Age-related Reference Data for Cortical and Trabecular 3D-DXA Parameters: the SEIOMM-3D-DXA Project

The SEIOMM-3D-DXA project established age- and sex-specific reference curves for 3D-DXA cortical and trabecular bone parameters using data from 1,366 Spanish adults, demonstrating that compartment-specific analysis can reveal significant bone imbalances missed by traditional areal BMD measurements to improve osteoporosis management.

The Gist

The "Bone Blueprint" Project: Seeing Inside the Wall

Imagine your bones aren't just solid blocks of white stone, but complex buildings made of two very different materials: a spongy, honeycomb-like interior (trabecular bone) and a hard, thick outer shell (cortical bone).

For decades, doctors have used a standard X-ray test (called DXA) to check if these buildings are safe. However, this old test is like looking at a building from the outside and measuring its shadow. It tells you how "dense" the shadow looks, but it can't tell you if the inside is crumbling while the outside looks fine, or vice versa. This is why many people who look "healthy" on a standard test still end up breaking a hip.

The SEIOMM-3D-DXA Project is like hiring a team of architects to build a 3D hologram of the bone based on that same 2D shadow. This new technology, called 3D-DXA, uses a special computer program (3D-Shaper) to reconstruct the bone in 3D, allowing doctors to measure the "sponge" and the "shell" separately.

Here is what this new study did, explained simply:

1. Building the "Normal" Map

To know if a house is in bad shape, you need a map of what a "healthy" house looks like at every age.

• The Mission: The researchers gathered data from 1,366 healthy men and women across Spain. They scanned their hips to create a massive library of "normal" bone data.
• The Result: They created a set of reference curves (like a growth chart for height, but for bone density). Now, when a doctor scans a 60-year-old woman, they can instantly see: "Is her bone sponge as strong as it should be for a 60-year-old? Is her outer shell thick enough?"

2. The "Two-Speed" Decline

One of the most interesting discoveries is that the "sponge" and the "shell" don't age at the same speed, and they don't age the same way for men and women.

• The Analogy: Think of a house where the roof (cortical bone) and the insulation (trabecular bone) are wearing out at different rates.
• The Finding: In women, the "insulation" (sponge) starts getting thin around age 50 (menopause), while the "roof" (shell) starts weakening a bit later, around age 60. In men, both start slowing down later in life, around age 60.
• Why it matters: If you only look at the whole house (the old 2D test), you might miss that the insulation is rotting away while the roof is still strong. This study gives doctors the tools to spot that specific problem.

3. The "Mismatch" Surprise

The study found something startling: About half of the people tested had a "mismatch."

• The Metaphor: Imagine a car where the engine is brand new, but the tires are bald. Or a car with a rusted engine but brand new tires.
• The Reality: In 52% of women and 49% of men, the "sponge" inside was much weaker (or stronger) than the "shell" outside.
• The Problem: The old 2D test averages these two out. It's like saying, "The car is fine," because the good engine cancels out the bad tires. But in reality, the car is unsafe. This study proves that checking the parts separately is crucial for catching hidden risks.

4. New Traffic Lights for Doctors

The researchers didn't just find data; they created a new traffic light system for doctors to use with this new 3D technology.

• Green Zone: Your bone parts are normal for your age.
• Yellow Zone: Your bone parts are a bit weak (Osteopenia).
• Red Zone: Your bone parts are very weak (Osteoporosis).
• The Innovation: They created specific "Red Light" numbers for the sponge and the shell separately, so doctors know exactly which part of the bone needs help.

The Bottom Line

This study is like upgrading from a black-and-white photo of a building to a color-coded 3D blueprint.

It tells us that bones are more complex than we thought. By measuring the "sponge" and the "shell" separately, doctors can now:

1. Catch bone weakness earlier.
2. Understand why a specific person is at risk (is it the inside or the outside?).
3. Treat patients more precisely, rather than guessing based on a single average number.

This is a big step forward in preventing hip fractures and keeping our "bone buildings" standing tall for longer.

Technical Summary

1. Problem Statement

• Limitations of Current Standard: Dual-energy X-ray Absorptiometry (DXA) is the clinical gold standard for diagnosing osteoporosis, but it provides only a 2D areal bone mineral density (aBMD) measurement. This method lacks sensitivity because many fractures occur in patients with T-scores above -2.5 (osteopenia or normal range).
• Lack of Compartmental Data: Standard DXA cannot distinguish between trabecular (spongy) and cortical (compact) bone, which have different metabolic rates and responses to treatment.
• Need for 3D Solutions: While Quantitative Computed Tomography (QCT) offers 3D volumetric data, it is limited by high radiation exposure and lack of standardization. 3D-DXA algorithms (like 3D-Shaper) can derive volumetric parameters from standard 2D DXA scans, but they lack age- and sex-specific reference data and clinical thresholds necessary for widespread clinical adoption.

2. Methodology

• Study Design: A multicenter, cross-sectional, population-based observational study (SEIOMM-3D-DXA project) conducted in Spain between 2018 and 2021.
• Cohort: 1,366 healthy adults (1,015 women, 351 men) aged 20–90 years.
  • Exclusion Criteria: Osteoporosis medication, metabolic bone diseases, history of fragility fractures, early/surgical menopause, hypogonadism, and spinal surgeries.
• Data Acquisition:
  • Hip DXA scans were acquired using GE Healthcare iDXA or Prodigy scanners.
  • Standard aBMD and T-scores were calculated using the NHANES III Caucasian reference database.
• 3D Analysis:
  • Scans were processed using 3D-Shaper software (v2.14).
  • The software fits a 3D statistical shape and density model (derived from QCT) to the 2D DXA image.
  • Parameters Generated:
    • Trabecular volumetric BMD (vBMD): Measured in mg/cm³.
    • Cortical surface BMD (sBMD): Measured in mg/cm² (calculated as cortical vBMD × cortical thickness).
• Statistical Analysis:
  • Reference Curves: Generated using the LMS method (Cole and Green) to model skewness, median, and coefficient of variation across age groups (20–90 years).
  • Thresholds: Regression analysis was performed to correlate 3D parameters with standard aBMD T-scores (-1.0 and -2.5) to establish clinical thresholds.
  • Discrepancy Analysis: Z-score differences between trabecular and cortical compartments were calculated to identify inter-compartmental imbalances. Precision was validated using Least Significant Change (LSC) calculations.
  • Cross-Manufacturer Conversion: Linear regression equations were derived to convert GE Healthcare 3D-Shaper data to Hologic-equivalent values.

3. Key Contributions

• First Reference Database: Established the first sex- and manufacturer-specific age-related reference curves and clinical thresholds for 3D-Shaper trabecular vBMD and cortical sBMD.
• Clinical Thresholds: Defined specific cut-off values for "Normal," "Low," and "Very Low" density corresponding to standard aBMD T-scores of -1.0 and -2.5.
• Inter-Compartmental Analysis: Provided evidence that a significant portion of the population exhibits clinically significant discrepancies between cortical and trabecular bone health, which standard DXA misses.
• Software Integration: These reference data and thresholds are integrated into 3D-Shaper software version 2.14, which has regulatory clearance in the US, Europe, Asia, and LATAM.

4. Key Results

• Reference Curves Alignment: The aBMD curves derived from the Spanish cohort closely matched the NHANES III Caucasian reference data, validating the study's representativeness.
• Age-Related Decline:
  • Women: Showed a steeper decline in both cortical and trabecular density, particularly after menopause (starting ~50 years). Cortical sBMD loss averaged -0.94 mg/cm²/year after age 65; Trabecular vBMD loss averaged -1.73 mg/cm³/year.
  • Men: Decline started later (~60 years) and was less pronounced. Cortical sBMD loss averaged -0.74 mg/cm²/year; Trabecular vBMD loss averaged -1.21 mg/cm³/year.
• Established Thresholds (Table 1):
  • Cortical sBMD (mg/cm²):
    • Women: Normal/Low (-1.0) = 146; Low/Very Low (-2.5) = 114.
    • Men: Normal/Low (-1.0) = 158; Low/Very Low (-2.5) = 122.
  • Trabecular vBMD (mg/cm³):
    • Women: Normal/Low (-1.0) = 138; Low/Very Low (-2.5) = 84.
    • Men: Normal/Low (-1.0) = 156; Low/Very Low (-2.5) = 98.
• Compartmental Imbalance:
  • 52.0% of women and 48.7% of men exhibited a clinically significant discrepancy (Z-score difference > LSC of 0.397) between trabecular and cortical compartments.
  • This indicates that nearly half of the population has a specific weakness in either the cortical or trabecular bone that would be masked by a standard aBMD T-score.

5. Significance and Conclusion

• Improved Risk Stratification: By providing compartment-specific reference data, clinicians can better stratify patients who may be at risk despite having "normal" or "osteopenic" aBMD T-scores.
• Personalized Treatment: The ability to detect imbalances (e.g., high cortical but low trabecular density, or vice versa) allows for more targeted therapeutic decisions, particularly for patients with secondary osteoporosis or those on specific treatments.
• Standardization: This study bridges the gap between advanced 3D modeling and clinical practice by aligning new 3D-DXA parameters with the established NHANES III T-score framework.
• Future Directions: The authors note the need for further research into race/ethnicity-specific norms and the clinical impact of inter-compartmental imbalances on fracture prediction and treatment monitoring.

In summary, the SEIOMM-3D-DXA project successfully translates 3D bone modeling from a research tool into a clinically actionable framework, offering a more nuanced view of bone health than traditional 2D DXA.