Temporal summation (TS) and offset analgesia (OA) are widely used psychophysical endpoints in pain research that index different components of central nociceptive processing. While crossover designs are commonly used in experimental pain studies to reduce between-participant variability, the design-stability of these endpoints under repeated testing during experimental sensitisation is not well characterised. This study compared the design-stability of mechanical TS (Sumscore) and offset analgesia magnitude (OffA) in a two-period, vehicle-controlled crossover trial of capsaicin-evoked secondary hyperalgesia in healthy adults. The primary aim was methodological: to determine whether session-order effects differ between TS and OffA when these are used as outcome measures in two-period crossover designs of capsaicin-induced central sensitisation. Topical capsaicin was used as a reversible experimental intervention to create a controlled, transient state of secondary hyperalgesia rather than as a therapeutic intervention. The study informs endpoint selection in future quantitative sensory testing (QST) crossover trials.

This was a single-site, two-period, vehicle-controlled, allocation- and formulation-blinded, randomised crossover study conducted at the Department of Human Health Science, University of Guelph, Canada. Sixteen healthy adult volunteers (10 female, 6 male; mean age 21.9 years) were recruited between May 2024 and April 2025. Participants were randomised in a 1:1 intended allocation to one of two sequences: capsaicin-first then vehicle (Sequence A-to-B) or vehicle-first then capsaicin (Sequence B-to-A); actual allocation was 6:10. The allocation sequence was computer-generated by a research assistant not involved in outcome collection. Sessions were separated by a minimum 1-week washout. Topical 0.075% capsaicin cream (Zostrix; Hi-Tech Pharmacal) was applied as a 5 mL dose to a 5 x 10 cm target area on the lateral elbow and to bilateral C5-C6 cervicothoracic dermatomes. The comparator was an equivalent 5 mL volume of inert vehicle lotion (Lubriderm; Johnson & Johnson) applied identically; creams were matched for colour and texture and prepared by a research assistant in unlabelled containers. Mechanical temporal summation (Sumscore; 256 mN pinprick, 16 stimuli at 1 Hz) and thermal offset analgesia (OA; 32-46°C stepped-stimulus protocol with three thermal hold durations of 5, 10, and 15 s) were assessed at baseline and at 10, 20, 30, and 40 min post-intervention. Continuous pain ratings during thermal stimulation were captured via CoVAS. Primary outcomes (Sumscore and the duration-averaged baseline-normalised OffA) were analysed using REML linear mixed-effects models with fixed effects for Intervention, Time, Intervention x Time, Period, and Sequence, with random intercepts for Participant; sex was included as a pre-specified covariate. Within-session temporal stability was assessed using ICC(2,1). A combined Measure x Intervention x Period model tested for differential design-sensitivity between TS and OffA. The study was approved by the University of Guelph Research Ethics Board (REB #19-06-003).

Inclusion Criteria:

• Healthy adult volunteers (≥18 years of age)
• Able to provide written informed consent in English

Exclusion Criteria:

• Current or chronic pain
• Neurological disease
• Skin conditions or sensitivity over the test sites to capsaicin
• Contraindications to thermal stimulation of the skin
• Musculoskeletal conditions that could alter somatosensory processing
• Use of analgesic medications within 24 hours prior to each session
• Use of caffeine within 24 hours prior to each session
• Use of alcohol within 24 hours prior to each session

De-identified individual participant data (IPD) and statistical analysis code will be available from the corresponding author on reasonable request, subject to a data-use agreement and institutional ethics/data-sharing requirements. Materials required to reproduce the intervention and assessment protocol are described in the published Methods and Supplementary Materials.

On request following publication of the primary manuscript.

Reasonable request to the corresponding author for non-commercial research use, subject to a data-use agreement consistent with University of Guelph data-sharing policy.

By design, participants, the investigator who applied the creams and collected the outcome data, and the data analyst were blinded to allocation. The randomisation sequence was computer-generated by a research assistant not involved in testing and concealed in sequentially sealed, opaque envelopes; capsaicin and vehicle creams were matched for colour and texture and dispensed in unlabelled containers. Because topical 0.075% capsaicin can produce perceptible burning and visible erythema, participant and investigator blinding may have been incompletely maintained. Participant allocation guesses were collected and investigator awareness of allocation was not formally assessed. The data analyst remained blinded through data lock and application of pre-specified exclusion criteria, and was unblinded only after the analysis dataset was finalised.