Identifying predictors of personal exposure to air temperature in peri-urban India

Our plain-English summary of Identifying predictors of personal exposure to air temperature in peri-urban India, by C. Milà, A. Curto, A. Dimitrova, et al (The Science of the Total Environment, 2020).

The study

• Authors: C. Milà, A. Curto, A. Dimitrova, V. Sreekanth, S. Kinra, J.D. Marshall, C. Tonne
• Journal: The Science of the Total Environment
• Published: 2020;707:136114
• Cited by: 16 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

Characterizing personal exposure to air temperature is critical to understanding exposure measurement error in epidemiologic studies using fixed-site exposure data and to identify strategies to protect public health. To date, no study evaluating personal air temperature in the general population has been conducted in a low-and-middle income country. We used data from the CHAI study consisting of 50 adults monitored in up to six non-consecutive 24 h sessions in peri-urban south India. We quantified the agreement and association between fixed-site ambient and personal air temperature, and identified predictors of personal air temperature based on housing assessment, self-reported, GPS, remote sensing, and wearable camera data. Mean (SD) daytime (6 am-10 pm) average personal air temperature was 31.2 (2.6) °C and mean nighttime (10 pm-6 am) average temperature was 28.8 (2.8) °C. Agreement between average personal air and fixed-site ambient temperatures was limited, especially at night when personal air temperatures were underestimated by fixed-site temperatures (MBE = -5.6 °C). The proportion of average personal nighttime temperature variability explained by ambient fixed-site temperatures was moderate (R2mar = 0.39); daytime associations were stronger for women (R2mar = 0.51) than for men (R2mar = 0.3). Other predictors of average nighttime personal air temperature included residential altitude, ceiling height, and household income. Predictors of average daytime personal air temperature included roof materials, GPS-tracked altitude, time working in agriculture (for women), and time travelling (for men). No biomass cooking, urban heat island, or greenspace effects were identified. R2mar between ambient fixed-site and personal air temperature indicate that ambient fixed-site temperature is only a moderately useful proxy of personal air temperature in the context of peri-urban India. Our findings suggest that people living in houses at lower altitude, with lower ceiling height and asbestos roofing sheets might be more vulnerable to heat. We also identified households with higher income, women working in agriculture and men with long commutes as disproportionately exposed to high temperatures.

Copyright © 2019. Published by Elsevier B.V.

In plain English

This study measured the air temperature people actually experienced. It did not study asbestos exposure or disease; asbestos appears only as a roofing material. The authors analysed 227 sessions from 50 adults in peri-urban south India who carried a temperature sensor for up to six 24-hour periods (May 2015 to February 2016). They compared these readings with a fixed outdoor monitor. At night, the fixed site read on average 5.6 °C below personal temperatures. By day, women with asbestos-sheet bedroom roofs had personal temperatures 1.34 °C higher than women under tile or grass roofs (95% CI 0.21 to 2.47). For men the estimate was 0.91 °C, with a confidence interval (-0.21 to 2.04) including zero. The authors believed roof type marked housing quality, since concrete-roofed houses had fans and white-painted ceilings. At night, higher household altitude and taller bedroom ceilings were linked to lower personal temperatures, as was higher GPS-tracked altitude for men by day. The authors concluded that fixed-site temperature is only a moderately useful proxy for personal temperature. They named low sensor accuracy as the main limitation.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Panel study with repeated 24-hour personal temperature monitoring sessions per participant, compared with a fixed-site ambient monitor and analysed with linear mixed models
• Who was studied: 50 adults from the CHAI panel (a subsample of the APCAPS cohort) in a peri-urban area south of Hyderabad, Telangana, India; 227 of 271 monitoring sessions were included after excluding sessions with non-compliance
• Exposure: Personal air temperature, measured by a sensor in a MicroPEM device worn on a rucksack. Predictors included housing features such as bedroom roof material (tiles/grass, asbestos sheets, concrete), altitude, income, occupation and time-activity from GPS, diaries and wearable cameras
• Main results:
  • Mean personal versus ambient fixed-site temperature: daytime 31.2 (SD 2.6) °C personal vs 30.3 (SD 3.5) °C ambient; nighttime 28.8 (SD 2.8) °C personal vs 23.3 (SD 3.7) °C ambient
  • At night, fixed-site temperatures underestimated personal temperatures (mean bias error -5.6 °C)
  • Fixed-site temperature explained a moderate share of variation in nighttime personal temperature (marginal R2 0.39). For daytime it explained more in women (0.51) than in men (0.30)
  • Nighttime: each 100 m higher household altitude was associated with personal temperature -1.17 °C (95% CI -1.87 to -0.46). Each 1 m greater bedroom ceiling height was associated with -1.61 °C (95% CI -3.08 to -0.15)
  • Nighttime: household income above 15,000 Indian rupees (vs below) was associated with 0.73 °C higher personal temperature (95% CI 0 to 1.47)
  • Daytime, women: an asbestos-sheet bedroom roof (vs tiles/grass) was associated with 1.34 °C higher personal temperature (95% CI 0.21 to 2.47). For men the estimate was 0.91 °C (95% CI -0.21 to 2.04)
  • Daytime, men: each 100 m higher GPS-tracked altitude was associated with -1.47 °C (95% CI -2.38 to -0.56). Each hour of GPS-recorded trips was associated with 0.21 °C higher temperature (95% CI 0.02 to 0.39)
  • No urban heat island, biomass cooking or greenspace effects were identified
  • The authors interpreted the asbestos-roof finding as reflecting housing quality: concrete-roofed houses had fans, ventilators and white-painted ceilings
• The authors' conclusion: The authors concluded that ambient fixed-site temperature is only a moderately useful proxy of personal air temperature in peri-urban India. They suggested that people in houses at lower altitude, with lower ceilings and with asbestos roofing sheets might be more vulnerable to heat
• Limitations the authors state:
  • The personal temperature sensor had low accuracy (plus or minus 3 °C) and a slow response time
  • The effect of the sensor lag on the results is difficult to predict, even after averaging
  • The authors knew who owned air conditioning but not whether it was used during monitoring
  • There were no data on personal radiation, humidity or air velocity

Related studies

• Quality assessment and public health status of harvested rainwater in a peri-urban community in Edo State of Nigeria — I.H. Igbinosa, I.T. Aighewi (2017)
• Impact of climate change on temperature variations and extrinsic incubation period of malaria parasites in Chennai, India: implications for its disease transmission potential — P.K. Kripa, P.S. Thanzeen, et al. (2024)
• Monitoring of airborne asbestos fibers in an urban ambient air of Shahryar City, Iran: levels, spatial distribution, seasonal variations, and health risk assessment — F. Taghizadeh, A.J. Jafari, et al. (2019)
• Does the roof type of a house influence the presence of adult Anopheles stephensi, urban malaria vector? - evidence from a few slum settings in Chennai, India — S. Ravishankaran, A. Asokan, et al. (2022)
• Indoor radon levels in urban Hyderabad area, Andhra Pradesh, India — M. Sreenath Reddy, P. Yadagiri Reddy, et al. (2008)
• Indoor air quality differences between urban and rural preschools in Korea — C. Yoon, K. Lee, et al. (2011)

Browse more research in The Science of the Total Environment, or return to medical research.