Sensitive thresholds of indoor temperature-humidity on arthritis in older adults: A year-round cohort time-series study with PELT change-point detection

Abstract

Sudden temperature and humidity shifts may trigger or worsen arthritis symptoms in older adults. However, indoor environmental exposure thresholds remain unclear. This study followed 28 elderly adults with osteoarthritis in Chongqing for one year, recording joint symptoms and indoor climate conditions. We used distributed lag non-linear models (DLNM) and generalized linear models (GLM) to assess how eight indoor environmental indicators—mean temperature (Tmean), mean humidity (Hmean), diurnal temperature range (DTR), diurnal humidity range (DHR), temperature change between neighboring days (TCN), humidity change (HCN), temperature variability, and humidity variability (HV)—affect symptoms over time. We used the Pruned Exact Linear Time (PELT) algorithm to detect change points and define three-level sensitivity thresholds. Results showed that rapid or unusual changes in indoor temperature and humidity increased joint pain risk. Risks were highest in autumn and winter when indoor conditions fluctuated sharply. Temperature-related symptoms appeared mostly within three days; humidity-related effects within four days. We identified the following acceptable sensitivity thresholds (Level Ⅲ): Tmean [14.6 °C, 32.4 °C]; DTR <4.4 °C; TCN [-2.2 °C, 2.5 °C]; TV <4.6 °C; Hmean [47.0 %, 81.4 %]; HCN [-8.1 %, 9.1 %]. DHR and HV showed no clear limits. Fluctuations beyond these values were more likely to worsen symptoms. Levels Ⅰ and Ⅱ exhibit progressively narrower threshold ranges, with Level Ⅰ representing the optimal zone in which symptom risk is minimal. These findings provide clear and quantified reference points for indoor climate control to help manage arthritis symptoms in older adults.