Development of a Screening Tool for Low Bone Mass in Asian Women

1
Osteoporosis Self-assessment Tool (OST) OSTA
for Asians
A simple clinical tool to identify women with
osteoporosis Professor J-Y Reginster
2
Background

• Cost and availability limit access to bone
  mineral density (BMD) in some communities
• A simple risk assessment tool could
• target BMD to high-risk women
• reduce the need for BMD measurements

3
Osteoporosis Self-assessment Tool for Asians
(OSTA)
Adapted from L. Koh, et al. A simple tool to
identify Asian women at increased risk of
osteoporosis. Osteoporos Int 12699-705, 2001.
4
Objective

• Develop and assess a simple tool for predicting
  osteoporosis (femoral neck BMD T lt -2.5) in
  postmenopausal Asian women, using risk factors
  obtained by questionnaire

5
Investigators

• The Osteoporosis Self-assessment Tool for Asia
  (OSTA) Research Group
• China Huang Qi Ren, Zhang Wei Bin, Huang Gong
  Yi, Lou Si Quan, Luo Xian Zheng, Liu Gui Lin
• South Korea In-Kwon Han, Ki-Hyun Park, Moo-Il
  Kang, Hyong-Moo Park, Hyoung-Woo Lee
• Taiwan Chun S Shih, Li H Chen, Keh-Sung Tsai,
  Yang Bor-Yau
• Hong Kong Annie Kung
• Philippines Tito P Torralba
• Thailand Rajata Rajatanavin
• Malaysia Siew-Pheng Chan
• Singapore Siok Bee Chionh, Tang Ching Lau,
  Paul Ho, Leonard Koh
• Belgium (WHO) Jean-Yves Reginster, Wafa Ben
  Sedrine
• Japan (validation sample) Saeko Fujiwara
• USA (MSD) Philip Ross, Larry Radican

6
Method

• Community-dwelling postmenopausal women
• 21 clinics in 8 Asian countries
• Consecutive subjects
• Routine check-up or follow-up of medical
  conditions
• 860 subjects completed questionnaires
• Study was performed under the auspices of the
  World Health Organization (WHO)

7
Recruitment Criteria

• Inclusion
• Postmenopausal
• Any race except Caucasian
• Hip anatomy suitable for BMD
• Willingness and ability to participate

• Exclusion
• History/evidence of metabolic bone disease
• Presence of cancer with known metastasis to bone
• Evidence of significant renal impairment
• One or both ovaries removed
• Both hips previously fractured or replaced
• History of bisphosphonate, fluoride or calcitonin
  use

8
Risk Factors Assessed in Questionnaire

• Ethnicity
• Weight
• Height (present/past)
• Physical activity
• Sunlight exposure
• Smoking
• Dairy / soybean intake
• Reproductive history

• Physical disability
• Rheumatoid arthritis
• Fracture history
• Family history of fracture
• Calcium supplements
• Postmenopausal estrogen
• Thyroid medication
• Corticosteroids

9
Results Study Population

• Demographics
• Chinese 59 Korean 18 Thai 11
• Filipino 9 Other 4
• Age (years) 62 (6)
• Weight (kg) 57 (9)
• Height (cm) 154 (6)
• Femoral neck BMD (g/cm2) 0.71 (0.14)
• 14 of women had T-scores lt -2.5

10
Multivariate Analysis

• Multivariable model - best fit
• 11 variables
• age
• weight
• estrogen
• thyroid medication
• any fracture since 45
• prior spine fracture
• Chinese
• Thai
• 3 countries (Malaysia, Hong Kong, Taiwan)

11
Simple Model (2 variables)Performs Well
Koh L, et al. Osteoporos Int 12699-705, 2001.
12
Calculation

• Model with only 2 variables
• (Weight - Age) x 0.2, drop decimal
• example Weight 59 kg, Age 72 yr
• (59-72) x 0.2 -2

Koh L, et al. Osteoporos Int 12699-705, 2001.
13
Risk Stratification
Koh L, et al. Osteoporos Int 12699-705, 2001.
14
Validation - JAPAN

• Adult Health Study, Hiroshima
• Postmenopausal women in existing cohort
• Ages 50 and older
• N 1123
• Sensitivity 98, specificity 29

Fujiwara S, et al. Curr Ther Res 62(8) 586-94,
2001.
15
Validation (Japan)
Development (Asian)
Risk level Low Medium High
Fujiwara S, et al. Curr Ther Res 62(8) 586-94,
2001.
16
Osteoporosis Self-assessment Tool for Asia (OSTA)
Weight (kg)
LOW RISK
Age (yr)
AT RISK measure BMD
HIGH RISK measure BMD treat
History of prior non-violent fracture consider
BMD measurement and treatment
17
Validation - KOREA

• Postmenopausal women
• Ages gt 55
• N 1101
• Sensitivity 87, specificity 67

Ben Sedrine W, et al. Arth Rheum 44(9
Suppl)S260, 2001.
18
Validation (KOREA)
Development (Asian)
Risk level Low Medium High
Fujiwara S, et al. Curr Ther Res 62(8) 586-94,
2001.
19
Validation - SINGAPORE

• 125 Postmenopausal women
• Mean 60 7.5 yr (range 50 - 89 yr)
• Sensitivity 94, specificity 64
• AUC 0.83
• 98 Men
• Mean 61 8.8 yr (range 50 - 88 yr)
• Sensitivity 50, specificity 78
• AUC 0.71

Koh L, et al. J Bone Miner Res 16(Suppl 1)S394
S396, 2001.
20
WOMEN
RESULTS - SINGAPORE
MEN
Risk level Low Medium High
Koh L, et al. J Bone Miner Res 16(Suppl 1)S394
S396, 2001.
21
Validation (SINGAPORE)
Risk level Low Medium High
Koh L, et al. J Bone Miner Res 16(Suppl 1)S394
S396, 2001.
22
Validation - CHINA

• 973 Postmenopausal women
• Clinic in Shanghai
• 91 women (9) had osteoporosis
• 557 (57) had osteopenia
• Sensitivity 92, specificity 54
• AUC 0.84

Ben Sedrine W, et al. Arth Rheum 44(9
Suppl)S258, 2001.
23
Validation (China)
Development (Asian)
Risk level Low Medium High
Ben Sedrine W, et al. Arth Rheum 44(9
Suppl)S258, 2001.
24
Summary - OSTA

• Age and weight alone performed well for
  identifying osteoporosis
• Consistent results among countries
• Including more variables did not substantially
  improve performance
• Using 3 categories may help to identify the
  highest and lowest risk groups
• BMD measurements are probably unnecessary in a
  substantial proportion of low risk women

25
Osteoporosis Self-assessment Tool (OST)
Performance in Caucasians
26
Validation - Caucasians

• Original SCORE population
• Postmenopausal women in the US
• Ages gt 45 (mean 61 yr)
• N 1102
• Sensitivity 88, specificity 52
• Adding other risk factors did not improve
  performance

Siris et al, J Bone Miner Res 16(Suppl 1)S341,
2001.
27
Validation - Caucasians

• Screening for Fracture Intervention Trial
• Postmenopausal women in the US
• Mean age 68 yr (SD 6)
• N 23,833
• Sensitivity 95, specificity 34
• Adding other risk factors did not improve
  performance

Hochberg et al, J Bone Miner Res 16(Suppl
1)S277, 2001.
28
SCORE population
FIT screen population
Risk level Low (gt1) Medium High (lt-3)
Hochberg et al, J Bone Miner Res 16(Suppl
1)S277, 2001.
Siris et al, J Bone Miner Res 16(Suppl 1)S341,
2001.
29
Osteoporosis Self-assessment Tool (OST)
Weight (kg)
LOW RISK
Age (yr)
AT RISK measure BMD
HIGH RISK measure BMD treat
History of prior non-violent fracture consider
BMD measurement and treatment
30
OST - CaucasiansSimple Rules
Age in yr minus weight in kg
31
SUMMARY

• Both OST and OSTA performed well
• Consistent results in different samples
• Using 3 categories may help to identify the
  highest and lowest risk groups
• BMD measurements are probably unnecessary in a
  substantial proportion of low risk women
• Category cut-points were slightly different in
  Caucasian versus Asian women