Shared Strides: Community-based, high-throughput biomechanics data collection in knee osteoarthritis

This study demonstrates that implementing a high-throughput, markerless biomechanics protocol in community-based settings is an acceptable and effective strategy for recruiting a diverse sample of knee osteoarthritis patients, thereby enhancing research engagement and reducing sample bias compared to traditional on-campus approaches.

The Gist

Imagine you are trying to understand how people with painful knees move, but you've been asking them to come to a fancy, high-tech lab deep inside a university. It's like asking someone to drive to the other side of town just to try on a pair of shoes. Only the most dedicated (or lucky) people show up, and they all look and act the same. You end up with a small, biased group of data that doesn't really tell you how everyone with knee pain moves.

This paper, titled "Shared Strides," is about a team of researchers who decided to flip the script. Instead of making people come to the lab, they packed up their high-tech equipment and took the lab to the people.

Here is the story of what they did, explained simply:

1. The Problem: The "Fortress" Lab

Traditionally, studying how people walk or squat with knee pain (osteoarthritis) happens in a "fortress" lab. You need special markers stuck to your skin, you have to wear tight clothes, and you have to drive to a specific building at a specific time.

• The Analogy: It's like a VIP club. Only people who know the code, have the time, and can afford the gas to get there get in. The result? The data looks like a photo of a single, very specific type of person, not the whole neighborhood.

2. The Solution: The "Pop-Up" Lab

The researchers created a portable, high-tech lab. Think of it like a food truck, but instead of tacos, it serves "movement data."

• The Tech: They used markerless motion capture. Imagine a movie set where actors don't need to wear those green suits with dots on them. The cameras just "see" you moving naturally.
• The Strategy: They set up this "food truck" in four different places:
  • Two spots on the university campus (the traditional "fortress").
  • Two spots in the community (a Senior Center and a Retirement Community).

3. The Experiment: Walking In vs. Driving In

They ran events where people could either:

• Schedule an appointment (like booking a dentist).
• Just walk in (like grabbing a coffee).

What happened?

• The "Walk-Ins": At the community sites, people saw the setup, got curious, and walked right in. About 40% of the people they studied were either new to research or just walked in that day.
• The "Regulars": At the university sites, most people had to schedule ahead.
• The Result: They collected data from 85 people (a big number for this type of study!). The people at the community sites felt more comfortable because the place was familiar to them, like their local library or community center, rather than a scary medical building.

4. The Big Takeaway: "Familiarity is Key"

The researchers asked the participants: "Would you come back if we did this at a community center instead of a university?"

• The Answer: People at the community centers said, "Yes, absolutely!"
• Why? Two main reasons:
  1. Familiarity: They knew the building. It felt safe.
  2. Convenience: It was closer, and parking was easier (or free!).

5. Why This Matters (The "So What?")

If we only study people who can drive to a university lab, we miss out on the people who take the bus, who have trouble walking far, or who just don't trust big institutions.

• The Metaphor: Imagine trying to design a new car seat. If you only ask professional race car drivers what they like, your new seat will be terrible for a grandma going to the grocery store.
• The Future: By taking the lab to the community, the researchers got a much more diverse and honest group of people. This means the solutions they find for knee pain will actually work for real people, not just the "lab regulars."

In a Nutshell

This paper proves that you don't need to build a bigger, fancier lab to get better science. Sometimes, you just need to pack up your tools, drive to the neighborhood, and knock on the door. By making research accessible and friendly, they got more people involved, got better data, and made science work for everyone, not just the few.

Technical Summary

1. Problem Statement

Knee osteoarthritis (OA) progression is influenced by mechanical loading, yet patient responsiveness to biomechanical interventions is highly variable due to inter-individual differences in symptoms and structural damage. Historically, biomechanics research has been limited by:

• Small Sample Sizes: Traditional optical motion capture studies typically involve 2–20 participants.
• Selection Bias: Laboratory-based studies require participants to travel to specific facilities, often excluding those with mobility issues or limited transportation, leading to homogeneous samples (often younger, healthier, or more affluent).
• Data Collection Constraints: The need for markers, wearable sensors, and constrictive clothing makes traditional methods time-consuming and difficult to scale for large, representative datasets.

The study aims to address these limitations by testing a high-throughput, community-based protocol using markerless motion capture to recruit a larger, more diverse sample of individuals with knee OA.

2. Methodology

Study Design:

• Type: Cross-sectional feasibility study (part of the "Shared Strides Study").
• Duration: 11 months with data collection events on 12 separate dates.
• Location: Four sites in Gainesville, FL:
  • Community Sites: Senior Center and Retirement Community (high foot traffic, public hallways).
  • On-Campus Sites: Clinical Laboratory (private hallway) and Student Union.
• Participants: Adults aged ≥40 with self-reported or ACR-diagnosed knee OA, capable of walking 20 feet without assistive devices.
• Exclusion: Neurological/musculoskeletal conditions affecting movement more than OA, recent lower-limb injury/surgery (<3 months), or pregnancy.

Data Collection Protocol:

• Technology: A portable biomechanics laboratory utilizing:
  • Markerless Motion Capture: Vicon Vero cameras with Nexus 2.16 software and Theia 3D software.
  • Force Platforms: Embedded in a portable walkway (AMTI).
  • Mobile Workstation: Custom-built computing unit.
• Throughput: Participants arrived 15–30 minutes apart. Average visit length was 65.8 ± 16.0 minutes. Multiple participants could be on-site simultaneously.
• Tasks: Activities of Daily Living (ADLs) including walking (usual/fast pace), balance (double/single-leg), squatting, sit-to-stand, step-up, and step-down.
• Recruitment Strategy:
  • Scheduled: Phone screening followed by specific appointment times.
  • Walk-in: On-site screening with immediate participation options (same-day or same-site).
• Acceptability Assessment: A custom 15-item questionnaire evaluated participant perceptions regarding site preference, transportation, familiarity, and motivations (altruism, compensation, health knowledge).

Statistical Analysis:

• Used Welch's ANOVA for continuous variables and Fisher's exact test for categorical variables to compare recruitment outcomes and demographics across the four sites.
• Pairwise post-hoc comparisons were conducted with Holm adjustment for significant findings.

3. Key Contributions

• Validation of High-Throughput Community Protocols: Demonstrated that portable, markerless biomechanics labs can be successfully deployed in non-laboratory settings (community centers) without compromising data collection efficiency.
• Recruitment Innovation: Showed that combining scheduled appointments with walk-in screening significantly increases enrollment, particularly among first-time research participants.
• Demographic Reach: While the overall sample remained predominantly White and female (reflecting the local catchment area), the community approach successfully recruited a 40% "new" participant rate (first-time researchers and/or walk-ins), a group often underrepresented in traditional lab studies.
• Acceptability Metrics: Identified familiarity with the site and distance/transportation as the primary drivers for participant engagement, rather than just the type of research.

4. Key Results

• Recruitment Outcomes:
  • Total Enrolled: 85 participants (from 118 eligible out of 187 screened).
  • Walk-in Success: 34 walk-ins were screened; 19 enrolled immediately at the community site.
  • Conversion Rates: Walk-in eligibility-to-enrollment conversion was 91%, compared to 68% for phone-screened participants.
  • New Participants: 21 participants had no prior research experience; 16 of these were recruited at community sites.
• Demographics:
  • Mean age: ~73 years; 76% female; predominantly White and non-Hispanic.
  • No statistically significant differences in demographics (age, sex, BMI, income, education) were found between the four sites ($p > 0.05$).
• Acceptability & Preferences:
  • Site Preference: Participants at the Senior Center (highest walk-in volume) were significantly more likely to agree they would prefer community locations over campus ($p = 0.019$).
  • Key Drivers: "Familiarity with the site" and "Distance/Transportation" were the top reasons for site preference.
  • Flexibility: 58% of participants reported that having a choice of location affected their desire to participate, and 36% reported it affected their ability to participate (significant difference across sites, $p = 0.032$).
  • Motivations: While altruism was common across all sites, on-campus participants cited compensation as a major motivator, whereas Senior Center participants cited increased health knowledge.

5. Significance and Implications

• Reducing Bias: This approach offers a scalable model to reduce selection bias and homogeneity in OA biomechanics research by lowering barriers to entry (travel, time, familiarity).
• Scalability: The "high-throughput" nature (65-minute visits with overlapping participants) proves that large-scale biomechanical data collection is feasible outside traditional labs.
• Community Engagement: The study highlights that "community-based" is not just a location change but a strategy to engage first-time researchers and those with limited mobility or transportation resources.
• Future Research: The authors suggest that future studies should target more diverse geographic areas to further reduce selection bias and that sharing research results with the public (health knowledge) is a potent recruitment tool for community settings.

Conclusion: Integrating markerless motion capture with community-based recruitment strategies enhances participant experience, facilitates larger and more heterogeneous sample sizes, and represents a viable path forward for inclusive, large-scale biomechanics research in knee osteoarthritis.