.459 0.056 <0.001 FN BMD, g/cm2 58 0.859 (0.566–1.123) 0.293 0.514 <0.001 0.345 0.478 <0.001 FN Z-score 58 −0.9 (−3.1 to 1.2) 0.199 0.056 <0.001 0.304 0.057 <0.001 Baseline Parameters n All Unadjusted Age Adjusted R2a B P Value R2a B P Value TBS 58 1.40 (1.10–1.69) — — — — — — Age, y 58 28.5 (20.0–49.0) 0.097 −0.004 0.017 — — — Ethnicity, Asian/white 58 6 (10.3)/52 (89.7) 0.003 0.020 0.708 0.098 0.008 0.882 Height, cm 58 148.0 (134.5–170.0) 0.008 −0.001 0.500 0.132 −0.003 0.142 Weight, kg 58 60.2 (29.4–121.8) 0.029 0.001 0.204 0.117 0.001 0.273 BMI, kg/m2 58 25.5 (13.8–49.5) 0.059 0.004 0.065 0.168 0.004 0.035 Karyotype, 45XO 50 15 (30.0) — 0.005 0.896 — 0.001 0.978 Age at TS diagnosis, y 58 11 (0–46) 0.048 −0.002 0.099 0.120 −0.002 0.241 Primary amenorrhea 58 47 (81.0) 0.003 −0.017 0.689 0.107 −0.031 0.435 Age at ERT initiation,b y 44 15 (11–30) 0.024 −0.002 0.263 0.190 −0.012 0.023 Current ERTc 51 44 (86.3) — 0.110 0.029 — 0.067 0.192 Prior GHT 58 35 (60.3) — 0.036 0.274 — 0.017 0.611 Current or ex-smoker 55 8 (14.5) — −0.011 0.812 — 0.027 0.579 Secondary osteoporosisd 51 9 (17.6) — 0.026 0.587 — 0.003 0.955 Hearing impairment 41 41 (80.4) — 0.011 0.768 — 0.020 0.589 Spine BMD, g/cm2 58 1.081 (0.681–1.411) 0.420 0.466 <0.001 0.500 0.457 <0.001 Spine Z-score 58 −1.0 (−3.9 to 1.9) 0.363 0.056 <0.001 0.459 0.056 <0.001 FN BMD, g/cm2 58 0.859 (0.566–1.123) 0.293 0.514 <0.001 0.345 0.478 <0.001 FN Z-score 58 −0.9 (−3.1 to 1.2) 0.199 0.056 <0.001 0.304 0.057 <0.001 Data are expressed as median (range) or n (%). Boldface indicates P values that are statistically significant. a R2 reported only for continuous variables. b In those with primary amenorrhea. c At time of DXA scan in those with amenorrhea (two patients with normal menses not included). d Including celiac disease (n = 3), inflammatory bowel disease (n = 2), hyperthyroidism (n = 2), primary hyperparathyroidism (n = 1), and type 1 diabetes mellitus (n = 1). View Large Using established adult norms for TBS (22), 39 women with TS (67%) had normal values (TBS ≥1.35); 15 women (26%) had partially degraded spinal microarchitecture (TBS 1.20 to 1.35); and four women (7%) had degraded microarchitecture (TBS ≤1.20). Low bone mass (Z-score ≤−2.0) at the spine was seen in 15 women (26%), whereas five women (9%) had low bone mass at the FN. The unadjusted and age-adjusted linear relationships between baseline spinal TBS and clinical, anthropometric, and bone density parameters are described in Table 1. Spinal TBS was significantly associated with spine BMD, spine Z-scores, FN BMD, and FN Z-scores in unadjusted and age-adjusted models. After adjustment for age, TBS was associated with BMI and inversely associated with age at ERT initiation. Longitudinal data Longitudinal data were available for 33 women with TS. Median duration of follow-up was 4 years (range, 1 to 10 years), and 48.5% of women had follow-up data available over at least 5 years. Over the follow-up period, 25 women with TS continued ERT; two women who were not receiving therapy at baseline commenced ERT; two women were never receiving ERT; and ERT status was unknown in four women. Longitudinal data were not available for the remaining 25 women because they were either lost to follow-up or did not have a second routine DXA scan at the time of study analysis. According to linear mixed model analysis of the longitudinal data, TBS was strongly associated with spine BMD, spine Z-score, FN BMD, and FN Z-score. The estimated changes (95% CI) in TBS per unit change in spine BMD, spine Z-score, FN BMD, and FN Z-score were 0.397 (0.223, 0.572); 0.049 (0.029, 0.070); 0.333 (0.099, 0.567); and 0.044 (0.017, 0.072), respectively (all age- and BMI-adjusted P < 0.01). During follow-up, no significant longitudinal changes per year in TBS, spine BMD, spine Z-score, FN BMD, and FN Z-score were observed (Table 2). Table 2. Longitudinal Changes in Bone Outcome Measurements Bone Measurements n Δ/y 95% CI P Value a TBS 33 0.003 (−0.003, 0.009) 0.353 Spine BMD, g/cm2 33 0.005 (−0.002, 0.013) 0.166 Spine Z-score 33 0.054 (−0.010, 0.118) 0.096 FN BMD, g/cm2 33 0.005 (−0.001, 0.011) 0.114 FN Z-score 33 0.028 (−0.021, 0.076) 0.256 Bone Measurements n Δ/y 95% CI P Value a TBS 33 0.003 (−0.003, 0.009) 0.353 Spine BMD, g/cm2 33 0.005 (−0.002, 0.013) 0.166 Spine Z-score 33 0.054 (−0.010, 0.118) 0.096 FN BMD, g/cm2 33 0.005 (−0.001, 0.011) 0.114 FN Z-score 33 0.028 (−0.021, 0.076) 0.256 Δ/y = estimated change in bone measurements per year of follow-up. a Adjusted for age and BMI. View Large Table 2. Longitudinal Changes in Bone Outcome Measurements Bone Measurements n Δ/y 95% CI P Value a TBS 33 0.003 (−0.003, 0.009) 0.353 Spine BMD, g/cm2 33 0.005 (−0.002, 0.013) 0.166 Spine Z-score 33 0.054 (−0.010, 0.118) 0.096 FN BMD, g/cm2 33 0.005 (−0.001, 0.011) 0.114 FN Z-score 33 0.028 (−0.021, 0.076) 0.256 Bone Measurements n Δ/y 95% CI P Value a TBS 33 0.003 (−0.003, 0.009) 0.353 Spine BMD, g/cm2 33 0.005 (−0.002, 0.013) 0.166 Spine Z-score 33 0.054 (−0.010, 0.118) 0.096 FN BMD, g/cm2 33 0.005 (−0.001, 0.011) 0.114 FN Z-score 33 0.028 (−0.021, 0.076) 0.256 Δ/y = estimated change in bone measurements per year of follow-up. a Adjusted for age and BMI. View Large TBS was inversely associated with age in this cohort [−0.004 per year (−0.007, −0.001); BMI-adjusted P = 0.014]. This remained statistically significant after adjusting separately for spine BMD, spine Z-score, and FN Z-score (all P < 0.05), but not for FN BMD (P = 0.087). Furthermore, after separate adjustment for clinical parameters, such as karyotype, current ERT, age at ERT initiation, GHT, smoking history, and secondary osteoporosis risk factors, TBS remained inversely associated with age (all P < 0.05). A decline in FN BMD with increasing age was also observed [−0.003 per year (−0.005, −0.001); BMI-adjusted P = 0.008], but no significant relationship with age was seen with spine BMD. The relationship between bone parameters and other clinical predictor variables is shown in Table 3. A delay in ERT initiation was associated with lower TBS, spine BMD, spine Z-score, FN BMD, and FN Z-score (all P < 0.05). Prior GHT was associated with a higher spine BMD, spine Z-score, FN BMD, and FN Z-score (all P < 0.05) but was not associated with TBS. An association between TBS and BMI was not observed in this longitudinal analysis. Table 3. Relationship Between Bone Outcome Measurements and Clinical Predictor Variables Using Multivariate Linear Mixed Model Analysis Bone Measurements Predictor Variable n Δ 95% CI P Value a TBS Karyotype, 45XO 29 −0.023 (−0.115, 0.068) 0.608 Current ERT 30 0.079 (−0.013, 0.171) 0.092 Age at ERT initiation,b y 24 −0.010 (−0.011, −0.010) <0.001 Prior GHT 33 0.024 (−0.067, 0.115) 0.589 Spine BMD, mg/cm2 Karyotype, 45XO 29 0.078 (−0.059, 0.215) 0.255 Current ERT 30 0.002 (0.028, 0.362) 0.955 Age of ERT initiation,b y 24 −0.021 (−0.039, −0.003) 0.024 Prior GHT 33 0.139 (0.018, 0.261) 0.027 Spine Z-score Karyotype, 45XO 29 0.740 (−0.404, 1.884) 0.195 Current ERT 30 0.010 (−0.677, 0.698) 0.975 Age at ERT initiation,b y 24 −1.183 (−1.875, −0.490) 0.001 Prior GHT 33 1.074 (0.196, 1.952) 0.018 FN BMD, mg/cm2 Karyotype, 45XO 29 0.065 (−0.050, 0.180) 0.256 Current ERT 30 −0.013 (−0.059, 0.032) 0.557 Age at ERT initiation,b y 24 −0.020 (−0.034, −0.007) 0.005 Prior GHT 33 0.114 (0.027, 0.202) 0.012 FN Z-score Karyotype, 45XO 29 0.803 (−0.233, 1.839) 0.124 Current ERT 30 −0.083 (−0.480, 0.315) 0.678 Age at ERT initiation,b y 24 −0.170 (−0.285, −0.054) 0.006 Prior GHT 33 0.947 (0.195, 1.700) 0.015 Bone Measurements Predictor Variable n Δ 95% CI P Value a TBS Karyotype, 45XO 29 −0.023 (−0.115, 0.068) 0.608 Current ERT 30 0.079 (−0.013, 0.171) 0.092 Age at ERT initiation,b y 24 −0.010 (−0.011, −0.010) <0.001 Prior GHT 33 0.024 (−0.067, 0.115) 0.589 Spine BMD, mg/cm2 Karyotype, 45XO 29 0.078 (−0.059, 0.215) 0.255 Current ERT 30 0.002 (0.028, 0.362) 0.955 Age of ERT initiation,b y 24 −0.021 (−0.039, −0.003) 0.024 Prior GHT 33 0.139 (0.018, 0.261) 0.027 Spine Z-score Karyotype, 45XO 29 0.740 (−0.404, 1.884) 0.195 Current ERT 30 0.010 (−0.677, 0.698) 0.975 Age at ERT initiation,b y 24 −1.183 (−1.875, −0.490) 0.001 Prior GHT 33 1.074 (0.196, 1.952) 0.018 FN BMD, mg/cm2 Karyotype, 45XO 29 0.065 (−0.050, 0.180) 0.256 Current ERT 30 −0.013 (−0.059, 0.032) 0.557 Age at ERT initiation,b y 24 −0.020 (−0.034, −0.007) 0.005 Prior GHT 33 0.114 (0.027, 0.202) 0.012 FN Z-score Karyotype, 45XO 29 0.803 (−0.233, 1.839) 0.124 Current ERT 30 −0.083 (−0.480, 0.315) 0.678 Age at ERT initiation,b y 24 −0.170 (−0.285, −0.054) 0.006 Prior GHT 33 0.947 (0.195, 1.700) 0.015 Δ = estimated change in bone measurements per unit change in predictor variable. Boldface indicates P values that are statistically significant. a P value adjusted for age and BMI. b In those with primary amenorrhea. View Large Table 3. Relationship Between Bone Outcome Measurements and Clinical Predictor Variables Using Multivariate Linear Mixed Model Analysis Bone Measurements Predictor Variable n Δ 95% CI P Value a TBS Karyotype, 45XO 29 −0.023 (−0.115, 0.068) 0.608 Current ERT 30 0.079 (−0.013, 0.171) 0.092 Age at ERT initiation,b y 24 −0.010 (−0.011, −0.010) <0.001 Prior GHT 33 0.024 (−0.067, 0.115) 0.589 Spine BMD, mg/cm2 Karyotype, 45XO 29 0.078 (−0.059, 0.215) 0.255 Current ERT 30 0.002 (0.028, 0.362) 0.955 Age of ERT initiation,b y 24 −0.021 (−0.039, −0.003) 0.024 Prior GHT 33 0.139 (0.018, 0.261) 0.027 Spine Z-score Karyotype, 45XO 29 0.740 (−0.404, 1.884) 0.195 Current ERT 30 0.010 (−0.677, 0.698) 0.975 Age at ERT initiation,b y 24 −1.183 (−1.875, −0.490) 0.001 Prior GHT 33 1.074 (0.196, 1.952) 0.018 FN BMD, mg/cm2 Karyotype, 45XO 29 0.065 (−0.050, 0.180) 0.256 Current ERT 30 −0.013 (−0.059, 0.032) 0.557 Age at ERT initiation,b y 24 −0.020 (−0.034, −0.007) 0.005 Prior GHT 33 0.114 (0.027, 0.202) 0.012 FN Z-score Karyotype, 45XO 29 0.803 (−0.233, 1.839) 0.124 Current ERT 30 −0.083 (−0.480, 0.315) 0.678 Age at ERT initiation,b y 24 −0.170 (−0.285, −0.054) 0.006 Prior GHT 33 0.947 (0.195, 1.700) 0.015 Bone Measurements Predictor Variable n Δ 95% CI P Value a TBS Karyotype, 45XO 29 −0.023 (−0.115, 0.068) 0.608 Current ERT 30 0.079 (−0.013, 0.171) 0.092 Age at ERT initiation,b y 24 −0.010 (−0.011, −0.010) <0.001 Prior GHT 33 0.024 (−0.067, 0.115) 0.589 Spine BMD, mg/cm2 Karyotype, 45XO 29 0.078 (−0.059, 0.215) 0.255 Current ERT 30 0.002 (0.028, 0.362) 0.955 Age of ERT initiation,b y 24 −0.021 (−0.039, −0.003) 0.024 Prior GHT 33 0.139 (0.018, 0.261) 0.027 Spine Z-score Karyotype, 45XO 29 0.740 (−0.404, 1.884) 0.195 Current ERT 30 0.010 (−0.677, 0.698) 0.975 Age at ERT initiation,b y 24 −1.183 (−1.875, −0.490) 0.001 Prior GHT 33 1.074 (0.196, 1.952) 0.018 FN BMD, mg/cm2 Karyotype, 45XO 29 0.065 (−0.050, 0.180) 0.256 Current ERT 30 −0.013 (−0.059, 0.032) 0.557 Age at ERT initiation,b y 24 −0.020 (−0.034, −0.007) 0.005 Prior GHT 33 0.114 (0.027, 0.202) 0.012 FN Z-score Karyotype, 45XO 29 0.803 (−0.233, 1.839) 0.124 Current ERT 30 −0.083 (−0.480, 0.315) 0.678 Age at ERT initiation,b y 24 −0.170 (−0.285, −0.054) 0.006 Prior GHT 33 0.947 (0.195, 1.700) 0.015 Δ = estimated change in bone measurements per unit change in predictor variable. Boldface indicates P values that are statistically significant. a P value adjusted for age and BMI. b In those with primary amenorrhea. View Large Fracture analysis Fracture history was available for 54 women with TS. Twenty-two fractures occurred in 17 of 54 women (31%), and nine of 54 women (16%) sustained a fracture after 20 years of age. Most fractures occurred in the upper limb (52.6%), followed by the lower limb (26.3%), vertebra (10.5%), clavicle (5.3%), and sternum (5.3%), with 64% occurring in predominantly trabecular bone (eight fractures at the end of long bones, four fractures in the short bones of the hands and feet, and two vertebral fractures). Baseline clinical characteristics and bone density data between fracture and nonfracture groups are described in Table 4. The fracture and nonfracture groups were similar except for height. No differences in spine BMD, spine Z-score, FN BMD, or FN Z-score were observed between groups even after adjusting for height. Table 4. Baseline Clinical Characteristics and Bone Parameters of Women With TS in Fracture and Nonfracture Groups Clinical Parameter n Fracture (n = 17) Nonfracture (n = 37) P Value Age, y 54 30.0 (20–47) 24.0 (20.0–49.0) 0.155 Ethnicity, Asian/white 54 1 (1.9)/16 (29.6) 5 (9.3)/32 (59.3) 0.407 Height, cm 54 151.0 (134.5–170.0) 147.2 (134.5–162.5) 0.041 Weight, kg 54 67.7 (29.4–121.8) 58.5 (39.8–110.8) 0.896 BMI, kg/m2 54 25.4 (13.8–39.0) 25.6 (18.4–49.5) 0.682 Karyotype, X monosomy 46 3 (6.5) 9 (19.6) 0.607 Age at TS diagnosis, y 54 12.0 (0.0–46.0) 11.0 (0.0–39.0) 0.581 Primary amenorrhea 54 13 (24.1) 30 (55.6) 0.696 Age at ERT initiation,a y 40 16.0 (12–46) 16.0 (0.0–39.0) 0.497 Current ERTb 50 12 (24.0) 31 (62.0) 0.124 Prior GHT 54 8 (14.8) 23 (42.6) 0.297 Current or ex-smoker 54 3 (5.6) 5 (9.3) 0.491 Secondary osteoporosis risk factorsc 49 4 (8.2) 5 (10.2) 0.322 Hearing impairment 47 9 (19.1) 28 (59.6) 0.272 TBS 54 1.34 (1.10–1.69) 1.42 (1.10–1.64) 0.101 Spine BMD, g/cm2 54 1.049 (0.721–1.411) 1.104 (0.681–1.343) 0.662 Spine Z-score 54 −0.9 (−3.9 to 1.9) −1.0 (−3.5 to 1.3) 0.823 FN BMD, g/cm2 54 0.833 (0.651–1.094) 0.911 (0.566–1.123) 0.635 FN Z-score 54 −0.8 (−2.1 to 1.2) −0.8 (−3.1 to 1.2) 0.716 Clinical Parameter n Fracture (n = 17) Nonfracture (n = 37) P Value Age, y 54 30.0 (20–47) 24.0 (20.0–49.0) 0.155 Ethnicity, Asian/white 54 1 (1.9)/16 (29.6) 5 (9.3)/32 (59.3) 0.407 Height, cm 54 151.0 (134.5–170.0) 147.2 (134.5–162.5) 0.041 Weight, kg 54 67.7 (29.4–121.8) 58.5 (39.8–110.8) 0.896 BMI, kg/m2 54 25.4 (13.8–39.0) 25.6 (18.4–49.5) 0.682 Karyotype, X monosomy 46 3 (6.5) 9 (19.6) 0.607 Age at TS diagnosis, y 54 12.0 (0.0–46.0) 11.0 (0.0–39.0) 0.581 Primary amenorrhea 54 13 (24.1) 30 (55.6) 0.696 Age at ERT initiation,a y 40 16.0 (12–46) 16.0 (0.0–39.0) 0.497 Current ERTb 50 12 (24.0) 31 (62.0) 0.124 Prior GHT 54 8 (14.8) 23 (42.6) 0.297 Current or ex-smoker 54 3 (5.6) 5 (9.3) 0.491 Secondary osteoporosis risk factorsc 49 4 (8.2) 5 (10.2) 0.322 Hearing impairment 47 9 (19.1) 28 (59.6) 0.272 TBS 54 1.34 (1.10–1.69) 1.42 (1.10–1.64) 0.101 Spine BMD, g/cm2 54 1.049 (0.721–1.411) 1.104 (0.681–1.343) 0.662 Spine Z-score 54 −0.9 (−3.9 to 1.9) −1.0 (−3.5 to 1.3) 0.823 FN BMD, g/cm2 54 0.833 (0.651–1.094) 0.911 (0.566–1.123) 0.635 FN Z-score 54 −0.8 (−2.1 to 1.2) −0.8 (−3.1 to 1.2) 0.716 Data are expressed as median (range) or n (%). Boldface indicates P values that are statistically significant. a In those with primary amenorrhea. b At time of DXA scan in those with amenorrhea (two patients with normal menses not included). c Including celiac disease (n = 3), inflammatory bowel disease (n = 2), hyperthyroidism (n = 2), primary hyperparathyroidism (n = 1), and type 1 diabetes mellitus (n = 1). View Large Table 4. Baseline Clinical Characteristics and Bone Parameters of Women With TS in Fracture and Nonfracture Groups Clinical Parameter n Fracture (n = 17) Nonfracture (n = 37) P Value Age, y 54 30.0 (20–47) 24.0 (20.0–49.0) 0.155 Ethnicity, Asian/white 54 1 (1.9)/16 (29.6) 5 (9.3)/32 (59.3) 0.407 Height, cm 54 151.0 (134.5–170.0) 147.2 (134.5–162.5) 0.041 Weight, kg 54 67.7 (29.4–121.8) 58.5 (39.8–110.8) 0.896 BMI, kg/m2 54 25.4 (13.8–39.0) 25.6 (18.4–49.5) 0.682 Karyotype, X monosomy 46 3 (6.5) 9 (19.6) 0.607 Age at TS diagnosis, y 54 12.0 (0.0–46.0) 11.0 (0.0–39.0) 0.581 Primary amenorrhea 54 13 (24.1) 30 (55.6) 0.696 Age at ERT initiation,a y 40 16.0 (12–46) 16.0 (0.0–39.0) 0.497 Current ERTb 50 12 (24.0) 31 (62.0) 0.124 Prior GHT 54 8 (14.8) 23 (42.6) 0.297 Current or ex-smoker 54 3 (5.6) 5 (9.3) 0.491 Secondary osteoporosis risk factorsc 49 4 (8.2) 5 (10.2) 0.322 Hearing impairment 47 9 (19.1) 28 (59.6) 0.272 TBS 54 1.34 (1.10–1.69) 1.42 (1.10–1.64) 0.101 Spine BMD, g/cm2 54 1.049 (0.721–1.411) 1.104 (0.681–1.343) 0.662 Spine Z-score 54 −0.9 (−3.9 to 1.9) −1.0 (−3.5 to 1.3) 0.823 FN BMD, g/cm2 54 0.833 (0.651–1.094) 0.911 (0.566–1.123) 0.635 FN Z-score 54 −0.8 (−2.1 to 1.2) −0.8 (−3.1 to 1.2) 0.716 Clinical Parameter n Fracture (n = 17) Nonfracture (n = 37) P Value Age, y 54 30.0 (20–47) 24.0 (20.0–49.0) 0.155 Ethnicity, Asian/white 54 1 (1.9)/16 (29.6) 5 (9.3)/32 (59.3) 0.407 Height, cm 54 151.0 (134.5–170.0) 147.2 (134.5–162.5) 0.041 Weight, kg 54 67.7 (29.4–121.8) 58.5 (39.8–110.8) 0.896 BMI, kg/m2 54 25.4 (13.8–39.0) 25.6 (18.4–49.5) 0.682 Karyotype, X monosomy 46 3 (6.5) 9 (19.6) 0.607 Age at TS diagnosis, y 54 12.0 (0.0–46.0) 11.0 (0.0–39.0) 0.581 Primary amenorrhea 54 13 (24.1) 30 (55.6) 0.696 Age at ERT initiation,a y 40 16.0 (12–46) 16.0 (0.0–39.0) 0.497 Current ERTb 50 12 (24.0) 31 (62.0) 0.124 Prior GHT 54 8 (14.8) 23 (42.6) 0.297 Current or ex-smoker 54 3 (5.6) 5 (9.3) 0.491 Secondary osteoporosis risk factorsc 49 4 (8.2) 5 (10.2) 0.322 Hearing impairment 47 9 (19.1) 28 (59.6) 0.272 TBS 54 1.34 (1.10–1.69) 1.42 (1.10–1.64) 0.101 Spine BMD, g/cm2 54 1.049 (0.721–1.411) 1.104 (0.681–1.343) 0.662 Spine Z-score 54 −0.9 (−3.9 to 1.9) −1.0 (−3.5 to 1.3) 0.823 FN BMD, g/cm2 54 0.833 (0.651–1.094) 0.911 (0.566–1.123) 0.635 FN Z-score 54 −0.8 (−2.1 to 1.2) −0.8 (−3.1 to 1.2) 0.716 Data are expressed as median (range) or n (%). Boldface indicates P values that are statistically significant. a In those with primary amenorrhea. b At time of DXA scan in those with amenorrhea (two patients with normal menses not included). c Including celiac disease (n = 3), inflammatory bowel disease (n = 2), hyperthyroidism (n = 2), primary hyperparathyroidism (n = 1), and type 1 diabetes mellitus (n = 1). View Large At baseline, prevalent fractures within the normal, partially degraded, and degraded TBS groups were recorded in eight of 36 women (22%), six of 14 women (43%), and three of four women (75%), respectively. In the degraded TBS group, two women recorded fractures at the wrist, and the third subject recorded a fracture involving the small bones of the hand. A statistically significant difference in fracture prevalence between normal and degraded TBS groups was found (height-adjusted P = 0.023) (Fig. 1), and this remained significant even after adjustments for BMD and Z-scores at the spine and FN, respectively (all P < 0.05). However, no difference in fracture prevalence was seen between normal and low bone mass groups at the spine and FN. Figure 1. View largeDownload slide Percentage of women with TS who fracture per bone outcome measurement. TBS: normal TBS ≥1.35; partially degraded TBS 1.20 to 1.35; and degraded TBS ≤1.20. Spine and FN Z-score: normal Z-score >−2.0; low Z-score ≤−2.0. Figure 1. View largeDownload slide Percentage of women with TS who fracture per bone outcome measurement. TBS: normal TBS ≥1.35; partially degraded TBS 1.20 to 1.35; and degraded TBS ≤1.20. Spine and FN Z-score: normal Z-score >−2.0; low Z-score ≤−2.0. The diagnostic value of spinal TBS in predicting prevalent fractures was assessed using AUROC analysis and was compared with AUROC for age, spine Z-score, and FN Z-score (Table 5). Although TBS alone was not predictive of prevalent fractures, TBS had a significantly higher AUROC than spine Z-score (0.643 vs 0.483; P = 0.013) and FN Z-score (0.643 vs 0.489; P = 0.033). The inclusion of TBS in a model containing age was predictive of prevalent fractures (P = 0.024) (Table 5). Table 5. AUROC Analysis for Fracture Outcomes n AUROC (95% CI) P Value All fractures  Age 58 0.621 (0.463, 0.779) 0.157  TBSa,b 58 0.643 (0.502, 0.784) 0.094  Spine Z-scorea 58 0.483 (0.316, 0.649) 0.116  FN Z-scoreb 58 0.489 (0.323, 0.655) 0.896 Composite models  Age + TBS 58 0.692 (0.542, 0.843) 0.024  Age + Spine Z-score 58 0.634 (0.484, 0.785) 0.116  Age + FN Z-score 58 0.624 (0.467, 0.781) 0.186 n AUROC (95% CI) P Value All fractures  Age 58 0.621 (0.463, 0.779) 0.157  TBSa,b 58 0.643 (0.502, 0.784) 0.094  Spine Z-scorea 58 0.483 (0.316, 0.649) 0.116  FN Z-scoreb 58 0.489 (0.323, 0.655) 0.896 Composite models  Age + TBS 58 0.692 (0.542, 0.843) 0.024  Age + Spine Z-score 58 0.634 (0.484, 0.785) 0.116  Age + FN Z-score 58 0.624 (0.467, 0.781) 0.186 Thresholds used in analysis: spine Z-score ≤−2.0, FN Z-score ≤−2.0, TBS <1.35. Boldface indicates P values that are statistically significant. a TBS had a significantly higher AUROC than spine Z-score (P = 0.0130). b TBS had a significantly higher AUROC than FN Z-score (P = 0.0327). View Large Table 5. AUROC Analysis for Fracture Outcomes n AUROC (95% CI) P Value All fractures  Age 58 0.621 (0.463, 0.779) 0.157  TBSa,b 58 0.643 (0.502, 0.784) 0.094  Spine Z-scorea 58 0.483 (0.316, 0.649) 0.116  FN Z-scoreb 58 0.489 (0.323, 0.655) 0.896 Composite models  Age + TBS 58 0.692 (0.542, 0.843) 0.024  Age + Spine Z-score 58 0.634 (0.484, 0.785) 0.116  Age + FN Z-score 58 0.624 (0.467, 0.781) 0.186 n AUROC (95% CI) P Value All fractures  Age 58 0.621 (0.463, 0.779) 0.157  TBSa,b 58 0.643 (0.502, 0.784) 0.094  Spine Z-scorea 58 0.483 (0.316, 0.649) 0.116  FN Z-scoreb 58 0.489 (0.323, 0.655) 0.896 Composite models  Age + TBS 58 0.692 (0.542, 0.843) 0.024  Age + Spine Z-score 58 0.634 (0.484, 0.785) 0.116  Age + FN Z-score 58 0.624 (0.467, 0.781) 0.186 Thresholds used in analysis: spine Z-score ≤−2.0, FN Z-score ≤−2.0, TBS <1.35. Boldface indicates P values that are statistically significant. a TBS had a significantly higher AUROC than spine Z-score (P = 0.0130). b TBS had a significantly higher AUROC than FN Z-score (P = 0.0327). View Large Discussion This study investigated the potential role of spinal TBS in the evaluation of skeletal health in an adult TS cohort. An age-related decline in TBS was observed in both cross-sectional and longitudinal analyses. During follow-up of up to 10 years, spinal TBS, spine BMD, spine Z-score, FN BMD, and FN Z-score remained stable in this TS cohort. Spinal TBS was positively associated with lumbar spine and FN BMD and Z-scores, and was adversely influenced by a delay in ERT initiation. Spine and FN bone parameters were also negatively associated with age at ERT initiation, whereas GHT appeared to have a protective effect. In this pilot study, TBS was a better predictor of prevalent fractures than low bone mass at the spine or FN. Spinal TBS is a textural parameter that provides an indirect measure of spinal microarchitecture that is independently associated with history of fracture and incidence of fracture and is complementary to BMD in estimating fracture risk in men and women (21). Most clinical studies have investigated TBS in older cohorts, and limited literature has reviewed changes in TBS in younger women. An age-related decline in TBS in healthy premenopausal women has been described in a small number of studies (24–28), whereas one North American study reported stable TBS values in women aged 30 to 45 years followed by a steady decline with age thereafter (29). The magnitude of the age-related decline in TBS described in a study of Thai women aged 30 to 50 years was 0.006 per year (27), whereas Iki et al. (28) reported a decline of 0.0013 per year in a Japanese cohort aged 20 to 40 years. The age-related decline in TBS of 0.004 per year seen in our TS cohort is not dissimilar to the decline described by Sritara et al. (27) and Iki et al. (28) and suggests that changes in TBS with age may not be accelerated in women with TS. In the longitudinal subgroup analysis, DXA bone parameters remained stable. This is possibly explained by the high rates of ERT (86%) use over the follow-up period, and it is reassuring that bone microarchitecture and density can be maintained in women with continued ERT exposure. This is consistent with published longitudinal studies in adult women with TS. Cleemann et al. (30) followed up 54 women with TS (aged 43.0 ± 9.95 years) over 6 years and found that BMD could be maintained in adults with TS by encouraging adequate ERT, a healthy lifestyle, and a high intake of calcium and vitamin D. Suganuma et al. (31) also showed that initiation of ERT in adult women with TS maintained bone density over 7 years of follow-up. Recently, Soucek et al. (7) showed that fractures were not increased in a cohort of women with TS receiving ERT. The observation that TBS in adulthood is adversely affected by delay in ERT initiation in women with TS is an important finding, although not unexpected. We also found in our longitudinal analysis that late pubertal induction detrimentally affected spine and FN bone density, a finding that is consistent with our previously published cross-sectional work showing that a delay in commencing ERT in primary amenorrhea was associated with lower BMD (6). Similar findings in bone density have been seen in other TS cohorts (32, 33), but the effect on microarchitecture has not been described. Gonadal failure manifesting as primary amenorrhea affects up to 85% of women with TS (34). Estrogen is a key hormonal regulator of bone health, with an important role in bone mass accrual during skeletal growth, skeletal homeostasis in adulthood, and accelerated bone loss during menopause (35). As such, estrogen deficiency is a well-described risk factor for osteoporosis in TS, and current professional TS guidelines recommend early pubertal induction with ERT by the age of 13 years in those with primary amenorrhea (36). Our finding supports current guidelines of early ERT initiation for adequate attainment of normal bone microarchitecture and peak bone mass. Limited literature has examined the effect of puberty on spinal trabecular microarchitecture. Two published studies of healthy pediatric female populations showed that pubertal development was associated with an increase in TBS. Shawwa et al. (37) reported a steady increase in TBS with age in a Lebanese cohort aged 10 to 17 years, with a higher mean TBS level in postmenarchal girls than in premenarchal girls (1.389 vs 1.293; P < 0.001). Dowthwaite et al. (38) showed that normal adult TBS levels were attained within 1 year postmenarche, 2 or more years before expected peak bone mass attainment at the spine. Correlation between TBS and pubertal stage was also seen in a small cohort study of adolescents with anorexia nervosa (39). Combined with our findings, this suggests that spinal TBS may be a valuable marker of inadequate trabecular microarchitecture attainment in girls with TS, particularly in those with delayed menarche. The positive effect of prior GHT on bone density seen in our TS cohort is interesting and has been reported in some studies (8, 9) but not in others (11, 13, 31). An association between GHT and TBS was not seen in our study. The discrepancy between GHT-related BMD and TBS changes in our study may be due to GHT-related increases in height and bone size, with this larger bone translating to higher areal BMD, as opposed to direct effects of GHT on volumetric bone density and microarchitecture. To date, the literature on the effects of GHT on bone in TS has remained inconsistent. In a case-control study of 28 women with TS, Nour et al. (40) found an increase in height and overall bone size in GH-treated women with TS compared with a non‒GH-treated group, but no difference in BMD or bone microarchitecture as measured by high-resolution pQCT. In contrast, a more recent study by Soucek et al. (7) showed a positive association between the duration of GHT and cortical volumetric BMD. We found that low TBS had a higher AUROC for prevalent fracture than traditional DXA parameters and that TBS in a model with age was predictive of fractures. This study adds to the body of evidence regarding the relationship between low TBS and fragility fractures in adults, which has already been described in older populations and in other conditions with secondary osteoporosis, such as diabetes mellitus, primary hyperparathyroidism, and rheumatoid arthritis (21). Our finding is of particular relevance to TS populations because TBS is not associated with height, which is a pitfall in the use of DXA-derived areal BMD in women with TS and short stature. Furthermore, because TBS uses the same region of interest as DXA-derived spinal images, no additional testing or radiation exposure is required, and spinal images can be retrospectively analyzed for TBS values. As such, spinal TBS is an attractive clinical tool that may be easily integrated into the routine care of women with TS to improve fracture risk assessment. We found a statistically significant positive correlation between TBS and BMI in an age-adjusted regression analysis; however, this relationship was not seen in our longitudinal data analysis. There may be a relationship between body composition and TBS in subjects with TS, and future studies should explore this relationship further. Although several significant findings were elucidated, we are mindful of the limitations of this study, particularly the retrospective study design and relatively small sample size. We did not have a control group, although we used established adult norms to interpret TBS and BMD Z-scores. Reliance on medical records may have underestimated our fracture prevalence, and we could not analyze incident fracture risk. Further, we were unable to reliably differentiate minimal trauma fractures from traumatic fractures. We were unable to longitudinally assess adherence to ERT or GHT; however, we expect nonadherence to ERT to be common because the prevalence of nonadherence was previously reported to be 15% to 37% in TS cohorts (41, 42). It is possible that nonadherence attenuated the relationship between TBS and bone density. Because we found a relationship between a delay in ERT initiation and TBS, we postulate that nonadherence to ERT may be associated with lower TBS values. Finally, we could not consistently obtain detailed data on type of ERT or dose and duration of GHT and hence could not directly assess the dose effect of ERT or GHT on TBS and fractures. This study has a number of strengths that make it relevant to the TS literature. In particular, the use of a novel fracture assessment tool and the longitudinal follow-up of up to 10 years make our study one of the longest TS studies with bone densitometric data in the literature. The majority of women (96%) attended the multidisciplinary adult TS clinic at our institution and thus had standardized screening questionnaire forms at each appointment, ensuring consistency in the recording of data. Conclusion In conclusion, we assessed changes in TBS in women with TS, and our preliminary findings have important clinical implications for the management of bone health in women with the condition. First, our findings suggest that TBS provides additional information for the diagnosis and risk stratification of skeletal fragility in adult women with TS, information that is independent of areal BMD measurements. A larger prospective study is required to validate spinal TBS as a fracture prediction tool in TS and to replicate our finding that it is superior to BMD. Second, a delay in ERT initiation adversely influenced all bone outcome measures, including TBS. This highlights the importance of early ERT to optimize bone density and trabecular microarchitecture in adulthood. Future pediatric studies should further explore the direct relationship between spinal TBS and pubertal induction in TS and whether an optimal ERT dose exists to provide adequate TBS attainment. Abbreviations: Abbreviations: AUROC area under the receiver operator curve BMD bone mineral density BMI body mass index DXA dual-energy X-ray absorptiometry ERT estrogen replacement therapy FN femoral neck GHT growth hormone therapy pQCT peripheral quantitative computed tomography TBS trabecular bone score TS Turner syndrome Acknowledgments The authors wish to acknowledge the Monash Health Adult Turner Syndrome Clinic and the patients with TS. We thank Ms. Stella May Gwini for assistance with the biostatistical analysis. Financial Support: This work was supported by the Amgen-Osteoporosis Australia-Australian New Zealand Bone & Mineral Society Clinical Grant (to A.V.). P.W. is supported by the National Health and Medical Research Council of Australia Early Career Fellowship. Disclosure Summary: The authors have nothing to disclose. References 1. Nielsen J , Wohlert M . Chromosome abnormalities found among 34,910 newborn children: results from a 13-year incidence study in Arhus, Denmark . Hum Genet . 1991 ; 87 ( 1 ): 81 – 83 . Google Scholar Crossref Search ADS PubMed 2. Shore RM , Chesney RW , Mazess RB , Rose PG , Bargman GJ . Skeletal demineralization in Turner’s syndrome . Calcif Tissue Int . 1982 ; 34 ( 1 ): 519 – 522 . Google Scholar Crossref Search ADS PubMed 3. Gravholt CH , Juul S , Naeraa RW , Hansen J . Morbidity in Turner syndrome . J Clin Epidemiol . 1998 ; 51 ( 2 ): 147 – 158 . Google Scholar Crossref Search ADS PubMed 4. Gravholt CH , Vestergaard P , Hermann AP , Mosekilde L , Brixen K , Christiansen JS . Increased fracture rates in Turner’s syndrome: a nationwide questionnaire survey . Clin Endocrinol (Oxf) . 2003 ; 59 ( 1 ): 89 – 96 . Google Scholar Crossref Search ADS PubMed 5. Davies MC , Gulekli B , Jacobs HS . Osteoporosis in Turner’s syndrome and other forms of primary amenorrhoea . Clin Endocrinol (Oxf) . 1995 ; 43 ( 6 ): 741 – 746 . Google Scholar Crossref Search ADS PubMed 6. Nguyen HH , Wong P , Strauss BJ , Jones G , Ebeling PR , Milat F , Vincent A . Delay in estrogen commencement is associated with lower bone mineral density in Turner syndrome . Climacteric . 2017 ; 20 ( 5 ): 436 – 441 . Google Scholar Crossref Search ADS PubMed 7. Soucek O , Schönau E , Lebl J , Willnecker J , Hlavka Z , Sumnik Z . 6-Year follow-up of fracture incidence and volumetric bone mineral density development in girls with Turner syndrome . J Clin Endocrinol Metab . 2018 ; 103 ( 3 ): 1188 – 1197 . Google Scholar Crossref Search ADS PubMed 8. Sas TC , de Muinck Keizer-Schrama SM , Stijnen T , van Teunenbroek A , van Leeuwen WJ , Asarfi A , van Rijn RR , Drop SL ; Dutch Advisory Group on Growth Hormone . Bone mineral density assessed by phalangeal radiographic absorptiometry before and during long-term growth hormone treatment in girls with Turner’s syndrome participating in a randomized dose-response study . Pediatr Res . 2001 ; 50 ( 3 ): 417 – 422 . Google Scholar Crossref Search ADS PubMed 9. Bertelloni S , Cinquanta L , Baroncelli GI , Simi P , Rossi S , Saggese G . Volumetric bone mineral density in young women with Turner’s syndrome treated with estrogens or estrogens plus growth hormone . Horm Res . 2000 ; 53 ( 2 ): 72 – 76 . Google Scholar PubMed 10. Prentice A , Parsons TJ , Cole TJ . Uncritical use of bone mineral density in absorptiometry may lead to size-related artifacts in the identification of bone mineral determinants . Am J Clin Nutr . 1994 ; 60 ( 6 ): 837 – 842 . Google Scholar Crossref Search ADS PubMed 11. Nadeem M , Roche EF . Bone mineral density in Turner’s syndrome and the influence of pubertal development . Acta Paediatr . 2014 ; 103 ( 1 ): e38 – e42 . Google Scholar Crossref Search ADS PubMed 12. Holroyd CR , Davies JH , Taylor P , Jameson K , Rivett C , Cooper C , Dennison EM . Reduced cortical bone density with normal trabecular bone density in girls with Turner syndrome . Osteoporos Int . 2010 ; 21 ( 12 ): 2093 – 2099 . Google Scholar Crossref Search ADS PubMed 13. Högler W , Briody J , Moore B , Garnett S , Lu PW , Cowell CT . Importance of estrogen on bone health in Turner syndrome: a cross-sectional and longitudinal study using dual-energy X-ray absorptiometry . J Clin Endocrinol Metab . 2004 ; 89 ( 1 ): 193 – 199 . Google Scholar Crossref Search ADS PubMed 14. Ross JL , Long LM , Feuillan P , Cassorla F , Cutler GB Jr . Normal bone density of the wrist and spine and increased wrist fractures in girls with Turner’s syndrome . J Clin Endocrinol Metab . 1991 ; 73 ( 2 ): 355 – 359 . Google Scholar Crossref Search ADS PubMed 15. Shaw NJ , Rehan VK , Husain S , Marshall T , Smith CS . Bone mineral density in Turner’s syndrome: a longitudinal study . Clin Endocrinol (Oxf) . 1997 ; 47 ( 3 ): 367 – 370 . Google Scholar Crossref Search ADS PubMed 16. Kushida K . Bone mineral density measurements for diagnosis of osteoporosis and the risk assessment [in Japanese] . Clin Calcium . 2004 ; 14 ( 12 ): 11 – 17 . Google Scholar PubMed 17. Bechtold S , Rauch F , Noelle V , Donhauser S , Neu CM , Schoenau E , Schwarz HP . Musculoskeletal analyses of the forearm in young women with Turner syndrome: a study using peripheral quantitative computed tomography . J Clin Endocrinol Metab . 2001 ; 86 ( 12 ): 5819 – 5823 . Google Scholar Crossref Search ADS PubMed 18. Soucek O , Lebl J , Snajderova M , Kolouskova S , Rocek M , Hlavka Z , Cinek O , Rittweger J , Sumnik Z . Bone geometry and volumetric bone mineral density in girls with Turner syndrome of different pubertal stages . Clin Endocrinol (Oxf) . 2011 ; 74 ( 4 ): 445 – 452 . Google Scholar Crossref Search ADS PubMed 19. Soucek O , Zapletalova J , Zemkova D , Snajderova M , Novotna D , Hirschfeldova K , Plasilova I , Kolouskova S , Rocek M , Hlavka Z , Lebl J , Sumnik Z . Prepubertal girls with Turner syndrome and children with isolated SHOX deficiency have similar bone geometry at the radius . J Clin Endocrinol Metab . 2013 ; 98 ( 7 ): E1241 – E1247 . Google Scholar Crossref Search ADS PubMed 20. Hansen S , Brixen K , Gravholt CH . Compromised trabecular microarchitecture and lower finite element estimates of radius and tibia bone strength in adults with turner syndrome: a cross-sectional study using high-resolution-pQCT . J Bone Miner Res . 2012 ; 27 ( 8 ): 1794 – 1803 . Google Scholar Crossref Search ADS PubMed 21. Silva BC , Broy SB , Boutroy S , Schousboe JT , Shepherd JA , Leslie WD . Fracture risk prediction by non-BMD DXA measures: the 2015 ISCD official positions part 2: trabecular bone score . J Clin Densitom . 2015 ; 18 ( 3 ): 309 – 330 . Google Scholar Crossref Search ADS PubMed 22. Silva BC , Leslie WD , Resch H , Lamy O , Lesnyak O , Binkley N , McCloskey EV , Kanis JA , Bilezikian JP . Trabecular bone score: a noninvasive analytical method based upon the DXA image . J Bone Miner Res . 2014 ; 29 ( 3 ): 518 – 530 . Google Scholar Crossref Search ADS PubMed 23. International Society for Clinical Densitometry . Official ISCD positions, 2015—adult & pediatric. Available at: iscd.app.box.com/OP-ISCD-2015-Adult. Accessed 5 April 2018 . 24. Looker AC , Sarafrazi Isfahani N , Fan B , Shepherd JA . Trabecular bone scores and lumbar spine bone mineral density of US adults: comparison of relationships with demographic and body size variables . Osteoporos Int . 2016 ; 27 ( 8 ): 2467 – 2475 . Google Scholar Crossref Search ADS PubMed 25. Cheng P , Qi HM , Di WJ , Liu J , Yu J , Lv S , Shen YL , Zha JM , Cai JM , Lai B , Ding GX . Establishment of TBS reference plots and correlation between TBS and BMD in healthy mainland Chinese women . Arch Osteoporos . 2016 ; 11 ( 1 ): 5 . Google Scholar Crossref Search ADS PubMed 26. Bazzocchi A , Ponti F , Diano D , Amadori M , Albisinni U , Battista G , Guglielmi G . Trabecular bone score in healthy ageing . Br J Radiol . 2015 ; 88 ( 1052 ): 20140865 . Google Scholar Crossref Search ADS PubMed 27. Sritara C , Thakkinstian A , Ongphiphadhanakul B , Amnuaywattakorn S , Utamakul C , Akrawichien T , Vathesatogkit P , Sritara P . Age-adjusted dual X-ray absorptiometry-derived trabecular bone score curve for the lumbar spine in Thai females and males . J Clin Densitom . 2016 ; 19 ( 4 ): 494 – 501 . Google Scholar Crossref Search ADS PubMed 28. Iki M , Tamaki J , Sato Y , Winzenrieth R , Kagamimori S , Kagawa Y , Yoneshima H . Age-related normative values of trabecular bone score (TBS) for Japanese women: the Japanese Population-based Osteoporosis (JPOS) study . Osteoporos Int . 2015 ; 26 ( 1 ): 245 – 252 . Google Scholar Crossref Search ADS PubMed 29. Simonelli C , Leib E , Mossman N , Winzenrieth R , Hans D , McClung M . Creation of an age-adjusted, dual-energy x-ray absorptiometry-derived trabecular bone score curve for the lumbar spine in non-Hispanic US white women . J Clin Densitom . 2014 ; 17 ( 2 ): 314 – 319 . Google Scholar Crossref Search ADS PubMed 30. Cleemann L , Hjerrild BE , Lauridsen AL , Heickendorff L , Christiansen JS , Mosekilde L , Gravholt CH . Long-term hormone replacement therapy preserves bone mineral density in Turner syndrome . Eur J Endocrinol . 2009 ; 161 ( 2 ): 251 – 257 . Google Scholar Crossref Search ADS PubMed 31. Suganuma N , Furuhashi M , Hirooka T , Moriwaki T , Hasegawa Y , Mori O , Ogawa M . Bone mineral density in adult patients with Turner’s syndrome: analyses of the effectiveness of GH and ovarian steroid hormone replacement therapies . Endocr J . 2003 ; 50 ( 3 ): 263 – 269 . Google Scholar Crossref Search ADS PubMed 32. Kodama M , Komura H , Kodama T , Nishio Y , Kimura T . Estrogen therapy initiated at an early age increases bone mineral density in Turner syndrome patients . Endocr J . 2012 ; 59 ( 2 ): 153 – 159 . Google Scholar Crossref Search ADS PubMed 33. Mora S , Weber G , Guarneri MP , Nizzoli G , Pasolini D , Chiumello G . Effect of estrogen replacement therapy on bone mineral content in girls with Turner syndrome . Obstet Gynecol . 1992 ; 79 ( 5 Pt 1 ): 747 – 751 . Google Scholar PubMed 34. Pasquino AM , Passeri F , Pucarelli I , Segni M , Municchi G ; Italian Study Group for Turner’s Syndrome . Spontaneous pubertal development in Turner’s syndrome . J Clin Endocrinol Metab . 1997 ; 82 ( 6 ): 1810 – 1813 . Google Scholar PubMed 35. Manolagas SC , O’Brien CA , Almeida M . The role of estrogen and androgen receptors in bone health and disease . Nat Rev Endocrinol . 2013 ; 9 ( 12 ): 699 – 712 . Google Scholar Crossref Search ADS PubMed 36. Gravholt CH , Andersen NH , Conway GS , Dekkers OM , Geffner ME , Klein KO , Lin AE, Mauras N, Quigley CA, Rubin K, Sandberg DE, Sas TCJ, Silberbach M, Söderström-Anttila V, Stochholm K, van Alfen-van derVelden JA, Woelfle J, Backeljauw PF; International Turner Syndrome Consensus Group. Clinical practice guidelines for the care of girls and women with Turner syndrome: proceedings from the 2016 Cincinnati International Turner Syndrome Meeting. Eur J Endocrinol. 2017;177(3):G1–G70 . 37. Shawwa K , Arabi A , Nabulsi M , Maalouf J , Salamoun M , Choucair M , Hans D , El-Hajj Fuleihan G . Predictors of trabecular bone score in school children . Osteoporos Int . 2016 ; 27 ( 2 ): 703 – 710 . Google Scholar Crossref Search ADS PubMed 38. Dowthwaite JN , Winzenrieth R , Binkley N , Krueger D , Scerpella TA . A focused evaluation of lumbar spine trabecular bone score in the first year post-menarche . Arch Osteoporos . 2017 ; 12 ( 1 ): 90 . Google Scholar Crossref Search ADS PubMed 39. Donaldson AA , Feldman HA , O’Donnell JM , Gopalakrishnan G , Gordon CM . Spinal bone texture assessed by trabecular bone score in adolescent girls with anorexia nervosa . J Clin Endocrinol Metab . 2015 ; 100 ( 9 ): 3436 – 3442 . Google Scholar Crossref Search ADS PubMed 40. Nour MA , Burt LA , Perry RJ , Stephure DK , Hanley DA , Boyd SK . Impact of growth hormone on adult bone quality in Turner Syndrome: a HR-pQCT study . Calcified Tissue Int. 2016 ; 98 ( 1 ): 49 – 59 . Google Scholar Crossref Search ADS 41. Freriks K , Timmermans J , Beerendonk CC , Verhaak CM , Netea-Maier RT , Otten BJ , Braat DD , Smeets DF , Kunst DH , Hermus AR , Timmers HJ . Standardized multidisciplinary evaluation yields significant previously undiagnosed morbidity in adult women with Turner syndrome . J Clin Endocrinol Metab . 2011 ; 96 ( 9 ): E1517 – E1526 . Google Scholar Crossref Search ADS PubMed 42. Vincent AJ , Nguyen HH , Ranasinha S , Vollenhoven B . Increased detection of co-morbidities with evaluation at a dedicated adult Turner syndrome clinic . Climacteric . 2017 ; 20 ( 5 ): 442 – 447 . Google Scholar Crossref Search ADS PubMed Copyright © 2018 Endocrine Society http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Clinical Endocrinology and Metabolism Oxford University Press

A Cross-Sectional and Longitudinal Analysis of Trabecular Bone Score in Adults With Turner Syndrome

Loading next page...
 
/lp/oxford-university-press/a-cross-sectional-and-longitudinal-analysis-of-trabecular-bone-score-KLsIDyOzjP
Publisher
Oxford University Press
Copyright
Copyright © 2018 Endocrine Society
ISSN
0021-972X
eISSN
1945-7197
D.O.I.
10.1210/jc.2018-00854
Publisher site
See Article on Publisher Site

Abstract

Abstract Context Turner syndrome (TS) is associated with short stature, gonadal failure, and fractures. Spinal trabecular bone score (TBS) is a novel bone imaging modality that has not been evaluated in TS. Objective To evaluate TBS in TS and its association with bone mineral density (BMD), prevalent fracture, and risk factors. Design and Setting Longitudinal study of TS from a single tertiary hospital between 2006 and 2017. Patients or Other Participants Fifty-eight subjects with TS aged 20 to 49 years who underwent dual-energy X-ray absorptiometry (DXA). Main Outcome Measures TBS, DXA parameters, and prevalent fractures were investigated. Results Normal, partially degraded, and degraded TBSs were observed in 39 (67%), 15 (26%), and four (7%) subjects, respectively. High rates of prescribed estrogen replacement therapy (ERT) with stable TBS and BMD were observed during follow-up. TBS was positively correlated with spine and femoral neck (FN) BMD and Z-scores (all P < 0.05) and negatively correlated with age (−0.004 per year; P = 0.014) and delay in ERT initiation in women with primary amenorrhea (−0.010 per year; P < 0.001). Fractures were present in 17 (31%) subjects. Low TBS had a significantly higher area under the receiver operator curve for predicting prevalent fracture than low bone mass at either the spine or FN (P < 0.05). Subjects with no history of fracture were more likely to have a normal TBS (P = 0.023). Conclusions BMD and TBS can be preserved with early initiation and continued use of ERT. TBS may provide additional fracture risk prediction to standard DXA parameters in TS and needs to be validated in larger prospective studies. Turner syndrome (TS) affects one in 2000 live-born girls and results from complete or partial X chromosome monosomy (1). Key features of TS include gonadal failure and short stature, whereas osteoporosis and an increased risk of fractures are associated complications. Osteoporosis has historically been associated with TS since initial reports of skeletal hypomineralization on radiographs of patients with TS (2). Subsequent studies noted an increased risk of fractures in TS, the relative risk ranging from 1.16 to 2.16 and most commonly affecting the upper limb, as well as low bone mineral density (BMD) as measured by dual-energy X-ray absorptiometry (DXA) (3–5). Gonadal failure is a major contributor to fracture risk (6), as fracture risk in patients with TS and adequate estrogen replacement therapy (ERT) is similar to that in control populations (7). Exogenous growth hormone therapy (GHT) is now standard treatment in TS for short stature and may also have beneficial effects on skeletal health (8, 9). The assessment of osteoporosis and fracture risk in women with TS is challenging. Bone density is routinely quantified by DXA, which provides a two-dimensional measure of areal BMD; however, it does not assess bone microarchitecture and can underestimate BMD in individuals with short stature and smaller bone size (10). Studies that adjusted for height or bone size have demonstrated a reduction in BMD in some TS cohorts (6, 11–13) but not others (8, 14, 15). As such, BMD alone may not be an accurate predictor of skeletal fragility and fracture risk in TS, particularly when short stature is prevalent. Although bone density is highly correlated with bone strength, it explains only 60% to 80% of the variance in bone (16). Information on other determinants of bone strength, such as bone geometry and microarchitecture, is not provided on routine BMD measurements. Peripheral quantitative computed tomography (pQCT) and high-resolution pQCT are noninvasive measures of bone microarchitecture that are not widely available for clinical use. The few TS studies using these modalities have shown cortical deficits in the forearm (12, 17–19), whereas compromised trabecular microarchitecture at the radius and tibia was identified in a study using high-resolution pQCT (20). Spinal trabecular bone score (TBS) has recently emerged as an indirect measure of trabecular microarchitecture at the spine (21). This novel software calculates a gray-level textural index derived from routine lumbar spine DXA images and does not require additional radiation exposure. A low TBS is associated with major osteoporotic fractures (defined as vertebral, hip, humeral, and wrist) in postmenopausal women, with similar findings in older men and other conditions associated with secondary osteoporosis (21). Given that estrogen deficiency is associated with fractures at sites rich in trabecular bone and a correlation between TBS and height has not been reported in the literature (22), TBS is an attractive bone imaging tool in TS populations. To date, no studies have investigated TBS in women with TS. Thus, in this cross-sectional and longitudinal study of an Australian adult TS cohort, we aimed to (1) assess spinal TBS and its relationship with clinical and densitometric parameters, (2) investigate longitudinal changes and clinical predictors of TBS and bone density, and (3) explore the capacity of TBS to predict fractures. Subjects and methods Patients This cross-sectional and longitudinal study involved women aged 20 to 50 years who had TS and underwent one or more DXA scans between the years 2006 and 2017 at Monash Medical Centre, Clayton, Victoria, Australia. Participants were identified from the Monash Medical Centre Adult TS Clinic, in addition to a comprehensive search of our institution’s Bone Density Department Database for all women documented as having hypogonadism and aged 20 to 50 years. Subsequent review of medical records confirmed 60 women with a history of TS. Two women taking osteoporosis-specific drugs were excluded, and none of the women had a history of oral glucocorticoid therapy. The study was approved by the Monash Health Human Research Ethics Committee (reference number 07062A). Informed consent was not required because this was a retrospective observational study, and DXA scans were performed as part of routine clinical practice. Clinical data Data regarding karyotype, menstrual history, medication use, and comorbidities were verified from hospital medical records and the Adult TS Clinic database. Data on age at ERT initiation and current use of ERT at the time of DXA scanning were collected, as well as any history of GHT use. Karyotype reports were available for 50 women and were coded as either “45XO” or “other X chromosome abnormalities.” A history of all documented fractures was established from medical records or by radiographically proven fractures from the hospital radiology database. Age at the time of fracture and the skeletal site were recorded. Measurement of bone density and TBS DXA scans were conducted on a single Lunar Prodigy device (GE Healthcare, Piscataway, NJ) at the Monash Medical Centre. For each woman, baseline and serial DXA scans during the study period were included. All women with TS had BMD (g/cm2) measured at the lumbar spine (L1-L4) and femoral neck (FN), and BMD Z-scores at each skeletal site were calculated. Low bone mass was defined as a Z-score ≤−2.0 as per the International Society for Clinical Densitometry guidelines for young premenopausal women (23). TBS measurements were obtained from DXA spinal images using TBS iNsight Software (version 3.0.2.0; Medimaps, Geneva, Switzerland). On the day of the DXA scan, each participant’s height and weight were measured with a wall-mounted stadiometer and electronic scale, respectively. The coefficients of variation for TBS and BMD using a lumbar spine phantom measured daily on the GE Lunar Prodigy were 0.94% and 0.92%, respectively. Statistical analysis The software program IBM SPSS Statistics v23 (IBM Corp, Armonk, NY) was used to conduct all statistical analyses, except when Stata v14.2 statistical software (StataCorp, College Station, TX) was used to test for significant differences between area under the receiver operator curve (AUROC) using pairwise comparison. A P value <0.05 was considered statistically significant. Baseline data Data from the baseline DXA scan of 58 women with TS were expressed as frequencies and median (range) values because continuous variables were not normally distributed. The relationship between baseline TBS, clinical variables, and densitometric variables was examined using univariate and age-adjusted multivariate linear regression analyses. Longitudinal data Women who had serial DXA scans were included in a longitudinal analysis. Longitudinal changes in bone parameters (TBS, spine BMD, spine Z-score, FN BMD, and FN Z-score) were assessed using linear mixed model analysis, where time was the number of years since the baseline DXA scan. All models were analyzed as random intercepts with fixed or random effects, and different candidate covariance models were considered, with the smallest Akaike information criterion used to choose the most appropriate model. The age- and body mass index (BMI)‒adjusted mean change per year (95% CI) for each bone parameter was calculated. The relationships between bone parameters and clinical predictor variables (age, karyotype, current ERT use, age at ERT initiation, and history of GHT) in the longitudinal data set were also assessed using linear mixed model analysis with adjustments for age and BMI. The model assessing the relationship between age and TBS was also further adjusted for potential confounders, including bone density parameters, karyotype, current ERT use, age at ERT initiation, history of GHT, smoking history, and secondary osteoporosis risk factors. The estimated change in bone outcome variable (95% CI) per unit change in clinical predictor variable is presented. Fracture analysis All fractures regardless of precipitating trauma or skeletal site were analyzed. Baseline clinical and densitometric data between fracture and nonfracture groups were compared using the Mann-Whitney U test for continuous variables and χ2 or Fisher’s exact test for categorical variables where appropriate. Because there was a significant difference in height between fracture and nonfracture groups, logistic regression was used to assess height-adjusted differences in bone parameters between groups. AUROC analysis was applied to determine the diagnostic value of age, TBS, spine Z-score, FN Z-score, and composite models to predict all fractures. TBS, spine Z-score, and FN Z-score were analyzed as dichotomous variables using established thresholds for abnormal values: TBS <1.35; spine Z-score ≤−2.0; and FN Z-score ≤−2.0. AUROC value and 95% CI are presented, and differences between AUROCs was tested using pairwise comparison. Results Baseline data Fifty-eight women with TS (median age, 28.5 years; range, 20 to 49 years) were included in this study. Their baseline characteristics, spinal TBS, and DXA-derived areal BMD and Z-scores of the spine and FN are described in Table 1. Table 1. Relationship Between Baseline TBS and Anthropometric, Clinical, and Bone Parameters Using Linear Regression in Women With TS Baseline Parameters n All Unadjusted Age Adjusted R2a B P Value R2a B P Value TBS 58 1.40 (1.10–1.69) — — — — — — Age, y 58 28.5 (20.0–49.0) 0.097 −0.004 0.017 — — — Ethnicity, Asian/white 58 6 (10.3)/52 (89.7) 0.003 0.020 0.708 0.098 0.008 0.882 Height, cm 58 148.0 (134.5–170.0) 0.008 −0.001 0.500 0.132 −0.003 0.142 Weight, kg 58 60.2 (29.4–121.8) 0.029 0.001 0.204 0.117 0.001 0.273 BMI, kg/m2 58 25.5 (13.8–49.5) 0.059 0.004 0.065 0.168 0.004 0.035 Karyotype, 45XO 50 15 (30.0) — 0.005 0.896 — 0.001 0.978 Age at TS diagnosis, y 58 11 (0–46) 0.048 −0.002 0.099 0.120 −0.002 0.241 Primary amenorrhea 58 47 (81.0) 0.003 −0.017 0.689 0.107 −0.031 0.435 Age at ERT initiation,b y 44 15 (11–30) 0.024 −0.002 0.263 0.190 −0.012 0.023 Current ERTc 51 44 (86.3) — 0.110 0.029 — 0.067 0.192 Prior GHT 58 35 (60.3) — 0.036 0.274 — 0.017 0.611 Current or ex-smoker 55 8 (14.5) — −0.011 0.812 — 0.027 0.579 Secondary osteoporosisd 51 9 (17.6) — 0.026 0.587 — 0.003 0.955 Hearing impairment 41 41 (80.4) — 0.011 0.768 — 0.020 0.589 Spine BMD, g/cm2 58 1.081 (0.681–1.411) 0.420 0.466 <0.001 0.500 0.457 <0.001 Spine Z-score 58 −1.0 (−3.9 to 1.9) 0.363 0.056 <0.001 0.459 0.056 <0.001 FN BMD, g/cm2 58 0.859 (0.566–1.123) 0.293 0.514 <0.001 0.345 0.478 <0.001 FN Z-score 58 −0.9 (−3.1 to 1.2) 0.199 0.056 <0.001 0.304 0.057 <0.001 Baseline Parameters n All Unadjusted Age Adjusted R2a B P Value R2a B P Value TBS 58 1.40 (1.10–1.69) — — — — — — Age, y 58 28.5 (20.0–49.0) 0.097 −0.004 0.017 — — — Ethnicity, Asian/white 58 6 (10.3)/52 (89.7) 0.003 0.020 0.708 0.098 0.008 0.882 Height, cm 58 148.0 (134.5–170.0) 0.008 −0.001 0.500 0.132 −0.003 0.142 Weight, kg 58 60.2 (29.4–121.8) 0.029 0.001 0.204 0.117 0.001 0.273 BMI, kg/m2 58 25.5 (13.8–49.5) 0.059 0.004 0.065 0.168 0.004 0.035 Karyotype, 45XO 50 15 (30.0) — 0.005 0.896 — 0.001 0.978 Age at TS diagnosis, y 58 11 (0–46) 0.048 −0.002 0.099 0.120 −0.002 0.241 Primary amenorrhea 58 47 (81.0) 0.003 −0.017 0.689 0.107 −0.031 0.435 Age at ERT initiation,b y 44 15 (11–30) 0.024 −0.002 0.263 0.190 −0.012 0.023 Current ERTc 51 44 (86.3) — 0.110 0.029 — 0.067 0.192 Prior GHT 58 35 (60.3) — 0.036 0.274 — 0.017 0.611 Current or ex-smoker 55 8 (14.5) — −0.011 0.812 — 0.027 0.579 Secondary osteoporosisd 51 9 (17.6) — 0.026 0.587 — 0.003 0.955 Hearing impairment 41 41 (80.4) — 0.011 0.768 — 0.020 0.589 Spine BMD, g/cm2 58 1.081 (0.681–1.411) 0.420 0.466 <0.001 0.500 0.457 <0.001 Spine Z-score 58 −1.0 (−3.9 to 1.9) 0.363 0.056 <0.001 0.459 0.056 <0.001 FN BMD, g/cm2 58 0.859 (0.566–1.123) 0.293 0.514 <0.001 0.345 0.478 <0.001 FN Z-score 58 −0.9 (−3.1 to 1.2) 0.199 0.056 <0.001 0.304 0.057 <0.001 Data are expressed as median (range) or n (%). Boldface indicates P values that are statistically significant. a R2 reported only for continuous variables. b In those with primary amenorrhea. c At time of DXA scan in those with amenorrhea (two patients with normal menses not included). d Including celiac disease (n = 3), inflammatory bowel disease (n = 2), hyperthyroidism (n = 2), primary hyperparathyroidism (n = 1), and type 1 diabetes mellitus (n = 1). View Large Table 1. Relationship Between Baseline TBS and Anthropometric, Clinical, and Bone Parameters Using Linear Regression in Women With TS Baseline Parameters n All Unadjusted Age Adjusted R2a B P Value R2a B P Value TBS 58 1.40 (1.10–1.69) — — — — — — Age, y 58 28.5 (20.0–49.0) 0.097 −0.004 0.017 — — — Ethnicity, Asian/white 58 6 (10.3)/52 (89.7) 0.003 0.020 0.708 0.098 0.008 0.882 Height, cm 58 148.0 (134.5–170.0) 0.008 −0.001 0.500 0.132 −0.003 0.142 Weight, kg 58 60.2 (29.4–121.8) 0.029 0.001 0.204 0.117 0.001 0.273 BMI, kg/m2 58 25.5 (13.8–49.5) 0.059 0.004 0.065 0.168 0.004 0.035 Karyotype, 45XO 50 15 (30.0) — 0.005 0.896 — 0.001 0.978 Age at TS diagnosis, y 58 11 (0–46) 0.048 −0.002 0.099 0.120 −0.002 0.241 Primary amenorrhea 58 47 (81.0) 0.003 −0.017 0.689 0.107 −0.031 0.435 Age at ERT initiation,b y 44 15 (11–30) 0.024 −0.002 0.263 0.190 −0.012 0.023 Current ERTc 51 44 (86.3) — 0.110 0.029 — 0.067 0.192 Prior GHT 58 35 (60.3) — 0.036 0.274 — 0.017 0.611 Current or ex-smoker 55 8 (14.5) — −0.011 0.812 — 0.027 0.579 Secondary osteoporosisd 51 9 (17.6) — 0.026 0.587 — 0.003 0.955 Hearing impairment 41 41 (80.4) — 0.011 0.768 — 0.020 0.589 Spine BMD, g/cm2 58 1.081 (0.681–1.411) 0.420 0.466 <0.001 0.500 0.457 <0.001 Spine Z-score 58 −1.0 (−3.9 to 1.9) 0.363 0.056 <0.001 0.459 0.056 <0.001 FN BMD, g/cm2 58 0.859 (0.566–1.123) 0.293 0.514 <0.001 0.345 0.478 <0.001 FN Z-score 58 −0.9 (−3.1 to 1.2) 0.199 0.056 <0.001 0.304 0.057 <0.001 Baseline Parameters n All Unadjusted Age Adjusted R2a B P Value R2a B P Value TBS 58 1.40 (1.10–1.69) — — — — — — Age, y 58 28.5 (20.0–49.0) 0.097 −0.004 0.017 — — — Ethnicity, Asian/white 58 6 (10.3)/52 (89.7) 0.003 0.020 0.708 0.098 0.008 0.882 Height, cm 58 148.0 (134.5–170.0) 0.008 −0.001 0.500 0.132 −0.003 0.142 Weight, kg 58 60.2 (29.4–121.8) 0.029 0.001 0.204 0.117 0.001 0.273 BMI, kg/m2 58 25.5 (13.8–49.5) 0.059 0.004 0.065 0.168 0.004 0.035 Karyotype, 45XO 50 15 (30.0) — 0.005 0.896 — 0.001 0.978 Age at TS diagnosis, y 58 11 (0–46) 0.048 −0.002 0.099 0.120 −0.002 0.241 Primary amenorrhea 58 47 (81.0) 0.003 −0.017 0.689 0.107 −0.031 0.435 Age at ERT initiation,b y 44 15 (11–30) 0.024 −0.002 0.263 0.190 −0.012 0.023 Current ERTc 51 44 (86.3) — 0.110 0.029 — 0.067 0.192 Prior GHT 58 35 (60.3) — 0.036 0.274 — 0.017 0.611 Current or ex-smoker 55 8 (14.5) — −0.011 0.812 — 0.027 0.579 Secondary osteoporosisd 51 9 (17.6) — 0.026 0.587 — 0.003 0.955 Hearing impairment 41 41 (80.4) — 0.011 0.768 — 0.020 0.589 Spine BMD, g/cm2 58 1.081 (0.681–1.411) 0.420 0.466 <0.001 0.500 0.457 <0.001 Spine Z-score 58 −1.0 (−3.9 to 1.9) 0.363 0.056 <0.001 0.459 0.056 <0.001 FN BMD, g/cm2 58 0.859 (0.566–1.123) 0.293 0.514 <0.001 0.345 0.478 <0.001 FN Z-score 58 −0.9 (−3.1 to 1.2) 0.199 0.056 <0.001 0.304 0.057 <0.001 Data are expressed as median (range) or n (%). Boldface indicates P values that are statistically significant. a R2 reported only for continuous variables. b In those with primary amenorrhea. c At time of DXA scan in those with amenorrhea (two patients with normal menses not included). d Including celiac disease (n = 3), inflammatory bowel disease (n = 2), hyperthyroidism (n = 2), primary hyperparathyroidism (n = 1), and type 1 diabetes mellitus (n = 1). View Large Using established adult norms for TBS (22), 39 women with TS (67%) had normal values (TBS ≥1.35); 15 women (26%) had partially degraded spinal microarchitecture (TBS 1.20 to 1.35); and four women (7%) had degraded microarchitecture (TBS ≤1.20). Low bone mass (Z-score ≤−2.0) at the spine was seen in 15 women (26%), whereas five women (9%) had low bone mass at the FN. The unadjusted and age-adjusted linear relationships between baseline spinal TBS and clinical, anthropometric, and bone density parameters are described in Table 1. Spinal TBS was significantly associated with spine BMD, spine Z-scores, FN BMD, and FN Z-scores in unadjusted and age-adjusted models. After adjustment for age, TBS was associated with BMI and inversely associated with age at ERT initiation. Longitudinal data Longitudinal data were available for 33 women with TS. Median duration of follow-up was 4 years (range, 1 to 10 years), and 48.5% of women had follow-up data available over at least 5 years. Over the follow-up period, 25 women with TS continued ERT; two women who were not receiving therapy at baseline commenced ERT; two women were never receiving ERT; and ERT status was unknown in four women. Longitudinal data were not available for the remaining 25 women because they were either lost to follow-up or did not have a second routine DXA scan at the time of study analysis. According to linear mixed model analysis of the longitudinal data, TBS was strongly associated with spine BMD, spine Z-score, FN BMD, and FN Z-score. The estimated changes (95% CI) in TBS per unit change in spine BMD, spine Z-score, FN BMD, and FN Z-score were 0.397 (0.223, 0.572); 0.049 (0.029, 0.070); 0.333 (0.099, 0.567); and 0.044 (0.017, 0.072), respectively (all age- and BMI-adjusted P < 0.01). During follow-up, no significant longitudinal changes per year in TBS, spine BMD, spine Z-score, FN BMD, and FN Z-score were observed (Table 2). Table 2. Longitudinal Changes in Bone Outcome Measurements Bone Measurements n Δ/y 95% CI P Value a TBS 33 0.003 (−0.003, 0.009) 0.353 Spine BMD, g/cm2 33 0.005 (−0.002, 0.013) 0.166 Spine Z-score 33 0.054 (−0.010, 0.118) 0.096 FN BMD, g/cm2 33 0.005 (−0.001, 0.011) 0.114 FN Z-score 33 0.028 (−0.021, 0.076) 0.256 Bone Measurements n Δ/y 95% CI P Value a TBS 33 0.003 (−0.003, 0.009) 0.353 Spine BMD, g/cm2 33 0.005 (−0.002, 0.013) 0.166 Spine Z-score 33 0.054 (−0.010, 0.118) 0.096 FN BMD, g/cm2 33 0.005 (−0.001, 0.011) 0.114 FN Z-score 33 0.028 (−0.021, 0.076) 0.256 Δ/y = estimated change in bone measurements per year of follow-up. a Adjusted for age and BMI. View Large Table 2. Longitudinal Changes in Bone Outcome Measurements Bone Measurements n Δ/y 95% CI P Value a TBS 33 0.003 (−0.003, 0.009) 0.353 Spine BMD, g/cm2 33 0.005 (−0.002, 0.013) 0.166 Spine Z-score 33 0.054 (−0.010, 0.118) 0.096 FN BMD, g/cm2 33 0.005 (−0.001, 0.011) 0.114 FN Z-score 33 0.028 (−0.021, 0.076) 0.256 Bone Measurements n Δ/y 95% CI P Value a TBS 33 0.003 (−0.003, 0.009) 0.353 Spine BMD, g/cm2 33 0.005 (−0.002, 0.013) 0.166 Spine Z-score 33 0.054 (−0.010, 0.118) 0.096 FN BMD, g/cm2 33 0.005 (−0.001, 0.011) 0.114 FN Z-score 33 0.028 (−0.021, 0.076) 0.256 Δ/y = estimated change in bone measurements per year of follow-up. a Adjusted for age and BMI. View Large TBS was inversely associated with age in this cohort [−0.004 per year (−0.007, −0.001); BMI-adjusted P = 0.014]. This remained statistically significant after adjusting separately for spine BMD, spine Z-score, and FN Z-score (all P < 0.05), but not for FN BMD (P = 0.087). Furthermore, after separate adjustment for clinical parameters, such as karyotype, current ERT, age at ERT initiation, GHT, smoking history, and secondary osteoporosis risk factors, TBS remained inversely associated with age (all P < 0.05). A decline in FN BMD with increasing age was also observed [−0.003 per year (−0.005, −0.001); BMI-adjusted P = 0.008], but no significant relationship with age was seen with spine BMD. The relationship between bone parameters and other clinical predictor variables is shown in Table 3. A delay in ERT initiation was associated with lower TBS, spine BMD, spine Z-score, FN BMD, and FN Z-score (all P < 0.05). Prior GHT was associated with a higher spine BMD, spine Z-score, FN BMD, and FN Z-score (all P < 0.05) but was not associated with TBS. An association between TBS and BMI was not observed in this longitudinal analysis. Table 3. Relationship Between Bone Outcome Measurements and Clinical Predictor Variables Using Multivariate Linear Mixed Model Analysis Bone Measurements Predictor Variable n Δ 95% CI P Value a TBS Karyotype, 45XO 29 −0.023 (−0.115, 0.068) 0.608 Current ERT 30 0.079 (−0.013, 0.171) 0.092 Age at ERT initiation,b y 24 −0.010 (−0.011, −0.010) <0.001 Prior GHT 33 0.024 (−0.067, 0.115) 0.589 Spine BMD, mg/cm2 Karyotype, 45XO 29 0.078 (−0.059, 0.215) 0.255 Current ERT 30 0.002 (0.028, 0.362) 0.955 Age of ERT initiation,b y 24 −0.021 (−0.039, −0.003) 0.024 Prior GHT 33 0.139 (0.018, 0.261) 0.027 Spine Z-score Karyotype, 45XO 29 0.740 (−0.404, 1.884) 0.195 Current ERT 30 0.010 (−0.677, 0.698) 0.975 Age at ERT initiation,b y 24 −1.183 (−1.875, −0.490) 0.001 Prior GHT 33 1.074 (0.196, 1.952) 0.018 FN BMD, mg/cm2 Karyotype, 45XO 29 0.065 (−0.050, 0.180) 0.256 Current ERT 30 −0.013 (−0.059, 0.032) 0.557 Age at ERT initiation,b y 24 −0.020 (−0.034, −0.007) 0.005 Prior GHT 33 0.114 (0.027, 0.202) 0.012 FN Z-score Karyotype, 45XO 29 0.803 (−0.233, 1.839) 0.124 Current ERT 30 −0.083 (−0.480, 0.315) 0.678 Age at ERT initiation,b y 24 −0.170 (−0.285, −0.054) 0.006 Prior GHT 33 0.947 (0.195, 1.700) 0.015 Bone Measurements Predictor Variable n Δ 95% CI P Value a TBS Karyotype, 45XO 29 −0.023 (−0.115, 0.068) 0.608 Current ERT 30 0.079 (−0.013, 0.171) 0.092 Age at ERT initiation,b y 24 −0.010 (−0.011, −0.010) <0.001 Prior GHT 33 0.024 (−0.067, 0.115) 0.589 Spine BMD, mg/cm2 Karyotype, 45XO 29 0.078 (−0.059, 0.215) 0.255 Current ERT 30 0.002 (0.028, 0.362) 0.955 Age of ERT initiation,b y 24 −0.021 (−0.039, −0.003) 0.024 Prior GHT 33 0.139 (0.018, 0.261) 0.027 Spine Z-score Karyotype, 45XO 29 0.740 (−0.404, 1.884) 0.195 Current ERT 30 0.010 (−0.677, 0.698) 0.975 Age at ERT initiation,b y 24 −1.183 (−1.875, −0.490) 0.001 Prior GHT 33 1.074 (0.196, 1.952) 0.018 FN BMD, mg/cm2 Karyotype, 45XO 29 0.065 (−0.050, 0.180) 0.256 Current ERT 30 −0.013 (−0.059, 0.032) 0.557 Age at ERT initiation,b y 24 −0.020 (−0.034, −0.007) 0.005 Prior GHT 33 0.114 (0.027, 0.202) 0.012 FN Z-score Karyotype, 45XO 29 0.803 (−0.233, 1.839) 0.124 Current ERT 30 −0.083 (−0.480, 0.315) 0.678 Age at ERT initiation,b y 24 −0.170 (−0.285, −0.054) 0.006 Prior GHT 33 0.947 (0.195, 1.700) 0.015 Δ = estimated change in bone measurements per unit change in predictor variable. Boldface indicates P values that are statistically significant. a P value adjusted for age and BMI. b In those with primary amenorrhea. View Large Table 3. Relationship Between Bone Outcome Measurements and Clinical Predictor Variables Using Multivariate Linear Mixed Model Analysis Bone Measurements Predictor Variable n Δ 95% CI P Value a TBS Karyotype, 45XO 29 −0.023 (−0.115, 0.068) 0.608 Current ERT 30 0.079 (−0.013, 0.171) 0.092 Age at ERT initiation,b y 24 −0.010 (−0.011, −0.010) <0.001 Prior GHT 33 0.024 (−0.067, 0.115) 0.589 Spine BMD, mg/cm2 Karyotype, 45XO 29 0.078 (−0.059, 0.215) 0.255 Current ERT 30 0.002 (0.028, 0.362) 0.955 Age of ERT initiation,b y 24 −0.021 (−0.039, −0.003) 0.024 Prior GHT 33 0.139 (0.018, 0.261) 0.027 Spine Z-score Karyotype, 45XO 29 0.740 (−0.404, 1.884) 0.195 Current ERT 30 0.010 (−0.677, 0.698) 0.975 Age at ERT initiation,b y 24 −1.183 (−1.875, −0.490) 0.001 Prior GHT 33 1.074 (0.196, 1.952) 0.018 FN BMD, mg/cm2 Karyotype, 45XO 29 0.065 (−0.050, 0.180) 0.256 Current ERT 30 −0.013 (−0.059, 0.032) 0.557 Age at ERT initiation,b y 24 −0.020 (−0.034, −0.007) 0.005 Prior GHT 33 0.114 (0.027, 0.202) 0.012 FN Z-score Karyotype, 45XO 29 0.803 (−0.233, 1.839) 0.124 Current ERT 30 −0.083 (−0.480, 0.315) 0.678 Age at ERT initiation,b y 24 −0.170 (−0.285, −0.054) 0.006 Prior GHT 33 0.947 (0.195, 1.700) 0.015 Bone Measurements Predictor Variable n Δ 95% CI P Value a TBS Karyotype, 45XO 29 −0.023 (−0.115, 0.068) 0.608 Current ERT 30 0.079 (−0.013, 0.171) 0.092 Age at ERT initiation,b y 24 −0.010 (−0.011, −0.010) <0.001 Prior GHT 33 0.024 (−0.067, 0.115) 0.589 Spine BMD, mg/cm2 Karyotype, 45XO 29 0.078 (−0.059, 0.215) 0.255 Current ERT 30 0.002 (0.028, 0.362) 0.955 Age of ERT initiation,b y 24 −0.021 (−0.039, −0.003) 0.024 Prior GHT 33 0.139 (0.018, 0.261) 0.027 Spine Z-score Karyotype, 45XO 29 0.740 (−0.404, 1.884) 0.195 Current ERT 30 0.010 (−0.677, 0.698) 0.975 Age at ERT initiation,b y 24 −1.183 (−1.875, −0.490) 0.001 Prior GHT 33 1.074 (0.196, 1.952) 0.018 FN BMD, mg/cm2 Karyotype, 45XO 29 0.065 (−0.050, 0.180) 0.256 Current ERT 30 −0.013 (−0.059, 0.032) 0.557 Age at ERT initiation,b y 24 −0.020 (−0.034, −0.007) 0.005 Prior GHT 33 0.114 (0.027, 0.202) 0.012 FN Z-score Karyotype, 45XO 29 0.803 (−0.233, 1.839) 0.124 Current ERT 30 −0.083 (−0.480, 0.315) 0.678 Age at ERT initiation,b y 24 −0.170 (−0.285, −0.054) 0.006 Prior GHT 33 0.947 (0.195, 1.700) 0.015 Δ = estimated change in bone measurements per unit change in predictor variable. Boldface indicates P values that are statistically significant. a P value adjusted for age and BMI. b In those with primary amenorrhea. View Large Fracture analysis Fracture history was available for 54 women with TS. Twenty-two fractures occurred in 17 of 54 women (31%), and nine of 54 women (16%) sustained a fracture after 20 years of age. Most fractures occurred in the upper limb (52.6%), followed by the lower limb (26.3%), vertebra (10.5%), clavicle (5.3%), and sternum (5.3%), with 64% occurring in predominantly trabecular bone (eight fractures at the end of long bones, four fractures in the short bones of the hands and feet, and two vertebral fractures). Baseline clinical characteristics and bone density data between fracture and nonfracture groups are described in Table 4. The fracture and nonfracture groups were similar except for height. No differences in spine BMD, spine Z-score, FN BMD, or FN Z-score were observed between groups even after adjusting for height. Table 4. Baseline Clinical Characteristics and Bone Parameters of Women With TS in Fracture and Nonfracture Groups Clinical Parameter n Fracture (n = 17) Nonfracture (n = 37) P Value Age, y 54 30.0 (20–47) 24.0 (20.0–49.0) 0.155 Ethnicity, Asian/white 54 1 (1.9)/16 (29.6) 5 (9.3)/32 (59.3) 0.407 Height, cm 54 151.0 (134.5–170.0) 147.2 (134.5–162.5) 0.041 Weight, kg 54 67.7 (29.4–121.8) 58.5 (39.8–110.8) 0.896 BMI, kg/m2 54 25.4 (13.8–39.0) 25.6 (18.4–49.5) 0.682 Karyotype, X monosomy 46 3 (6.5) 9 (19.6) 0.607 Age at TS diagnosis, y 54 12.0 (0.0–46.0) 11.0 (0.0–39.0) 0.581 Primary amenorrhea 54 13 (24.1) 30 (55.6) 0.696 Age at ERT initiation,a y 40 16.0 (12–46) 16.0 (0.0–39.0) 0.497 Current ERTb 50 12 (24.0) 31 (62.0) 0.124 Prior GHT 54 8 (14.8) 23 (42.6) 0.297 Current or ex-smoker 54 3 (5.6) 5 (9.3) 0.491 Secondary osteoporosis risk factorsc 49 4 (8.2) 5 (10.2) 0.322 Hearing impairment 47 9 (19.1) 28 (59.6) 0.272 TBS 54 1.34 (1.10–1.69) 1.42 (1.10–1.64) 0.101 Spine BMD, g/cm2 54 1.049 (0.721–1.411) 1.104 (0.681–1.343) 0.662 Spine Z-score 54 −0.9 (−3.9 to 1.9) −1.0 (−3.5 to 1.3) 0.823 FN BMD, g/cm2 54 0.833 (0.651–1.094) 0.911 (0.566–1.123) 0.635 FN Z-score 54 −0.8 (−2.1 to 1.2) −0.8 (−3.1 to 1.2) 0.716 Clinical Parameter n Fracture (n = 17) Nonfracture (n = 37) P Value Age, y 54 30.0 (20–47) 24.0 (20.0–49.0) 0.155 Ethnicity, Asian/white 54 1 (1.9)/16 (29.6) 5 (9.3)/32 (59.3) 0.407 Height, cm 54 151.0 (134.5–170.0) 147.2 (134.5–162.5) 0.041 Weight, kg 54 67.7 (29.4–121.8) 58.5 (39.8–110.8) 0.896 BMI, kg/m2 54 25.4 (13.8–39.0) 25.6 (18.4–49.5) 0.682 Karyotype, X monosomy 46 3 (6.5) 9 (19.6) 0.607 Age at TS diagnosis, y 54 12.0 (0.0–46.0) 11.0 (0.0–39.0) 0.581 Primary amenorrhea 54 13 (24.1) 30 (55.6) 0.696 Age at ERT initiation,a y 40 16.0 (12–46) 16.0 (0.0–39.0) 0.497 Current ERTb 50 12 (24.0) 31 (62.0) 0.124 Prior GHT 54 8 (14.8) 23 (42.6) 0.297 Current or ex-smoker 54 3 (5.6) 5 (9.3) 0.491 Secondary osteoporosis risk factorsc 49 4 (8.2) 5 (10.2) 0.322 Hearing impairment 47 9 (19.1) 28 (59.6) 0.272 TBS 54 1.34 (1.10–1.69) 1.42 (1.10–1.64) 0.101 Spine BMD, g/cm2 54 1.049 (0.721–1.411) 1.104 (0.681–1.343) 0.662 Spine Z-score 54 −0.9 (−3.9 to 1.9) −1.0 (−3.5 to 1.3) 0.823 FN BMD, g/cm2 54 0.833 (0.651–1.094) 0.911 (0.566–1.123) 0.635 FN Z-score 54 −0.8 (−2.1 to 1.2) −0.8 (−3.1 to 1.2) 0.716 Data are expressed as median (range) or n (%). Boldface indicates P values that are statistically significant. a In those with primary amenorrhea. b At time of DXA scan in those with amenorrhea (two patients with normal menses not included). c Including celiac disease (n = 3), inflammatory bowel disease (n = 2), hyperthyroidism (n = 2), primary hyperparathyroidism (n = 1), and type 1 diabetes mellitus (n = 1). View Large Table 4. Baseline Clinical Characteristics and Bone Parameters of Women With TS in Fracture and Nonfracture Groups Clinical Parameter n Fracture (n = 17) Nonfracture (n = 37) P Value Age, y 54 30.0 (20–47) 24.0 (20.0–49.0) 0.155 Ethnicity, Asian/white 54 1 (1.9)/16 (29.6) 5 (9.3)/32 (59.3) 0.407 Height, cm 54 151.0 (134.5–170.0) 147.2 (134.5–162.5) 0.041 Weight, kg 54 67.7 (29.4–121.8) 58.5 (39.8–110.8) 0.896 BMI, kg/m2 54 25.4 (13.8–39.0) 25.6 (18.4–49.5) 0.682 Karyotype, X monosomy 46 3 (6.5) 9 (19.6) 0.607 Age at TS diagnosis, y 54 12.0 (0.0–46.0) 11.0 (0.0–39.0) 0.581 Primary amenorrhea 54 13 (24.1) 30 (55.6) 0.696 Age at ERT initiation,a y 40 16.0 (12–46) 16.0 (0.0–39.0) 0.497 Current ERTb 50 12 (24.0) 31 (62.0) 0.124 Prior GHT 54 8 (14.8) 23 (42.6) 0.297 Current or ex-smoker 54 3 (5.6) 5 (9.3) 0.491 Secondary osteoporosis risk factorsc 49 4 (8.2) 5 (10.2) 0.322 Hearing impairment 47 9 (19.1) 28 (59.6) 0.272 TBS 54 1.34 (1.10–1.69) 1.42 (1.10–1.64) 0.101 Spine BMD, g/cm2 54 1.049 (0.721–1.411) 1.104 (0.681–1.343) 0.662 Spine Z-score 54 −0.9 (−3.9 to 1.9) −1.0 (−3.5 to 1.3) 0.823 FN BMD, g/cm2 54 0.833 (0.651–1.094) 0.911 (0.566–1.123) 0.635 FN Z-score 54 −0.8 (−2.1 to 1.2) −0.8 (−3.1 to 1.2) 0.716 Clinical Parameter n Fracture (n = 17) Nonfracture (n = 37) P Value Age, y 54 30.0 (20–47) 24.0 (20.0–49.0) 0.155 Ethnicity, Asian/white 54 1 (1.9)/16 (29.6) 5 (9.3)/32 (59.3) 0.407 Height, cm 54 151.0 (134.5–170.0) 147.2 (134.5–162.5) 0.041 Weight, kg 54 67.7 (29.4–121.8) 58.5 (39.8–110.8) 0.896 BMI, kg/m2 54 25.4 (13.8–39.0) 25.6 (18.4–49.5) 0.682 Karyotype, X monosomy 46 3 (6.5) 9 (19.6) 0.607 Age at TS diagnosis, y 54 12.0 (0.0–46.0) 11.0 (0.0–39.0) 0.581 Primary amenorrhea 54 13 (24.1) 30 (55.6) 0.696 Age at ERT initiation,a y 40 16.0 (12–46) 16.0 (0.0–39.0) 0.497 Current ERTb 50 12 (24.0) 31 (62.0) 0.124 Prior GHT 54 8 (14.8) 23 (42.6) 0.297 Current or ex-smoker 54 3 (5.6) 5 (9.3) 0.491 Secondary osteoporosis risk factorsc 49 4 (8.2) 5 (10.2) 0.322 Hearing impairment 47 9 (19.1) 28 (59.6) 0.272 TBS 54 1.34 (1.10–1.69) 1.42 (1.10–1.64) 0.101 Spine BMD, g/cm2 54 1.049 (0.721–1.411) 1.104 (0.681–1.343) 0.662 Spine Z-score 54 −0.9 (−3.9 to 1.9) −1.0 (−3.5 to 1.3) 0.823 FN BMD, g/cm2 54 0.833 (0.651–1.094) 0.911 (0.566–1.123) 0.635 FN Z-score 54 −0.8 (−2.1 to 1.2) −0.8 (−3.1 to 1.2) 0.716 Data are expressed as median (range) or n (%). Boldface indicates P values that are statistically significant. a In those with primary amenorrhea. b At time of DXA scan in those with amenorrhea (two patients with normal menses not included). c Including celiac disease (n = 3), inflammatory bowel disease (n = 2), hyperthyroidism (n = 2), primary hyperparathyroidism (n = 1), and type 1 diabetes mellitus (n = 1). View Large At baseline, prevalent fractures within the normal, partially degraded, and degraded TBS groups were recorded in eight of 36 women (22%), six of 14 women (43%), and three of four women (75%), respectively. In the degraded TBS group, two women recorded fractures at the wrist, and the third subject recorded a fracture involving the small bones of the hand. A statistically significant difference in fracture prevalence between normal and degraded TBS groups was found (height-adjusted P = 0.023) (Fig. 1), and this remained significant even after adjustments for BMD and Z-scores at the spine and FN, respectively (all P < 0.05). However, no difference in fracture prevalence was seen between normal and low bone mass groups at the spine and FN. Figure 1. View largeDownload slide Percentage of women with TS who fracture per bone outcome measurement. TBS: normal TBS ≥1.35; partially degraded TBS 1.20 to 1.35; and degraded TBS ≤1.20. Spine and FN Z-score: normal Z-score >−2.0; low Z-score ≤−2.0. Figure 1. View largeDownload slide Percentage of women with TS who fracture per bone outcome measurement. TBS: normal TBS ≥1.35; partially degraded TBS 1.20 to 1.35; and degraded TBS ≤1.20. Spine and FN Z-score: normal Z-score >−2.0; low Z-score ≤−2.0. The diagnostic value of spinal TBS in predicting prevalent fractures was assessed using AUROC analysis and was compared with AUROC for age, spine Z-score, and FN Z-score (Table 5). Although TBS alone was not predictive of prevalent fractures, TBS had a significantly higher AUROC than spine Z-score (0.643 vs 0.483; P = 0.013) and FN Z-score (0.643 vs 0.489; P = 0.033). The inclusion of TBS in a model containing age was predictive of prevalent fractures (P = 0.024) (Table 5). Table 5. AUROC Analysis for Fracture Outcomes n AUROC (95% CI) P Value All fractures  Age 58 0.621 (0.463, 0.779) 0.157  TBSa,b 58 0.643 (0.502, 0.784) 0.094  Spine Z-scorea 58 0.483 (0.316, 0.649) 0.116  FN Z-scoreb 58 0.489 (0.323, 0.655) 0.896 Composite models  Age + TBS 58 0.692 (0.542, 0.843) 0.024  Age + Spine Z-score 58 0.634 (0.484, 0.785) 0.116  Age + FN Z-score 58 0.624 (0.467, 0.781) 0.186 n AUROC (95% CI) P Value All fractures  Age 58 0.621 (0.463, 0.779) 0.157  TBSa,b 58 0.643 (0.502, 0.784) 0.094  Spine Z-scorea 58 0.483 (0.316, 0.649) 0.116  FN Z-scoreb 58 0.489 (0.323, 0.655) 0.896 Composite models  Age + TBS 58 0.692 (0.542, 0.843) 0.024  Age + Spine Z-score 58 0.634 (0.484, 0.785) 0.116  Age + FN Z-score 58 0.624 (0.467, 0.781) 0.186 Thresholds used in analysis: spine Z-score ≤−2.0, FN Z-score ≤−2.0, TBS <1.35. Boldface indicates P values that are statistically significant. a TBS had a significantly higher AUROC than spine Z-score (P = 0.0130). b TBS had a significantly higher AUROC than FN Z-score (P = 0.0327). View Large Table 5. AUROC Analysis for Fracture Outcomes n AUROC (95% CI) P Value All fractures  Age 58 0.621 (0.463, 0.779) 0.157  TBSa,b 58 0.643 (0.502, 0.784) 0.094  Spine Z-scorea 58 0.483 (0.316, 0.649) 0.116  FN Z-scoreb 58 0.489 (0.323, 0.655) 0.896 Composite models  Age + TBS 58 0.692 (0.542, 0.843) 0.024  Age + Spine Z-score 58 0.634 (0.484, 0.785) 0.116  Age + FN Z-score 58 0.624 (0.467, 0.781) 0.186 n AUROC (95% CI) P Value All fractures  Age 58 0.621 (0.463, 0.779) 0.157  TBSa,b 58 0.643 (0.502, 0.784) 0.094  Spine Z-scorea 58 0.483 (0.316, 0.649) 0.116  FN Z-scoreb 58 0.489 (0.323, 0.655) 0.896 Composite models  Age + TBS 58 0.692 (0.542, 0.843) 0.024  Age + Spine Z-score 58 0.634 (0.484, 0.785) 0.116  Age + FN Z-score 58 0.624 (0.467, 0.781) 0.186 Thresholds used in analysis: spine Z-score ≤−2.0, FN Z-score ≤−2.0, TBS <1.35. Boldface indicates P values that are statistically significant. a TBS had a significantly higher AUROC than spine Z-score (P = 0.0130). b TBS had a significantly higher AUROC than FN Z-score (P = 0.0327). View Large Discussion This study investigated the potential role of spinal TBS in the evaluation of skeletal health in an adult TS cohort. An age-related decline in TBS was observed in both cross-sectional and longitudinal analyses. During follow-up of up to 10 years, spinal TBS, spine BMD, spine Z-score, FN BMD, and FN Z-score remained stable in this TS cohort. Spinal TBS was positively associated with lumbar spine and FN BMD and Z-scores, and was adversely influenced by a delay in ERT initiation. Spine and FN bone parameters were also negatively associated with age at ERT initiation, whereas GHT appeared to have a protective effect. In this pilot study, TBS was a better predictor of prevalent fractures than low bone mass at the spine or FN. Spinal TBS is a textural parameter that provides an indirect measure of spinal microarchitecture that is independently associated with history of fracture and incidence of fracture and is complementary to BMD in estimating fracture risk in men and women (21). Most clinical studies have investigated TBS in older cohorts, and limited literature has reviewed changes in TBS in younger women. An age-related decline in TBS in healthy premenopausal women has been described in a small number of studies (24–28), whereas one North American study reported stable TBS values in women aged 30 to 45 years followed by a steady decline with age thereafter (29). The magnitude of the age-related decline in TBS described in a study of Thai women aged 30 to 50 years was 0.006 per year (27), whereas Iki et al. (28) reported a decline of 0.0013 per year in a Japanese cohort aged 20 to 40 years. The age-related decline in TBS of 0.004 per year seen in our TS cohort is not dissimilar to the decline described by Sritara et al. (27) and Iki et al. (28) and suggests that changes in TBS with age may not be accelerated in women with TS. In the longitudinal subgroup analysis, DXA bone parameters remained stable. This is possibly explained by the high rates of ERT (86%) use over the follow-up period, and it is reassuring that bone microarchitecture and density can be maintained in women with continued ERT exposure. This is consistent with published longitudinal studies in adult women with TS. Cleemann et al. (30) followed up 54 women with TS (aged 43.0 ± 9.95 years) over 6 years and found that BMD could be maintained in adults with TS by encouraging adequate ERT, a healthy lifestyle, and a high intake of calcium and vitamin D. Suganuma et al. (31) also showed that initiation of ERT in adult women with TS maintained bone density over 7 years of follow-up. Recently, Soucek et al. (7) showed that fractures were not increased in a cohort of women with TS receiving ERT. The observation that TBS in adulthood is adversely affected by delay in ERT initiation in women with TS is an important finding, although not unexpected. We also found in our longitudinal analysis that late pubertal induction detrimentally affected spine and FN bone density, a finding that is consistent with our previously published cross-sectional work showing that a delay in commencing ERT in primary amenorrhea was associated with lower BMD (6). Similar findings in bone density have been seen in other TS cohorts (32, 33), but the effect on microarchitecture has not been described. Gonadal failure manifesting as primary amenorrhea affects up to 85% of women with TS (34). Estrogen is a key hormonal regulator of bone health, with an important role in bone mass accrual during skeletal growth, skeletal homeostasis in adulthood, and accelerated bone loss during menopause (35). As such, estrogen deficiency is a well-described risk factor for osteoporosis in TS, and current professional TS guidelines recommend early pubertal induction with ERT by the age of 13 years in those with primary amenorrhea (36). Our finding supports current guidelines of early ERT initiation for adequate attainment of normal bone microarchitecture and peak bone mass. Limited literature has examined the effect of puberty on spinal trabecular microarchitecture. Two published studies of healthy pediatric female populations showed that pubertal development was associated with an increase in TBS. Shawwa et al. (37) reported a steady increase in TBS with age in a Lebanese cohort aged 10 to 17 years, with a higher mean TBS level in postmenarchal girls than in premenarchal girls (1.389 vs 1.293; P < 0.001). Dowthwaite et al. (38) showed that normal adult TBS levels were attained within 1 year postmenarche, 2 or more years before expected peak bone mass attainment at the spine. Correlation between TBS and pubertal stage was also seen in a small cohort study of adolescents with anorexia nervosa (39). Combined with our findings, this suggests that spinal TBS may be a valuable marker of inadequate trabecular microarchitecture attainment in girls with TS, particularly in those with delayed menarche. The positive effect of prior GHT on bone density seen in our TS cohort is interesting and has been reported in some studies (8, 9) but not in others (11, 13, 31). An association between GHT and TBS was not seen in our study. The discrepancy between GHT-related BMD and TBS changes in our study may be due to GHT-related increases in height and bone size, with this larger bone translating to higher areal BMD, as opposed to direct effects of GHT on volumetric bone density and microarchitecture. To date, the literature on the effects of GHT on bone in TS has remained inconsistent. In a case-control study of 28 women with TS, Nour et al. (40) found an increase in height and overall bone size in GH-treated women with TS compared with a non‒GH-treated group, but no difference in BMD or bone microarchitecture as measured by high-resolution pQCT. In contrast, a more recent study by Soucek et al. (7) showed a positive association between the duration of GHT and cortical volumetric BMD. We found that low TBS had a higher AUROC for prevalent fracture than traditional DXA parameters and that TBS in a model with age was predictive of fractures. This study adds to the body of evidence regarding the relationship between low TBS and fragility fractures in adults, which has already been described in older populations and in other conditions with secondary osteoporosis, such as diabetes mellitus, primary hyperparathyroidism, and rheumatoid arthritis (21). Our finding is of particular relevance to TS populations because TBS is not associated with height, which is a pitfall in the use of DXA-derived areal BMD in women with TS and short stature. Furthermore, because TBS uses the same region of interest as DXA-derived spinal images, no additional testing or radiation exposure is required, and spinal images can be retrospectively analyzed for TBS values. As such, spinal TBS is an attractive clinical tool that may be easily integrated into the routine care of women with TS to improve fracture risk assessment. We found a statistically significant positive correlation between TBS and BMI in an age-adjusted regression analysis; however, this relationship was not seen in our longitudinal data analysis. There may be a relationship between body composition and TBS in subjects with TS, and future studies should explore this relationship further. Although several significant findings were elucidated, we are mindful of the limitations of this study, particularly the retrospective study design and relatively small sample size. We did not have a control group, although we used established adult norms to interpret TBS and BMD Z-scores. Reliance on medical records may have underestimated our fracture prevalence, and we could not analyze incident fracture risk. Further, we were unable to reliably differentiate minimal trauma fractures from traumatic fractures. We were unable to longitudinally assess adherence to ERT or GHT; however, we expect nonadherence to ERT to be common because the prevalence of nonadherence was previously reported to be 15% to 37% in TS cohorts (41, 42). It is possible that nonadherence attenuated the relationship between TBS and bone density. Because we found a relationship between a delay in ERT initiation and TBS, we postulate that nonadherence to ERT may be associated with lower TBS values. Finally, we could not consistently obtain detailed data on type of ERT or dose and duration of GHT and hence could not directly assess the dose effect of ERT or GHT on TBS and fractures. This study has a number of strengths that make it relevant to the TS literature. In particular, the use of a novel fracture assessment tool and the longitudinal follow-up of up to 10 years make our study one of the longest TS studies with bone densitometric data in the literature. The majority of women (96%) attended the multidisciplinary adult TS clinic at our institution and thus had standardized screening questionnaire forms at each appointment, ensuring consistency in the recording of data. Conclusion In conclusion, we assessed changes in TBS in women with TS, and our preliminary findings have important clinical implications for the management of bone health in women with the condition. First, our findings suggest that TBS provides additional information for the diagnosis and risk stratification of skeletal fragility in adult women with TS, information that is independent of areal BMD measurements. A larger prospective study is required to validate spinal TBS as a fracture prediction tool in TS and to replicate our finding that it is superior to BMD. Second, a delay in ERT initiation adversely influenced all bone outcome measures, including TBS. This highlights the importance of early ERT to optimize bone density and trabecular microarchitecture in adulthood. Future pediatric studies should further explore the direct relationship between spinal TBS and pubertal induction in TS and whether an optimal ERT dose exists to provide adequate TBS attainment. Abbreviations: Abbreviations: AUROC area under the receiver operator curve BMD bone mineral density BMI body mass index DXA dual-energy X-ray absorptiometry ERT estrogen replacement therapy FN femoral neck GHT growth hormone therapy pQCT peripheral quantitative computed tomography TBS trabecular bone score TS Turner syndrome Acknowledgments The authors wish to acknowledge the Monash Health Adult Turner Syndrome Clinic and the patients with TS. We thank Ms. Stella May Gwini for assistance with the biostatistical analysis. Financial Support: This work was supported by the Amgen-Osteoporosis Australia-Australian New Zealand Bone & Mineral Society Clinical Grant (to A.V.). P.W. is supported by the National Health and Medical Research Council of Australia Early Career Fellowship. Disclosure Summary: The authors have nothing to disclose. References 1. Nielsen J , Wohlert M . Chromosome abnormalities found among 34,910 newborn children: results from a 13-year incidence study in Arhus, Denmark . Hum Genet . 1991 ; 87 ( 1 ): 81 – 83 . Google Scholar Crossref Search ADS PubMed 2. Shore RM , Chesney RW , Mazess RB , Rose PG , Bargman GJ . Skeletal demineralization in Turner’s syndrome . Calcif Tissue Int . 1982 ; 34 ( 1 ): 519 – 522 . Google Scholar Crossref Search ADS PubMed 3. Gravholt CH , Juul S , Naeraa RW , Hansen J . Morbidity in Turner syndrome . J Clin Epidemiol . 1998 ; 51 ( 2 ): 147 – 158 . Google Scholar Crossref Search ADS PubMed 4. Gravholt CH , Vestergaard P , Hermann AP , Mosekilde L , Brixen K , Christiansen JS . Increased fracture rates in Turner’s syndrome: a nationwide questionnaire survey . Clin Endocrinol (Oxf) . 2003 ; 59 ( 1 ): 89 – 96 . Google Scholar Crossref Search ADS PubMed 5. Davies MC , Gulekli B , Jacobs HS . Osteoporosis in Turner’s syndrome and other forms of primary amenorrhoea . Clin Endocrinol (Oxf) . 1995 ; 43 ( 6 ): 741 – 746 . Google Scholar Crossref Search ADS PubMed 6. Nguyen HH , Wong P , Strauss BJ , Jones G , Ebeling PR , Milat F , Vincent A . Delay in estrogen commencement is associated with lower bone mineral density in Turner syndrome . Climacteric . 2017 ; 20 ( 5 ): 436 – 441 . Google Scholar Crossref Search ADS PubMed 7. Soucek O , Schönau E , Lebl J , Willnecker J , Hlavka Z , Sumnik Z . 6-Year follow-up of fracture incidence and volumetric bone mineral density development in girls with Turner syndrome . J Clin Endocrinol Metab . 2018 ; 103 ( 3 ): 1188 – 1197 . Google Scholar Crossref Search ADS PubMed 8. Sas TC , de Muinck Keizer-Schrama SM , Stijnen T , van Teunenbroek A , van Leeuwen WJ , Asarfi A , van Rijn RR , Drop SL ; Dutch Advisory Group on Growth Hormone . Bone mineral density assessed by phalangeal radiographic absorptiometry before and during long-term growth hormone treatment in girls with Turner’s syndrome participating in a randomized dose-response study . Pediatr Res . 2001 ; 50 ( 3 ): 417 – 422 . Google Scholar Crossref Search ADS PubMed 9. Bertelloni S , Cinquanta L , Baroncelli GI , Simi P , Rossi S , Saggese G . Volumetric bone mineral density in young women with Turner’s syndrome treated with estrogens or estrogens plus growth hormone . Horm Res . 2000 ; 53 ( 2 ): 72 – 76 . Google Scholar PubMed 10. Prentice A , Parsons TJ , Cole TJ . Uncritical use of bone mineral density in absorptiometry may lead to size-related artifacts in the identification of bone mineral determinants . Am J Clin Nutr . 1994 ; 60 ( 6 ): 837 – 842 . Google Scholar Crossref Search ADS PubMed 11. Nadeem M , Roche EF . Bone mineral density in Turner’s syndrome and the influence of pubertal development . Acta Paediatr . 2014 ; 103 ( 1 ): e38 – e42 . Google Scholar Crossref Search ADS PubMed 12. Holroyd CR , Davies JH , Taylor P , Jameson K , Rivett C , Cooper C , Dennison EM . Reduced cortical bone density with normal trabecular bone density in girls with Turner syndrome . Osteoporos Int . 2010 ; 21 ( 12 ): 2093 – 2099 . Google Scholar Crossref Search ADS PubMed 13. Högler W , Briody J , Moore B , Garnett S , Lu PW , Cowell CT . Importance of estrogen on bone health in Turner syndrome: a cross-sectional and longitudinal study using dual-energy X-ray absorptiometry . J Clin Endocrinol Metab . 2004 ; 89 ( 1 ): 193 – 199 . Google Scholar Crossref Search ADS PubMed 14. Ross JL , Long LM , Feuillan P , Cassorla F , Cutler GB Jr . Normal bone density of the wrist and spine and increased wrist fractures in girls with Turner’s syndrome . J Clin Endocrinol Metab . 1991 ; 73 ( 2 ): 355 – 359 . Google Scholar Crossref Search ADS PubMed 15. Shaw NJ , Rehan VK , Husain S , Marshall T , Smith CS . Bone mineral density in Turner’s syndrome: a longitudinal study . Clin Endocrinol (Oxf) . 1997 ; 47 ( 3 ): 367 – 370 . Google Scholar Crossref Search ADS PubMed 16. Kushida K . Bone mineral density measurements for diagnosis of osteoporosis and the risk assessment [in Japanese] . Clin Calcium . 2004 ; 14 ( 12 ): 11 – 17 . Google Scholar PubMed 17. Bechtold S , Rauch F , Noelle V , Donhauser S , Neu CM , Schoenau E , Schwarz HP . Musculoskeletal analyses of the forearm in young women with Turner syndrome: a study using peripheral quantitative computed tomography . J Clin Endocrinol Metab . 2001 ; 86 ( 12 ): 5819 – 5823 . Google Scholar Crossref Search ADS PubMed 18. Soucek O , Lebl J , Snajderova M , Kolouskova S , Rocek M , Hlavka Z , Cinek O , Rittweger J , Sumnik Z . Bone geometry and volumetric bone mineral density in girls with Turner syndrome of different pubertal stages . Clin Endocrinol (Oxf) . 2011 ; 74 ( 4 ): 445 – 452 . Google Scholar Crossref Search ADS PubMed 19. Soucek O , Zapletalova J , Zemkova D , Snajderova M , Novotna D , Hirschfeldova K , Plasilova I , Kolouskova S , Rocek M , Hlavka Z , Lebl J , Sumnik Z . Prepubertal girls with Turner syndrome and children with isolated SHOX deficiency have similar bone geometry at the radius . J Clin Endocrinol Metab . 2013 ; 98 ( 7 ): E1241 – E1247 . Google Scholar Crossref Search ADS PubMed 20. Hansen S , Brixen K , Gravholt CH . Compromised trabecular microarchitecture and lower finite element estimates of radius and tibia bone strength in adults with turner syndrome: a cross-sectional study using high-resolution-pQCT . J Bone Miner Res . 2012 ; 27 ( 8 ): 1794 – 1803 . Google Scholar Crossref Search ADS PubMed 21. Silva BC , Broy SB , Boutroy S , Schousboe JT , Shepherd JA , Leslie WD . Fracture risk prediction by non-BMD DXA measures: the 2015 ISCD official positions part 2: trabecular bone score . J Clin Densitom . 2015 ; 18 ( 3 ): 309 – 330 . Google Scholar Crossref Search ADS PubMed 22. Silva BC , Leslie WD , Resch H , Lamy O , Lesnyak O , Binkley N , McCloskey EV , Kanis JA , Bilezikian JP . Trabecular bone score: a noninvasive analytical method based upon the DXA image . J Bone Miner Res . 2014 ; 29 ( 3 ): 518 – 530 . Google Scholar Crossref Search ADS PubMed 23. International Society for Clinical Densitometry . Official ISCD positions, 2015—adult & pediatric. Available at: iscd.app.box.com/OP-ISCD-2015-Adult. Accessed 5 April 2018 . 24. Looker AC , Sarafrazi Isfahani N , Fan B , Shepherd JA . Trabecular bone scores and lumbar spine bone mineral density of US adults: comparison of relationships with demographic and body size variables . Osteoporos Int . 2016 ; 27 ( 8 ): 2467 – 2475 . Google Scholar Crossref Search ADS PubMed 25. Cheng P , Qi HM , Di WJ , Liu J , Yu J , Lv S , Shen YL , Zha JM , Cai JM , Lai B , Ding GX . Establishment of TBS reference plots and correlation between TBS and BMD in healthy mainland Chinese women . Arch Osteoporos . 2016 ; 11 ( 1 ): 5 . Google Scholar Crossref Search ADS PubMed 26. Bazzocchi A , Ponti F , Diano D , Amadori M , Albisinni U , Battista G , Guglielmi G . Trabecular bone score in healthy ageing . Br J Radiol . 2015 ; 88 ( 1052 ): 20140865 . Google Scholar Crossref Search ADS PubMed 27. Sritara C , Thakkinstian A , Ongphiphadhanakul B , Amnuaywattakorn S , Utamakul C , Akrawichien T , Vathesatogkit P , Sritara P . Age-adjusted dual X-ray absorptiometry-derived trabecular bone score curve for the lumbar spine in Thai females and males . J Clin Densitom . 2016 ; 19 ( 4 ): 494 – 501 . Google Scholar Crossref Search ADS PubMed 28. Iki M , Tamaki J , Sato Y , Winzenrieth R , Kagamimori S , Kagawa Y , Yoneshima H . Age-related normative values of trabecular bone score (TBS) for Japanese women: the Japanese Population-based Osteoporosis (JPOS) study . Osteoporos Int . 2015 ; 26 ( 1 ): 245 – 252 . Google Scholar Crossref Search ADS PubMed 29. Simonelli C , Leib E , Mossman N , Winzenrieth R , Hans D , McClung M . Creation of an age-adjusted, dual-energy x-ray absorptiometry-derived trabecular bone score curve for the lumbar spine in non-Hispanic US white women . J Clin Densitom . 2014 ; 17 ( 2 ): 314 – 319 . Google Scholar Crossref Search ADS PubMed 30. Cleemann L , Hjerrild BE , Lauridsen AL , Heickendorff L , Christiansen JS , Mosekilde L , Gravholt CH . Long-term hormone replacement therapy preserves bone mineral density in Turner syndrome . Eur J Endocrinol . 2009 ; 161 ( 2 ): 251 – 257 . Google Scholar Crossref Search ADS PubMed 31. Suganuma N , Furuhashi M , Hirooka T , Moriwaki T , Hasegawa Y , Mori O , Ogawa M . Bone mineral density in adult patients with Turner’s syndrome: analyses of the effectiveness of GH and ovarian steroid hormone replacement therapies . Endocr J . 2003 ; 50 ( 3 ): 263 – 269 . Google Scholar Crossref Search ADS PubMed 32. Kodama M , Komura H , Kodama T , Nishio Y , Kimura T . Estrogen therapy initiated at an early age increases bone mineral density in Turner syndrome patients . Endocr J . 2012 ; 59 ( 2 ): 153 – 159 . Google Scholar Crossref Search ADS PubMed 33. Mora S , Weber G , Guarneri MP , Nizzoli G , Pasolini D , Chiumello G . Effect of estrogen replacement therapy on bone mineral content in girls with Turner syndrome . Obstet Gynecol . 1992 ; 79 ( 5 Pt 1 ): 747 – 751 . Google Scholar PubMed 34. Pasquino AM , Passeri F , Pucarelli I , Segni M , Municchi G ; Italian Study Group for Turner’s Syndrome . Spontaneous pubertal development in Turner’s syndrome . J Clin Endocrinol Metab . 1997 ; 82 ( 6 ): 1810 – 1813 . Google Scholar PubMed 35. Manolagas SC , O’Brien CA , Almeida M . The role of estrogen and androgen receptors in bone health and disease . Nat Rev Endocrinol . 2013 ; 9 ( 12 ): 699 – 712 . Google Scholar Crossref Search ADS PubMed 36. Gravholt CH , Andersen NH , Conway GS , Dekkers OM , Geffner ME , Klein KO , Lin AE, Mauras N, Quigley CA, Rubin K, Sandberg DE, Sas TCJ, Silberbach M, Söderström-Anttila V, Stochholm K, van Alfen-van derVelden JA, Woelfle J, Backeljauw PF; International Turner Syndrome Consensus Group. Clinical practice guidelines for the care of girls and women with Turner syndrome: proceedings from the 2016 Cincinnati International Turner Syndrome Meeting. Eur J Endocrinol. 2017;177(3):G1–G70 . 37. Shawwa K , Arabi A , Nabulsi M , Maalouf J , Salamoun M , Choucair M , Hans D , El-Hajj Fuleihan G . Predictors of trabecular bone score in school children . Osteoporos Int . 2016 ; 27 ( 2 ): 703 – 710 . Google Scholar Crossref Search ADS PubMed 38. Dowthwaite JN , Winzenrieth R , Binkley N , Krueger D , Scerpella TA . A focused evaluation of lumbar spine trabecular bone score in the first year post-menarche . Arch Osteoporos . 2017 ; 12 ( 1 ): 90 . Google Scholar Crossref Search ADS PubMed 39. Donaldson AA , Feldman HA , O’Donnell JM , Gopalakrishnan G , Gordon CM . Spinal bone texture assessed by trabecular bone score in adolescent girls with anorexia nervosa . J Clin Endocrinol Metab . 2015 ; 100 ( 9 ): 3436 – 3442 . Google Scholar Crossref Search ADS PubMed 40. Nour MA , Burt LA , Perry RJ , Stephure DK , Hanley DA , Boyd SK . Impact of growth hormone on adult bone quality in Turner Syndrome: a HR-pQCT study . Calcified Tissue Int. 2016 ; 98 ( 1 ): 49 – 59 . Google Scholar Crossref Search ADS 41. Freriks K , Timmermans J , Beerendonk CC , Verhaak CM , Netea-Maier RT , Otten BJ , Braat DD , Smeets DF , Kunst DH , Hermus AR , Timmers HJ . Standardized multidisciplinary evaluation yields significant previously undiagnosed morbidity in adult women with Turner syndrome . J Clin Endocrinol Metab . 2011 ; 96 ( 9 ): E1517 – E1526 . Google Scholar Crossref Search ADS PubMed 42. Vincent AJ , Nguyen HH , Ranasinha S , Vollenhoven B . Increased detection of co-morbidities with evaluation at a dedicated adult Turner syndrome clinic . Climacteric . 2017 ; 20 ( 5 ): 442 – 447 . Google Scholar Crossref Search ADS PubMed Copyright © 2018 Endocrine Society

Journal

Journal of Clinical Endocrinology and MetabolismOxford University Press

Published: Oct 1, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off