OSTA as a screening tool to predict osteoporosis in Indian postmenopausal women - a nationwide study

A cross-sectional study comparing machine learning and logistic regression techniques for predicting osteoporosis in a group at high risk of cardiovascular disease among old adults

BACKGROUND

Osteoporosis has become a significant public health concern that necessitates the application of appropriate techniques to calculate disease risk. Traditional methods, such as logistic regression,have been widely used to identify risk factors and predict disease probability. However,with the advent of advanced statistics techniques,machine learning models offer promising alternatives for improving prediction accuracy. What's more, studies that use risk factors and prediction models for osteoporosis in high-risk groups for cardiovascular diseases are scarce. We aimed to explore the risk factors and disease probability of osteoporosis by comparing logistic regression with four machine learning models. By doing so,we seek to provide insights into the most effective methods for osteoporosis risk assessment and contribute to the development of tailored prevention strategies at high risk of cardiovascular disease among old adults.

METHODS

We carried out a cross-sectional investigation of a high-risk group in cardiovascular patients. A logistic regression model and four common machine learning methods,DT,RF,SVM,and XGBoost were implemented to create a prediction model using information from 211 participants who met the inclusion requirements. Metrics for calibration and discrimination were used to compare the models.

RESULTS

In total,211 patients were enrolled. The AUCs were 0.751 for the logistic regression model,0.72 for the SVM model,0.70 for the random forest model,0.697 for the model XGBoost,and 0.69 for the decision tree model. The logistic regression model outperforms other models for machine learning. According to the logistic regression model,there were nine predictors,including age,sex,glucose,TG (triglyceride),fracture history,stroke history,and CNV (copy number variation) nssv659422, and low-sodium salt. A well-calibrated result of 0.199 on the Brier scale. The findings of the internal validation demonstrated the high degree of repeatability of the prediction model employed in this study.

CONCLUSIONS

In this study, we discovered that when predicting osteoporosis,a number of machine learning techniques fell short of logistic regression. In a specific population, we have innovatively developed a risk prediction model for osteoporosis events that integrates genetic and environmental factors, is an effective tool for assessing osteoporosis risk and can serve as the basis for specialized intervention approaches.

Osteoporosis risk and its association with all-cause and cause-specific mortality among the elderly: a 16-year nationwide cohort study

BACKGROUND

Aged osteoporosis poses a significant threat to the well-being and longevity of older individuals, yet evidence regarding the relationship between osteoporosis risk and mortality among the elderly population in Asia remains unknown.

AIMS

Our study aimed to investigate associations between osteoporosis risk and all-cause mortality, as well as cause-specific mortality, among the Chinese elderly population.

METHODS

Pooled data from the Chinese Longitudinal Healthy Longevity Survey (CLHLS) conducted between 2002 and 2018 were utilized to analyze the associations between osteoporosis risk and all-cause, heart disease, cardiovascular disease (CVD), respiratory disease, and cancer mortality. Cox proportional hazards models were employed for this analysis. Osteoporosis risk was assessed using the Osteoporosis Self-Assessment Tool for Asians (OSTA). Restricted cubic spline (RCS) functions were applied to explore the nonlinear relationship between OSTA and mortality. The robustness of the Cox models was evaluated through internal verification, subgroup analyses, and sensitivity analyses.

RESULTS

A total of 12,711 elderly individuals aged ≥ 65 years were included for analysis at baseline. During a 16-year follow-up, 7,963 individuals in the cohort were identified as deceased. Compared to those with low osteoporosis risk, elderly individuals with high osteoporosis risk demonstrated a significantly elevated risk of all-cause, heart disease, CVD, respiratory disease and cancer mortality. The relationship between OSTA level and all-cause and cause-specific mortality exhibited a significant L-shaped pattern.

CONCLUSIONS

The risk of osteoporosis is independently associated with the prediction of mortality. The OSTA may serve as a suitable predictor for mortality related to osteoporosis among the Asian population.

Osteoporosis screening using X-ray assessment and osteoporosis self-assessment tool for Asians in hip surgery patients

OBJECTIVES

As many patients with osteoporosis remain undiagnosed, we aimed to develop a simple method to efficiently screen for osteoporosis using a combination of anteroposterior hip X-ray assessment and the Osteoporosis Self-Assessment Tool for Asians (OSTA), which is calculated as (body weight - age) × 0.2.

METHODS

One hundred Japanese women (age: 73 ± 11 years, body weight: 54.4 ± 11.1 kg) who underwent hip surgery, anteroposterior hip X-ray, and DXA were included. Based on the DXA results of the total proximal femur, 35 cases were diagnosed with osteoporosis. Fifteen orthopaedic surgeons visually inspected the hip X-ray images and scored the suspicion of osteoporosis on a scale of 1-4 (1: very unlikely, 4: very suspicious), which is referred to as "pred-score." In addition, OSTA was calculated as a continuous variable (OSTA score). Osteoporosis was screened using the pred-score and OSTA score, and both scores were analyzed using the receiver operating characteristic curves.

RESULTS

The area under the curves (AUCs) of the pred-score and OSTA score were 0.626-0.875 and 0.817 across surgeons, respectively. When both scores were used, the AUC for screening osteoporosis ranged from 0.821 to 0.915 across surgeons. Significant improvement from AUCs calculated with the pred-score or OSTA score was found in 11 surgeons (73.3%).

CONCLUSION

The combination of X-ray assessment and OSTA can be used to screen for osteoporosis and has the potential to be used as a new simple screening tool in daily clinical practice.

Current status and dilemmas of osteoporosis screening tools: A narrative review

OBJECTIVE

This review aims to explore the strengths and dilemmas of existing osteoporosis screening tools and suggest possible ways of optimization, in addition to exploring the potential of AI-integrated X-ray imaging in osteoporosis screening, especially its ability to improve accuracy and applicability to different populations. To break through the dilemma of low accessibility, poor clinical translation, complexity of use, and apparent limitations of screening results of existing osteoporosis screening tools.

DATA SOURCES

A comprehensive literature search was performed using PubMed, Web of Science, and CNKI databases. The search included articles published between 2000 and 2023, focusing on studies evaluating osteoporosis screening tools, Artificial intelligence applications in medical imaging, and implementing AI technologies in clinical settings.

STUDY SELECTION

The Osteoporosis Risk Assessment Tool for Asians (OSTA), the Simple Calculated Osteoporosis Risk Estimator (SCORE), age, body size, one or no estrogen ever (ABONE), and the Osteoporosis Risk Index (OSIRIS) are the six commonly used screening tools for osteoporosis that are discussed in this review. In addition, the performance of AI-integrated imaging systems is explored in light of relevant research advances in Artificial intelligence in osteoporosis screening. Studies of the use of these tools in different populations and their advantages and disadvantages were included in the selection criteria.

RESULTS

The results highlight that AI-integrated X-ray imaging technologies offer significant improvements over traditional osteoporosis screening tools. Artificial intelligence systems demonstrated higher accuracy by incorporating complex clinical data and providing personalized assessments for diverse populations. The studies showed that AI-driven imaging could enhance sensitivity and specificity, particularly in detecting early-stage bone density loss in patients with complex clinical profiles. The findings also suggest that Artificial intelligence technologies have the potential to be effectively applied in resource-limited settings through the use of mobile devices and remote diagnostics.

CONCLUSIONS

AI-integrated X-ray imaging technology significantly advances osteoporosis screening, offering more accurate and adaptable solutions than traditional tools. Its ability to incorporate complex clinical data and apply it across various demographic groups makes it particularly promising in diverse and resource-limited environments. Further research is needed to explore the full potential of AI in enhancing screening accessibility and effectiveness, particularly in underserved populations.

Combining OSTA and BMR to predict osteoporosis in Chinese population

INTRODUCTION

Osteoporosis is a debilitating bone disease that significantly contributes to disability and a loss of autonomy among older adults. This study aimed to characterize osteoporosis and explore the feasibility of combining OSTA and BMR for osteoporosis prediction.

METHODS

A cross-sectional study involving 1435 participants (1300 women and 135 men) was conducted. Spearman's correlation, simple linear regression analyses, and multiple linear regression models were utilized to investigate the association between OSTA, BMR, and bone mineral density (BMD). Furthermore, the efficacy of integrating OSTA with BMR for osteoporosis screening and prediction was assessed through receiver operating characteristic (ROC) curves.

RESULTS

In the total population, the sensitivity of combination variable W was 58.63%, and the specificity was 70.90%. When OSTA and BMR were employed separately to diagnose osteoporosis, the sensitivity was 47.70% and 55.34%, respectively, while the specificity was 63.80% and 69.80%, respectively.

CONCLUSIONS

The combined utilization of OSTA and BMR formula represents an effective screening method for osteoporosis.

Prevalence of osteoporosis and associated factors among people aged 50 years and older in the Madhesh province of Nepal: a community-based cross-sectional study

BACKGROUND

The high prevalence of osteoporosis has increased the economic burden on the health system globally. The burden of osteoporosis and its associated factors have not been adequately assessed in community settings in the Nepalese context thus far. Therefore, this study aimed to assess the prevalence of osteoporosis and its associated factors, lifestyle behaviors, and dietary calcium intake.

METHODS

A community-based cross-sectional study was conducted among 395 people aged 50 years and older in the Madhesh Province of Nepal between July 2022 and August 2023. The Osteoporosis Self-assessment Tools for Asians (OSTA) index was used to measure osteoporosis. A structured questionnaire was used to collect sociodemographic information, anthropometric data, lifestyle behavior, daily dietary calcium intake, and frequency of calcium-rich food consumption. A food frequency questionnaire and 24-hour recall methods were used to assess dietary intake. The chi-square test, binary logistic regression and Mann‒Whitney U test were applied to measure the association between predictors and the outcome of interest.

RESULTS

The prevalence of no risk, moderate risk and high risk of osteoporosis were 38.7%, 39%, and 22.3% respectively. The risk of osteoporosis was higher in females (aOR = 5.18, CI: 2.10-12.75, p < 0.001) and increased risk with advancing age (aOR = 32.49, CI: 14.02-75.28, p < 0.001). Similarly, underweight was associated with increased odds of having osteoporosis (aOR = 13.42, CI = 4.58-39.30, p < 0.001). The incidence of osteoporosis was strongly associated with daily calcium intake of 225 mg (100, 386).

CONCLUSION

This study revealed a high prevalence of osteoporosis among people aged 50 years and older due to the combined effect of being underweight and having inadequate calcium intake. Nutritional counselling services encourage people to consume sufficient calcium-rich food and adopt an appropriate lifestyle behaviours to maintain healthy body weight so that osteoporosis and osteoporotic fractures could be prevented. Further research can explore the impact of socioeconomic status and medical comorbidities on a large scale.

Assessing the utility of osteoporosis self-assessment tool for Asians in patients undergoing hip surgery

Objectives

Diagnosis and treatment of osteoporosis are instrumental in obtaining good outcomes of hip surgery. Measuring bone mineral density (BMD) using dual-energy X-ray absorptiometry (DXA) is the gold standard for diagnosing osteoporosis. However, due to limited access to DXA, there is a need for a screening tool to identify patients at a higher risk of osteoporosis. We analyzed the potential utility of the Osteoporosis Self-assessment Tool for Asians (OSTA) as a screening tool for osteoporosis.

Methods

A total of 1378 female patients who underwent hip surgery at 8 institutions were analyzed. For each patient, the BMD of the proximal femoral region was measured by DXA (DXA-BMD), and the correlation with OSTA score (as a continuous variable) was assessed. Receiver operating characteristic (ROC) curve analysis was performed to assess the ability of OSTA score to predict osteoporosis. Lastly, the OSTA score was truncated to yield an integer (OSTA index) to clarify the percentage of patients with osteoporosis for each index.

Results

DXA-BMD showed a strong correlation with OSTA (r = 0.683; P < 0.001). On ROC curve analysis, the optimal OSTA score cut-off value of -5.4 was associated with 73.8% sensitivity and 80.9% specificity for diagnosis of osteoporosis (area under the curve: 0.842). A decrease in the OSTA index by 1 unit was associated with a 7.3% increase in the probability of osteoporosis.

Conclusions

OSTA is a potentially useful tool for screening osteoporosis in patients undergoing hip surgery. Our findings may help identify high-risk patients who require further investigation using DXA.

The Significance of Combined OSTA, HU Value and VBQ Score in Osteoporosis Screening Before Spinal Surgery

OBJECTIVE

This study aimed to assess the utility of a combined assessment using the Osteoporosis Self-Assessment Tool for Asians (OSTA), Hounsfield unit (HU) value, and vertebral bone quality (VBQ) score for preoperative osteoporosis (OP) screening in patients scheduled for spinal surgery.

METHODS

This study encompassed 288 participants, including 128 males and 160 females. Patients were stratified into 2 groups: the OP group (T-score ≤ -2.5) and the non-OP group (T-score > -2.5), determined by dual-energy X-ray absorptiometry (DEXA). Binary logistic regression was used to construct a combined diagnostic model, and the receiver operating characteristic (ROC) curve evaluated the sensitivity, specificity, accuracy, positive predictive value (PPV), and negative predictive value (NPV) of these metrics individually or in combination to screen for OP.

RESULTS

Osteoporosis patients exhibited significantly lower OSTA and HU values in comparison to non-OP patients, while their VBQ scores were significantly higher (P < 0.001). The ROC curve analysis results indicated that within the male group, the combined diagnosis had a sensitivity of 93.8%, specificity of 82.3%, accuracy of 85.2%, PPV of 63.8%, and NPV of 97.5%. In the female group, the combined diagnosis had a sensitivity of 93.9%, specificity of 87.4%, accuracy of 90.0%, PPV of 83.6%, and NPV of 95.4%.

CONCLUSIONS

The combined use of OSTA, HU values, and VBQ scores in preoperative OP screening for spinal surgery demonstrates significantly higher accuracy and superior screening value compared to individual assessments. These results establish a robust scientific foundation for conducting preoperative OP screening in patients undergoing spinal surgery.

Can TBS reference values be a valid indicator for clinical vertebral fracture? A cross-sectional study

INTRODUCTION

Trabecular bone score (TBS) is partially independent of fracture risk. Reference values for TBS have not been established in official guidelines, and thus clinicians often have difficulty interpreting TBS results. This study aimed to investigate whether reference values for TBS could be a valid indicator for clinical vertebral fracture (CVF).

MATERIALS AND METHODS

This cross-sectional study involved 231 women with CVF and 563 women without CVF aged 60-90 years who underwent dual-energy X-ray absorptiometry during 2019-2023. They were divided into osteoporosis, osteopenia, and normal groups according to bone mineral density of the lumbar spine. Reference values for TBS were defined as low (≤ 1.23), intermediate (1.23-1.31), and high (≥ 1.31).

RESULTS

Among patients without anti-osteoporosis treatment (n = 476), the proportion with low TBS was 36.7% in the CVF group and 10.7% in the control group. The proportion with CVF was higher in the low TBS group than in the intermediate and high TBS groups, especially in the osteoporosis group (p < 0.001). The odds ratio for CVF was higher in the low TBS group than in the intermediate and high especially in patients with normal BMD and osteoporosis. The TBS cut-off values for incidence of CVF in the osteoporosis, osteopenia, and normal groups were 1.224, 1.319, and 1.322, respectively.

CONCLUSIONS

The reference value for low TBS (≤ 1.23) was useful as an indicator for CVF, especially in patients with osteoporosis. It is expected that reference values for TBS will be established in official guidelines in the future.

Cost-effectiveness of selective bone densitometry using the osteoporosis self-assessment tool for Asians in multi-ethnic Asian population

Early detection and treatment of osteoporosis can help to prevent debilitating fractures in the elderly. The osteoporosis self-assessment tool for Asians can be used as a screening tool to stratify patients for bone densitometry. It is most cost-effective for post-menopausal women aged 70 and males aged 75.

PURPOSE

To determine the cost-effectiveness of selective bone densitometry (SBD) using the Osteoporosis Self-Assessment Tool for Asians (OSTA) as a risk-stratifying tool for the three predominant races (Chinese, Malay and Indian) in Singapore.

METHODS

Decision analytical models were developed using a Markov model. Three scenarios were compared: no bone densitometry, SBD using the OSTA as a pre-screening tool and universal bone densitometry. Those diagnosed with osteoporosis were treated with five years of alendronate therapy. Data sources were from Singapore epidemiological studies, healthcare cost figures and published literature. Measurements include life years, quality-adjusted life years (QALYs), costs and incremental cost-effectiveness ratios (ICER).

RESULTS

Compared to no bone densitometry, SBD using the OSTA would cost between $40,679 and $73,909 per QALY gained for men aged 75-80 and $22,386 to $58,185 per QALY gained for post-menopausal women aged 70-80. Universal bone densitometry would cost $157,955 to $177,127 per QALY gained for men aged 75-80 and $40,179 to $66,112 per QALY gained for post-menopausal women aged 70 to 80 compared to SBD.

CONCLUSION

In general, osteoporosis screening was the most cost-effective for Malays and the least cost-effective for Indians. However, a general guideline should still be applied to the Singaporean population, as further explained later. Overall, the most cost-effective strategy for males would be using OSTA as a risk-stratifying tool at age 75. For post-menopausal women, SBD should be used for women aged 70, while universal bone densitometry should be used for women aged 75-80.