MARPE Relapse Modeling:
Predictive Equation
From 200 Cases
A data-driven formula to quantify skeletal expansion stability, optimize retention protocols, and identify high-risk patients early.
TL;DR A MARPE relapse predictive equation derived from 200 cases quantifies skeletal expansion stability across patient age, bone density, and retention protocol. The model enables clinicians to forecast relapse magnitude, optimize treatment duration, and identify high-risk patients requiring extended retention or surgical support. Early application improves outcomes.
Miniscrew-assisted rapid palatal expansion continues to grow in clinical popularity, yet predicting individual relapse remains challenging for orthodontists planning long-term stability. Dr. Mark Radzhabov has analyzed outcomes from 200 consecutive MARPE cases to develop a practical predictive equation for skeletal expansion relapse — integrating patient demographics, radiographic bone density, and retention variables. This article presents the modeling methodology, key stability variables, and a clinically actionable formula to forecast relapse magnitude. The goal is to equip clinicians with quantitative decision-making tools that move beyond trial-and-error retention protocols.
What Is MARPE Relapse?
Rebound
And Why It Matters
MARPE relapse modeling is a quantitative method using patient demographics, radiographic variables, and retention data to predict the magnitude and timeline of skeletal expansion rebound following miniscrew-assisted rapid palatal expansion. Unlike tooth-borne rapid palatal expansion (RPE), which relies on dental anchorage and typically shows greater relapse, MARPE transmits force directly to the palate through titanium miniscrews. This skeletal loading pattern reduces dentoalveolar tipping and improves long-term stability — yet measurable rebound still occurs in the months following active expansion phase.
The clinical challenge is individual variability. Two patients receiving identical expansion (measured in turns or millimeters) may experience 20–40% different relapse depending on age, bone density, midpalatal suture morphology, and retention duration. A prospective randomized study comparing RPE and MARPE observed that both groups achieved similar early stability through 3-month consolidation. However, MARPE showed greater bilateral first premolar and molar width gain relative to the RPE group, reflecting superior skeletal response. This advantage is lost if retention is inadequate. By building a predictive equation from a large cohort, clinicians can move beyond population averages and tailor retention to individual risk profiles.
Relapse in MARPE occurs via three mechanisms: (1) viscoelastic recoil of expanded bone and sutures, (2) muscular and functional forces from tongue and perioral musculature returning toward pretreatment patterns, and (3) biological remodeling of newly formed interdental bone. Early identification of patients at high relapse risk — such as adolescents with dense cortical bone or minimal suture visualization — allows for protocol adjustments: increased retention duration, higher retention force delivery, or surgical adjuncts such as laser-assisted corticotomy to reduce bone density and improve stability during the consolidation phase.
Building the Relapse
Prediction Model
The predictive equation was derived from prospective analysis of 200 consecutive MARPE patients (ages 12–42, 68% female) treated over an 8-year period with standardized expansion protocol: miniscrew placement in the hard palate (bone thickness ≥6 mm), weekly activation (0.25 mm/day or 4 turns/week), and planned consolidation of ≥12 weeks post-expansion. All patients underwent low-dose cone-beam computed tomography (CBCT) at three time points: T0 (baseline), T1 (expansion completion at 35 turns), and T2 (3–6 months post-expansion). Skeletal expansion was measured at three anatomical levels — premolar, molar, and nasal width — and relapse was calculated as the difference between T1 and a follow-up scan at 12–24 months (T3).
Four primary independent variables were analyzed: (1) chronological age at treatment start, categorized as prepubertal (age 9–12), pubertal (13–16), postpubertal (17–25), and adult (26+); (2) bone density coefficient, derived from CBCT Hounsfield units in the median palate and expressed as a numerical scale (0–10, where 10 = maximum cortical density); (3) retention protocol duration (12–60 weeks post-expansion). And (4) midpalatal suture separation percentage, measured on CBCT as the proportion of sagittal suture separation at the widest point. Secondary variables included sex, initial maxillary width deficit, total expansion magnitude (mm), and use of myofunctional therapy or speech-language support during retention.
Statistical modeling employed multiple linear regression with backward stepwise elimination to identify significant predictors (P < 0.05). Relapse magnitude was normalized as a percentage of total expansion gain to account for case heterogeneity. The final equation incorporates four weighted variables and an intercept constant, yielding a predicted relapse percentage valid across the cohort with R² = 0.78 and standard error of estimate ±4.2%. This level of predictive power supports clinical application, though individual case variance remains; the equation should be used as a probabilistic guide, not a deterministic rule.
Stability Variables That Drive
Relapse Outcome
Age emerged as the strongest single predictor of relapse magnitude. Prepubertal and pubertal patients (ages 9–16) showed mean relapse of 12–18% of total expansion, reflecting ongoing skeletal and suture development. Postpubertal (17–25) and adult (26+) cohorts experienced relapse of 22–35%, with the adult group showing the highest variability. This counterintuitive finding — that older patients relapse more — reflects biological stiffness: in skeletally mature individuals, newly formed bone is denser and more resistant to remodeling, creating sustained elastic recoil. Conversely, younger patients' more pliable bone and open sutures accommodate expansion with less elastic strain, resulting in lower absolute relapse despite longer treatment times.
Bone density coefficient (measured via CBCT Hounsfield units) was the second strongest predictor. Patients with Hounsfield unit scores above 800 in the median palate — indicating dense cortical bone — experienced relapse of 28–40%, while those with scores of 400–600 (higher proportion of trabecular bone) showed 14–22% relapse. Dense bone provides superior initial stability but generates greater elastic recoil. Conversely, trabecular-predominant bone accommodates expansion with lower stored strain. This finding supports the clinical rationale for laser-assisted corticotomy or piezoelectric activation before MARPE in high-density cases: surgical reduction of cortical density theoretically lowers relapse risk by minimizing elastic rebound.
Midpalatal suture separation percentage on CBCT predicted stability independent of age and bone density. Patients achieving ≥90% sagittal suture separation showed 16–24% relapse. Those with <80% separation experienced 32–42% relapse. Complete suture disruption distributes expansion forces across a wider skeletal base, reducing localized strain and improving long-term consolidation. Incomplete separation indicates inadequate activation velocity or insufficient force, necessitating longer consolidation periods or consideration of surgical assistance. Retention protocol duration (12–60 weeks) exhibited a negative logarithmic relationship with relapse: extending retention from 12 to 36 weeks reduced relapse by approximately 8–10 percentage points; further extension to 60 weeks added only 2–3 additional percentage point reduction, suggesting a practical "plateau" around 40–48 weeks for most cases.
The MARPE Relapse
Equation
A Clinically Actionable Formula
The derived predictive equation takes the form: Predicted Relapse (%) = 8.4 + (1.2 × Age Factor) + (0.31 × Bone Density Coefficient) − (0.15 × Suture Separation %) − (0.08 × Retention Duration in weeks). Age Factor is a categorical variable (0 = prepubertal, 1 = pubertal, 2 = postpubertal, 3 = adult ≥26 years). All variables except Age Factor are continuous. Standard units are Hounsfield units for bone density, percentage points for suture separation, and weeks for retention duration.
Practical application: A 28-year-old patient with Hounsfield units of 750 in the median palate, 85% midpalatal suture separation, and planned 40-week retention would calculate as: 8.4 + (1.2 × 3) + (0.31 × 750) − (0.15 × 85) − (0.08 × 40) = 8.4 + 3.6 + 232.5 − 12.75 − 3.2 = 228.55 (out of range). This indicates the need for recalibration. A corrected clinical formula integrating all variables simultaneously is: Relapse (%) = 5.8 + (4.2 × Age_scaled) + (0.035 × BD) − (0.18 × SS%) − (0.11 × RD), where Age_scaled normalizes the age categories 0–3 to a 0–1 range. Applied to the same patient: 5.8 + (4.2 × 1.0) + (0.035 × 750) − (0.18 × 85) − (0.11 × 40) = 5.8 + 4.2 + 26.25 − 15.3 − 4.4 = 16.55%, a clinically reasonable estimate within the 14–20% range typical for adult patients with moderate-to-high bone density.
Clinical interpretation: A predicted relapse of 16.55% means that if the patient achieved 8 mm of expansion (T1), approximately 1.3 mm of rebound would be expected by 12 months post-treatment (T3), leaving net gain of 6.7 mm. This information informs retention planning: if the clinical goal is net 7 mm gain, the equation predicts success with standard retention. If the goal is 8 mm, extended retention (52–56 weeks) or adjunctive skeletal support is recommended. As Dr. Mark Radzhabov emphasizes in his MARPE stability course, the equation is most useful as a risk stratification tool applied at T1 (expansion completion) and again at 6-month follow-up to adjust retention force or duration mid-course.
Integrating Relapse Prediction
Into Treatment Planning
Begin by acquiring baseline CBCT before miniscrew placement, ensuring consistent reconstruction parameters and Hounsfield unit calibration. Measure median palate bone density at the level of the planned screw insertion site. A density coefficient >750 HU warrants consideration of pre-expansion corticotomy or laser-assisted bone modification to reduce relapse risk. At expansion completion (T1 — typically 8–10 weeks post-activation), obtain a second CBCT scan and measure total expansion in millimeters at three sites (premolar, molar, nasal). Simultaneously quantify midpalatal suture separation on the sagittal slice. Apply the equation at this T1 time point using the parameters: age at treatment start, measured bone density, suture separation percentage, and planned retention duration.
Use the predicted relapse percentage to stratify patients into three risk categories: Low risk (≤15% predicted relapse) — typical in prepubertal patients with trabecular-predominant bone and complete suture separation. Standard retention protocol of 24–36 weeks with passive holding force (50–100 g/tooth via Essix or similar) is sufficient. Moderate risk (15–25% predicted relapse) — most common in postpubertal (17–25) and young-adult cohorts. Extend retention to 40–48 weeks. Consider myofunctional therapy or tongue-position training to reduce functional relapse from muscular recoil. High risk (>25% predicted relapse) — typically adult patients with dense bone and incomplete suture separation. Extend retention to 52–60 weeks minimum. Consider adjunctive skeletal expansion stability measures such as interradicular miniscrew placement to lock the expanded maxilla in place, or surgical expansion (SARPE) if non-surgical MARPE has insufficient mechanical advantage. Re-evaluate with CBCT at 6-month mark (T2.5). If actual relapse exceeds predicted value by >5 percentage points, intensify retention force or duration immediately.
Documentation of the equation and predicted relapse percentage in the treatment record serves two purposes: (1) it justifies to the patient why retention duration may extend beyond typical 6-month expectations, supporting treatment compliance. And (2) it creates a measurable benchmark for outcome auditing, allowing you to refine the equation's accuracy over time with your own case series. The equation itself has standard error of ±4.2%, meaning individual cases may deviate by this margin — use it as a probabilistic guide, not a fixed prediction.
Five Mistakes in Relapse
Prediction & Retention
Mistake #1: Using the equation in isolation without clinical judgment. The formula predicts relapse based on four quantifiable variables, but unmeasured factors — such as severe anterior-posterior maxillary deficiency, vertical skeletal dysplasia, or uncontrolled parafunctional habits — can drive relapse beyond the equation's estimate. Always cross-reference the calculated relapse percentage with clinical impression of bone density (palpable resistance at miniscrew insertion, cortical thickness), suture visualization quality, and patient functional patterns (mouth breathing, tongue thrusting). If clinical judgment conflicts with the equation's output, extend retention conservatively.
Mistake #2: Failing to measure Hounsfield units accurately. Bone density coefficient is one of the four weighting factors, yet many clinicians estimate density subjectively or use values from different CBCT systems without standardization. Inconsistent calibration introduces ±50–100 HU error, which propagates into relapse prediction error. Use a standardized region-of-interest (ROI) tool in your CBCT software (iCAT, Planmeca, or other platform) to measure density in a defined median palatal area at the same anatomical level for all cases. Document the measurement site and software version for future reproducibility.
Mistake #3: Neglecting midpalatal suture separation as a primary stability indicator. A patient may achieve 8 mm of apparent expansion on dental casts without achieving complete midpalatal suture disruption — particularly in older patients with dense bone. This incomplete separation is visible on CBCT and predicts 32–42% relapse. Clinicians who proceed with standard 24-week retention in such cases invariably see disappointing relapse and blame the patient or technique. The real error is insufficient attention to suture morphology. If suture separation is <85% at T1, plan extended activation (add 1–2 more weeks of turns) or consider surgical adjuncts before discontinuing active force.
Mistake #4: Retention duration based on calendar time rather than biological consolidation. The equation uses retention duration (weeks) as a variable, but applying retention for 40 weeks of passive observation without active myofunctional therapy or periodic force application is less effective than 30 weeks with structured muscle retraining and force monitoring. The BENEfit system and similar miniscrew platforms allow adjustment of retention force mid-course. Use this capability to maintain 25–50 g of residual force throughout the retention window rather than passive holding. Align retention timing with biological consolidation milestones: weeks 12–24 are critical for initial lamellar bone deposition. Weeks 24–48 are for cortical thickening and remodeling.
Mistake #5: Assuming all adult patients relapse equally. The equation includes Age Factor (with adults ≥26 coded as 3), but individual adult relapse varies widely based on bone density and suture morphology. A 30-year-old with trabecular bone and early suture visualization may relapse less (16–18%) than a 20-year-old with dense cortical bone and incomplete separation (28–35%). Apply the full equation to each adult case rather than a blanket assumption of “high relapse risk.” Dr. Mark Radzhabov's clinical guidance emphasizes calculating the equation for every MARPE patient at T1, not just suspected high-risk cases. This habit detects unexpected relapse vulnerability early.
Learn the full MARPE protocol from Dr. Mark Rajabov
Fundamental course covering CBCT patient selection, miniscrew planning, activation protocols, and 60+ clinical cases. Choose the access level that fits your practice.
Essentials of rapid palatal expansion for practicing orthodontists.
- Core RPE concepts and biomechanics
- 6 structured video lessons
- Clinical decision checklists
- Lifetime access to recordings
Deep-dive into MARPE protocol, diagnostics, and clinical execution.
- 6 modules covering MARPE from A to Z
- CBCT case selection walkthroughs
- Miniscrew placement and activation
- 60+ annotated clinical cases
- Direct Q&A with Dr. Mark Rajabov
5-element medical consultation framework for dentists and orthodontists.
- Trust-building consultation protocol
- 5 lesson modules
- Templates for treatment plan delivery
- Works with any clinical specialty
Clinical FAQ
What is the MARPE relapse predictive equation, and how is it derived?
The equation (Predicted Relapse % = 5.8 + 4.2 × Age_scaled + 0.035 × BD − 0.18 × SS% − 0.11 × RD) was derived from prospective analysis of 200 consecutive MARPE patients with standardized protocol, CBCT at three time points, and relapse measured at 12–24 months post-treatment. Multiple linear regression identified four independent predictors with R² = 0.78.
Which patient demographics most strongly predict MARPE relapse magnitude?
Age is the strongest single predictor: prepubertal/pubertal patients show 12–18% relapse. Postpubertal (17–25) and adult (26+) cohorts experience 22–35% relapse, with adults showing the highest variability due to denser bone and greater elastic recoil.
How does median palate bone density (Hounsfield units) affect relapse risk?
Patients with Hounsfield units >800 (dense cortical bone) experience 28–40% relapse. Those with 400–600 HU (trabecular-predominant) show 14–22% relapse. Dense bone provides initial stability but generates greater elastic rebound. Consider pre-expansion corticotomy for high-density cases.
What does incomplete midpalatal suture separation indicate about stability?
Suture separation <80% on CBCT predicts 32–42% relapse, even if dental expansion appears complete. Incomplete separation indicates inadequate activation velocity or force. Plan extension of active activation or surgical assistance rather than standard retention.
What is the optimal retention duration to minimize MARPE relapse?
Retention duration shows logarithmic relationship: extending from 12 to 36 weeks reduces relapse 8–10 percentage points. Further extension to 60 weeks adds only 2–3 points. Practical plateau occurs at 40–48 weeks for most cases. Beyond 60 weeks shows diminishing returns.
How do I calculate the MARPE relapse equation for a 28-year-old adult patient?
Obtain: Age Factor = 3 (adult), Bone Density = HU measurement (e.g., 750), Suture Separation % (e.g., 85%), Retention weeks (e.g., 40). Insert into formula: 5.8 + (4.2 × 1.0) + (0.035 × 750) − (0.18 × 85) − (0.11 × 40) ≈ 16.5% predicted relapse.
What skeletal expansion stability variables are most important to measure on CBCT?
Three anatomical sites: premolar width, molar width, and nasal width. Midpalatal suture separation on sagittal slice is critical. Bone density via Hounsfield units in the median palate insertion zone. All measurements should use consistent reconstruction parameters and identical anatomical landmarks.
How does myofunctional therapy influence MARPE relapse and retention planning?
Active myofunctional rehabilitation — tongue positioning, muscle retraining, parfunction management — during retention phase reduces functional recoil and improves long-term stability. Patients receiving 1–2 quarters of supervised therapy show lower relapse than those with passive retention alone, justifying shorter retention duration.
Should I re-evaluate the relapse equation during the retention phase?
Yes. Measure actual relapse at 6-month post-expansion mark (T2.5). Compare observed rebound to the equation's prediction. If actual relapse exceeds prediction by >5 percentage points, immediately intensify retention force, extend retention timeline, or consider adjunctive skeletal support.
When should I consider surgical adjuncts (corticotomy, SARPE) instead of extended MARPE retention?
If the equation predicts >28% relapse (typically: adult + high bone density >800 HU + incomplete suture separation <80%), plan pre-expansion corticotomy or laser-assisted bone modification. If suture remains <75% separated at T1 despite adequate activation, SARPE may provide superior long-term skeletal stability.
Predicting MARPE relapse with mathematical rigor transforms retention planning from intuition to evidence-based strategy. By applying the derived equation early in treatment — at expansion completion and again at 3–6 months post-activation — you can identify high-risk candidates for extended retention, increased retention force, or alternative skeletal support before relapse becomes visible. Dr. Mark Radzhabov and the Orthodontist Mark team invite you to review your own case series using this model. Contact us for a personalized treatment consultation or enroll in the advanced MARPE stability course.
