Are there Hawthorne-like participation effects in oral-hygiene research? Evidence from longitudinal plaque mapping

To investigate Hawthorne-like research participation effects on toothbrushing-related plaque outcomes. The data source was a longitudinal flossing study without toothbrushing instruction.

This secondary analysis used disclosed-plaque intraoral scans from 37 subjects (23.1±3.2 years) obtained at four pre-flossing time points over 4 weeks. Plaque coverage (P%) was quantified planimetrically on oral (i.e. lingual/palatal) and vestibular (i.e. buccal/labial) surfaces of the Ramfjord teeth (FDI 16, 21, 24, 36, 41, 44). Linear mixed-effects models tested changes over time and, in a more differentiated model, surface and tooth-group interactions. Baseline tertiles of participant-level P% were used to assess heterogeneity in trajectories. Plaque dynamics (Gain/Loss/Stable areas) were derived from binary plaque masks and summarised across baseline-referenced intervals (T1-T2, T1-T3, T1-T4).

Overall P% did not change from T1-T4 (p=0.527). P% was higher on oral than vestibular surfaces (p<0.001), with no evidence that time trends differed by surface or tooth group (all interactions p≥0.315). Trajectories were similar across baseline P% tertiles (time×tertile p=0.794). Loss and Gain areas were stable across intervals (Loss p=0.435; Gain p=0.343) and did not differ by baseline tertile (p≥0.327). Stable areas decreased from 38.3% (T1-T2) to 34.9% (T1-T3) and 33.8% (T1-T4) (interval p<0.001).

Across multiple analytical perspectives, there was little evidence that study participation produced clinically relevant changes in toothbrushing-related plaque outcomes.

Over short follow-up, toothbrushing-related plaque outcomes appeared largely stable despite repeated study participation, providing context for interpreting longitudinal plaque measurements in oral-hygiene research.