Equitable Pain Assessment - GSO Research Showcase 2026

Toward Equitable Pain Assessment: Quantitative Sensory Testing in Racially and Ethnically Diverse Pain-Free Adults

Mariam Maqsood, DPT, MSc¹; Ericka N. Merriwether, PT, DPT, PhD¹,²

¹Department of Physical Therapy, NYU Steinhardt School of Culture, Education, and Human Development, New York University, New York, NY

²Department of Medicine, NYU Grossman School of Medicine, New York University, New York, NY

Abstract

Normative quantitative sensory testing (QST) data from pain-free adults are predominantly derived from racially and ethnically homogeneous samples (i.e., White/Caucasian), limiting interpretation of pain mechanisms in diverse patient populations. Existing datasets provide values primarily for hand and foot regions but lack data for clinically relevant sites such as trapezius, lumbar spine, and knee. This study characterized pressure pain sensitivity, mechanical pain facilitation, and endogenous pain modulation in a racially and ethnically diverse sample across multiple anatomical regions. Twenty pain-free adults (age 30 ± 6.6 years; 10M/10F; 60% Asian, 15% White, 10% Black, 15% multiracial/other) completed standardized QST across two sessions spaced one week apart. Pressure pain thresholds were assessed at trapezius (370 ± 130 kPa), lumbar paraspinal muscle (380 ± 133 kPa), and knee (304 ± 106 kPa). Temporal summation was evaluated at first dorsal webspace, third proximal phalanx, forearm, lumbar spine, and knee with difference scores of 1.2-1.9. Conditioned pain modulation showed mean pain inhibition of 43.4% (parallel) and 42.0% (sequential). Substantial individual variability was observed across all measures (coefficient of variation 35-88%), highlighting diverse pain processing patterns. Test-retest reliability was good to excellent for pressure pain thresholds (ICC 0.93-0.99) and temporal summation (ICC 0.83-0.93). Conditioned pain modulation raw values showed excellent reliability (ICC 0.83-0.94), though percent-change scores were less stable. These findings demonstrate rigorous, reliable QST is feasible in racially and ethnically diverse populations and addresses critical representation gaps in pain research.

Background and Significance

Quantitative sensory testing (QST) provides standardized, objective assessment of somatosensory function and pain processing mechanisms, including mechanical pain sensitivity, temporal summation (central facilitation), and conditioned pain modulation (descending inhibition). These measures are essential for characterizing pain phenotypes, identifying underlying mechanisms in chronic pain conditions, and evaluating treatment response in clinical and research settings.

Current QST normative datasets lack racial and ethnic diversity, limiting applicability to diverse clinical populations. Foundational references—including the widely-used German DFNS protocol (Rolke et al., 2006) and datasets from Israel (Yarnitsky & Sprecher, 1994), Australia (Hafner et al., 2015), and Germany (Magerl et al., 2010)—do not report participant race or ethnicity. Despite documented ethnic differences in QST responses across populations (Yang et al., 2013; González-Duarte et al., 2016; Campbell et al., 2005), comprehensive diverse normative data remain limited. This study characterizes QST responses in a racially and ethnically diverse sample across clinically relevant musculoskeletal sites, providing preliminary data to inform equitable normative reference development.

Research Questions

1. What is the magnitude and variability of QST responses at clinically relevant anatomical sites (trapezius, lumbar spine, knee) in a racially and ethnically diverse sample of pain-free adults?

2. What are the test-retest reliability estimates for QST measures (PPT, TS, CPM) in a racially and ethnically diverse sample of pain-free adults?

Methods

Participants

Twenty pain-free adults participated in this study (age 30.0 ± 6.6 years; BMI 24.3 ± 3.8 kg/m²; M=10, F=10). The sample was racially and ethnically diverse: 12 Asian (60%), 3 White (15%), 2 Black (10%), and 3 multiracial/mixed (15%) participants. (Figure 1).

Ethnicity was self-reported as 95% Not Hispanic/Latino (19), 5% Hispanic/Latino (1)

All participants were free of chronic pain conditions, neurological disorders, and sensory impairments at the time of testing.

Two pie charts comparing racial and ethnic composition of study samples. Left chart titled "Typical QST Normative Studies" shows a predominantly gray circle representing approximately 90 percent homogeneous samples with a small lighter gray segment for other or not reported. Right chart titled "This Study (N=20)" shows four colored segments: dark purple representing 60 percent Asian (12 participants), gold representing 15 percent White (3 participants), dark purple representing 10 percent Black (2 participants), and light purple representing 15 percent multiracial or mixed (3 participants). Legend in top right identifies body region color coding.

Figure 1. Racial composition of study sample compared to typical QST normative research populations. Existing normative datasets (left) are predominantly derived from homogeneous samples, while this study (right) included 60% Asian, 15% White, 10% Black, and 15% multiracial/mixed participants.

Study Design

Participants completed standardized quantitative sensory testing (QST) protocols across two sessions spaced one week apart to assess test-retest reliability. All testing was conducted in a temperature-controlled laboratory environment following standardized verbal instructions and anatomical landmark identification procedures.

QST Assessment Protocol

See Figures 2A-C below for detailed assessment procedures.

Medical illustration showing pressure pain threshold assessment from posterior view. Brown-skinned person with short dark hair standing with back to viewer, wearing dark green shorts. Left upper back shows teal circle labeled "Trapezius (non-dominant)" with arrow pointing to site. Lower back shows blue circle labeled "L2 Paraspinal (2 cm lateral)" with arrow pointing to site positioned to the right of spine. Right side of image shows brown hand holding blue rectangular Wagner FPIX algometer device with digital display reading "000.0" and three red buttons labeled "FORCE ONE". Black probe extends from bottom of device toward the body. Text beside device reads "Wagner FPIX 50 algometer" and "50 kPa/s, 3 trials/site, averaged value". Small inset box shows close-up of black circular probe tip. Larger inset box in lower left shows side view of bent knee with yellow circle on inner knee labeled "Knee (medial side)". Bottom text reads "Pressure Pain Threshold (PPT) equals Pressure applied until first pain sensation reported". White background with clear black labels and arrows.

Figure 2A. Pressure pain threshold (PPT) assessment protocol. PPT was assessed at three anatomical sites using pressure algometry with ascending stimulation rate (50 kPa/s, Wagner FPIX-50 algometer, Greenwich, CT).

Testing sites: non-dominant trapezius (teal circle), L2 paraspinal muscle 2 cm lateral to spinal midline (blue circle), and medial knee (yellow circle, shown in inset). Three trials per site were obtained and averaged to determine PPT values. Pressure applied until participant reported first pain sensation. This protocol follows standardized guidelines for mechanical pain sensitivity assessment.

Figure 2B. Temporal summation (TS) assessment protocol. TS assessed at five anatomical sites using repetitive pinprick stimulation (256 mN, weighted pinprick device).

Middle panel shows three testing sites: 1st dorsal webspace (green circle, Hand 1 inset showing dorsal surface), 3rd metacarpal (light blue circle labeled on Hand 1 inset showing dorsal surface), forearm (dark blue circle, Arm 2 inset showing volar surface). The left panel shows patella (yellow circle). Right panel (Posterior View) shows L2 spine testing site (orange circle).

Timeline at bottom illustrates stimulus progression from first (circle, 0) to tenth (triangle, 10) stimulus. Wind-up difference calculated as change in pain rating between 10th and 1st stimulus (0-10 numeric rating scale).

Note: L2 spine site (on vertebra) for TS differs from L2 paraspinal site (2 cm lateral) used for PPT assessment.

Figure 2C. Conditioned pain modulation (CPM) assessment protocol showing parallel and sequential paradigms.

Left panel (Parallel Paradigm): Non-dominant hand immersed in cold water (8-10°C/46.4-50°F) while simultaneously assessing PPT at dominant trapezius (blue circle) using pressure algometry. Test stimulus (trapezius PPT) and conditioning stimulus (cold water) delivered concurrently.

Right panel (Sequential Paradigm): PPT assessed at dominant trapezius (yellow circle) immediately following withdrawal from cold water.

Sequential test was delivered after parallel test completion, indicated by curved arrow showing temporal sequence.

CPM was calculated as absolute change or percentage change. Both paradigms assess endogenous pain modulation capacity via descending inhibitory pathways.

Statistical Analysis

Descriptive statistics (means, standard deviations, ranges, coefficients of variation) were calculated for all QST parameters. Test-retest reliability was assessed using intraclass correlation coefficients (ICC) with two-way mixed effects model for absolute agreement. ICC values were interpreted as poor (<0.50), moderate (0.50-0.74), good (0.75-0.89), or excellent (≥0.90) according to established guidelines (Koo & Li, 2016).

Results

All QST measures demonstrated substantial individual variability with good to excellent test-retest reliability (ICC 0.83-0.96). See Figures 3-6 for complete results.

Scatter plot with error bars showing pressure pain thresholds at three anatomical sites. Green-teal filled circles show 20 individual trapezius values scattered vertically from approximately 180 to 680 kilopascals, with a white diamond marker at the mean of 370.0 kilopascals. Blue filled circles show 20 individual L2 lumbar values scattered vertically from approximately 190 to 680 kilopascals, with a white diamond marker at the mean of 380.1 kilopascals. Gold-orange filled circles show 20 individual knee values scattered vertically from approximately 140 to 520 kilopascals, with a white diamond marker at the mean of 303.8 kilopascals. Thick colored error bars extend vertically from each mean showing plus or minus one standard deviation: green for trapezius ranging approximately 240 to 500 kPa, blue for L2 ranging approximately 250 to 510 kPa, and gold for knee ranging approximately 200 to 410 kPa. Text box in upper left states "n equals 20 per region". Legend in upper right shows filled gray circle labeled "Individual values" and white diamond labeled "Mean plus or minus SD". Y-axis labeled "Pressure Pain Threshold (kPa)" ranges from 100 to 700. X-axis labeled "Body Region" shows three categories: Trapezius, L2, and Knee. Light gray horizontal gridlines aid readability.

Figure 3. Pressure pain thresholds across anatomical sites with individual participant values. Mean ± SD: Trapezius 370.0 ± 130.1 kPa (bluish green), L2 lumbar 380.1 ± 133.0 kPa (dark blue), Medial knee 303.8 ± 105.8 kPa (orange). Individual dots show substantial variability (coefficient of variation 35%) in pain processing patterns.

Error bars represent ± 1 standard deviation. N=20.

Scatter plot with error bars showing temporal summation wind-up difference scores at five anatomical sites. Green-teal filled circles show 20 individual 1st dorsal webspace values scattered vertically from approximately 0.8 to 3.6, with a white diamond marker at the mean of 1.6. Light blue filled circles show 20 individual 3rd MCP values scattered vertically from approximately 0.3 to 4.3, with a white diamond marker at the mean of 1.8. Dark blue filled circles show 20 individual forearm values scattered vertically from approximately 0.6 to 4.0, with a white diamond marker at the mean of 1.9. Orange filled circles show 20 individual L2 lumbar values scattered vertically from approximately 0.2 to 5.0, with a white diamond marker at the mean of 1.5. Gold-yellow filled circles show 20 individual knee values scattered vertically from approximately 0.3 to 2.8, with a white diamond marker at the mean of 1.2. Thick colored error bars extend vertically from each mean showing plus or minus one standard deviation. Text box in upper left states "n equals 20 per region". Legend in upper right shows filled gray circle labeled "Individual values" and white diamond labeled "Mean plus or minus SD". Y-axis labeled "Wind-Up Difference (delta Pain Rating)" ranges from 0 to 5. X-axis labeled "Body Region" shows five categories: 1st Dorsal Webspace, 3rd MCP, Forearm, L2, and Knee. Light gray horizontal gridlines aid readability.

Figure 4. Temporal summation scores across five anatomical sites with individual participant values.

Mean ± SD: 1st Dorsal Webspace 1.6 ± 0.8 (bluish green), 3rd MCP 1.8 ± 1.3 (sky blue), Forearm 1.9 ± 1.0 (dark blue), L2 lumbar 1.5 ± 1.3 (vermillion), Knee 1.2 ± 0.7 (orange).

Individual dots show high variability (coefficient of variation 51-88%) in facilitated pain processing.

Error bars represent ± 1 standard deviation. N=20.

Scatter plot with error bars showing conditioned pain modulation percentage effects for two experimental paradigms. A horizontal dashed gray line crosses the entire plot at zero percent labeled "No CPM Effect (0%)" dividing the graph into pain inhibition above and pain facilitation below. Left column labeled "Parallel" with "n equals 15" shows blue filled circles representing 15 individual participant responses scattered vertically from approximately negative 3 percent to positive 88 percent, with most dots clustered between 20 and 80 percent. A white diamond marker at 43.4 percent indicates the mean with a thick blue error bar extending vertically showing plus or minus one standard deviation ranging approximately from 15 to 71 percent. Right column labeled "Sequential" with "n equals 18" shows gold-orange filled circles representing 18 individual participant responses scattered vertically from approximately negative 22 percent to positive 94 percent, with dots spread across the range. A white diamond marker at 42.0 percent indicates the mean with a thick gold error bar extending vertically showing plus or minus one standard deviation ranging approximately from 10 to 74 percent. Text box in upper left states "Session 1". Legend in upper right shows filled gray circle labeled "Individual values", white diamond labeled "Mean plus or minus SD", and dashed line labeled "No CPM Effect (0%)". Y-axis labeled "CPM Effect (%)" ranges from negative 20 to positive 120. X-axis labeled "CPM Protocol" shows two categories: Parallel and Sequential. Light gray horizontal gridlines aid readability.

Figure 5. Conditioned pain modulation (CPM) percentage effects comparing parallel and sequential protocols (Session 1) with individual participant values.

Mean ± SD: Parallel 43.4% ± 28.0% (dark blue, n=15), Sequential 42.0% ± 32.1% (orange, n=18).

Dashed line indicates no CPM effect (0%). Positive values indicate pain inhibition, negative values indicate pain facilitation.

Individual dots show substantial variability in endogenous pain modulation capacity.

Error bars represent ± 1 standard deviation.

Grouped bar chart showing test-retest reliability ICC values for eleven quantitative sensory testing measures organized in three visually separated groups. First group shows three blue bars for pressure pain thresholds: trapezius at 0.96, L2 at 0.95, and knee at 0.93, with "PPT" label below in blue text. A thin vertical gray line separates this from the second group showing five purple bars for temporal summation: 1st DWS at 0.93, 3rd MCP at 0.90, forearm at 0.84, L2 at 0.83, and knee at 0.90, with "TS" label below in purple text. Another thin vertical gray line separates this from the third group showing three orange bars for conditioned pain modulation: baseline at 0.92, parallel at 0.83, and sequential at 0.94, with "CPM" label below in orange text. All ICC values are displayed in bold black text above each bar. Two horizontal dashed gray reference lines cross the chart: one at 0.75 labeled "Good (0.75)" on the right side, and one at 0.90 labeled "Excellent (0.90)" on the right side. The y-axis is labeled "Intraclass Correlation Coefficient (ICC)" and ranges from 0.70 to 1.00. The x-axis shows individual measure abbreviations below each bar. A horizontal legend at bottom center shows three colored boxes with labels: blue box "PPT (Pressure Pain Threshold)", purple box "TS (Temporal Summation)", orange box "CPM (Conditioned Pain Modulation)". Light gray horizontal gridlines aid readability.

Figure 6. Test-retest reliability (ICC) for quantitative sensory testing measures. All measures demonstrated good to excellent reliability (ICC ≥0.83). Dashed lines indicate good (0.75) and excellent (0.90) reliability thresholds.

ICC interpretation: <0.50=poor, 0.50-0.75=moderate, 0.75-0.90=good, >0.90=excellent (Koo & Li, 2016).

Conclusion

QST demonstrates excellent reliability in diverse populations (ICC ≥0.83). Individual variability reflects genuine biological diversity, emphasizing the need for personalized assessment and diverse normative databases to advance equitable pain care.

Clinical Implications

This study has important implications for equitable pain assessment:

QST is reliable in diverse populations: Clinicians can confidently use PPT, TS and CPM protocols in racially and ethnically diverse patients, as these measures demonstrate excellent reliability (ICC 0.83-0.99) across demographic groups.
Individual variability is real biological diversity: Substantial individual differences (CV 34-76% for reliable measures) emphasize the importance of individualized pain assessment rather than rigid normative comparisons. This variability represents genuine diversity in pain processing mechanisms, not measurement error.
Current normative datasets may perpetuate inequities: Existing QST reference values derive from predominantly homogeneous samples and may inadequately represent pain processing in patients from underrepresented racial and ethnic groups, given documented ethnic differences in pain responses (Campbell et al., 2005; González-Duarte et al., 2016).
Inclusive pain research is both feasible and rigorous: Inclusion of diverse participants in pain research is both feasible and methodologically rigorous, supporting the development of inclusive normative databases for equitable clinical care.

Acknowledgments

Funding: This work was supported by the NYU Steinhardt Spring 2025 Doctoral Research and Travel Grant.

Affiliation: This research was conducted through the Inclusive and Translational Research in Pain (ITRIP) Laboratory at NYU Steinhardt School of Culture, Education, and Human Development.

We extend our gratitude to all members of the ITRIP Laboratory for their unwavering commitment to advancing equity in pain research and their dedication to promoting inclusive, diverse, and rigorous scientific inquiry.

We are deeply grateful to all study participants for their time, dedication, and contribution to advancing equitable pain research.

Author Information

Mariam Maqsood, DPT, MSc

Contact: mm13154@nyu.edu

Mariam Maqsood is a PhD candidate in Physical Therapy Research at NYU Steinhardt School of Culture, Education, and Human Development. She earned her Doctor of Physical Therapy from Dow University (Pakistan) and Master of Science in Recreation from the University of Pécs (Hungary). As someone with lived experience navigating chronic pain, she brings both professional expertise and personal insight to her research examining chronic musculoskeletal pain in persons with blindness and low vision. Her work develops accessible pain assessment protocols grounded in health equity and critical disability theory. She is a Steinhardt Doctoral Fellow and recipient of the NYU Steinhardt Spring 2025 and 2026 Doctoral Research and Travel Grant.

Ericka N. Merriwether, PT, DPT, PhD

Contact: em3766@nyu.edu

Ericka N. Merriwether is an Associate Professor in the Department of Physical Therapy and Director of the Inclusive and Translational Research in Pain (ITRIP) Laboratory at NYU Steinhardt. She earned her DPT from Mayo Clinic School of Health Sciences and PhD in Movement Science from Washington University in St. Louis. Dr. Merriwether's research examines multi-level mechanisms of chronic musculoskeletal pain in racially and ethnically diverse adults with higher BMI, integrating biophysical pain measurement, motion analysis, and clinical assessment to characterize unique pain phenotypes. Her work has been supported by the National Institutes of Health (NIH-NIAMS K23AR080846), the Foundation for Physical Therapy, and NYU Steinhardt. She serves on the Board of Directors of the United States Association for the Study of Pain (USASP) and is a recipient of the NYU Steinhardt Gabriel Carras Award and Champions of Equity Distinguished Service Award.

Page updated

Report abuse