Estimating climate change and mental health impacts in Canada: A cross-sectional survey protocol

climate change anxiety was made up of two latent constructs: cognitive emotional impairment, which encompassed rumination, sleep-related concerns, difficulties concentrating, and crying, and functional impairment, which encompassed impairments in day-to-day functioning, including tasks at work or school. Some subsequent studies validated the original two-factor structure of the climate change and anxiety scale in other countries and languages [e.g., 34, 35], whereas others could not replicate the two-factor structure [e.g., 36, 37]. Wullenkord et al. (2021) could not reproduce the two-factor structure in the German population, leading them to question the construct validity of the climate change anxiety scale [37]. They recommended further scale development to encompass varying levels of intensity of climate change anxiety, as well as other related emotional experiences. Specifically, they argued for incorporating an emotional factor to measure feelings associated with anxiety (e.g., worry and fear) and for add- ing items that capture experiences of uncertainty that are traditionally associated with anxiety. Therefore, in our study, we incorporated Clayton and Karasizia’s (2020) climate change anxi- ety scale [33] as well as Searle and Gow’s (2010) [38] scale of climate-related emotions, and we also selected items from Reser et al. (2012) that measure feelings of unpredictability related to climate change [43] (Table 2). Survey instrument The survey consists of six sections: (1) self-reported past experiences of climate change; (2)

self-reported climate-related emotions; (3) self-reported past and current impacts, anticipatory impacts, and vicarious experiences of climate change; (4) self-reported subclinical mental health outcomes; (5) self-reported behavioural responses; and (6) demographics (Table 2). The questionnaire was written in English and translated into French and Inuktitut. Translations were back-translated into English to ensure accuracy. The questionnaire was designed to take approximately 10 minutes to complete over the telephone. The English version of the full ques- tionnaire is available in S1 Table. Over a one-week period, we used a convenience sample (n = 35) to pretest the online ques- tionnaire and a random sample (n = 20) to pretest the telephone questionnaire for length and content. We refined the survey based on estimated completion times as well as respondent feedback. The data collected during pretesting will not be included in the final dataset. Eligible participants The questionnaire will be administered in all provinces and territories in Canada. Youth and adults who are 13 years old or older and have a Canadian postal address or telephone number will be eligible to participate. The survey will be offered in English, French, or Inuktitut. Attempts will be made to accommodate participants that speak languages other than English, French, or Inuktitut on an as-requested basis. Sampling Sample size. There are no published prevalence estimates of climate change-related men- tal health outcomes that are representative of Canada. Therefore, we calculated our target sam-

ple sizes using an a priori estimate of 25% prevalence of subclinical outcomes related to climate change based on estimates from the United States of America [33] to calculate a target sample size of 289 people per province/territory. We used an allowable error of 5% and a confi- dence level of 95% to calculate a target sample size of 289 people per province/territory. Reflecting our objective to estimate prevalence by province/territory, our total target sample size is therefore 3,757 people [i.e. (289 people per province or territory) x (10 provinces + 3 ter- ritories) = 3,757 people]. Then, reflecting our objective to estimate the national prevalence by PLOS ONE Climate change and mental health survey protocol PLOS ONE | https://doi.org/10.1371/journal.pone.0291303 October 11, 2023 4 / 12