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

month, we distributed the target sample of 3,757 people over 12 months; that is, our monthly target sample size is 313 people per month [i.e. (3,757 people) / (12 months) = 313 people per month] or 24 people per province/territory per month [i.e. (313 people per month) / (13 prov- inces or territories) = 24 people per province or territory per month]. Sampling procedure. A multi-stage, multi-stratified random probability sampling method will be used to obtain a sample representative of Canada (Fig 1). Participants will be randomly selected via a two-stage approach. First, households will be randomly selected and recruited either by an addressed letter or a telephone call (landlines and cellular). Then, within each randomly selected household, the person who is 13 years old or older and has the next birthday will be invited to participate in the survey. Table 2. A summary of the questionnaire sections, with a description of the items in the section, an overview of the purpose of the section, and information on how the items were developed. Questionnaire Section Goal of Questions Measures Validity 1) Past experiences of climate change To capture information about individual experiences of climate change impacts, which can be a precursor to subclinical mental health outcomes. 2 x 5-point Likert scale questions capturing direct and indirect exposures to climate change. We selected two of the three questions assessing personal experiences of climate change validated by Clayton and Karazsia [33]. 2) Climate-related emotions To capture a range of subclinical

outcomes and emotions that can be experienced in response to climate change. 13 x 5-point Likert scale questions capturing emotional responses to climate change (i.e. sad, angry, grief, helpless, hopeless, worried, anxious, depressed, tense, concerned, stressed, scared, and powerless). Our questions cover the 12 emotions assessed by Searle and Gow [38], and we added an additional question about grief. 3) Past and current impacts, anticipatory impacts, and vicarious experiences To measure the drivers of respondents’ emotional reactions to climate change. 3 x 5-point Likert scale questions capturing the extent to which respondents’ reactions to climate change were related to past and current climate change impacts, anticipated future climate change impacts, or news and social media coverage. We developed new questions based on the identified pathways of ecological grief defined by Cunsolo and Ellis [25]. 4) Subclinical outcomes To measure cognitive and emotional impairment from climate change exposures. 8 x 5-point Likert scale questions capturing climate change impacts on concentration and rumination. We adapted the cognitive and emotional impairment questions in Clayton and Karazsia’s [33] climate change anxiety scale. This scale has been used and/or validated in subsequent studies by Cruz and High [39], Innocenti et al. [36], Larionow et al. [34], Mouguima-Daouda et al. [35], Simon et al. [40], and Joo Jang et al. [41]. To measure functional impairment from climate change exposures. 1 x 5-point Likert scale question capturing climate change impacts on daily life.

We adapted the five functional impairment questions from Clayton and Karazsia’s [33] climate anxiety scale to create one proxy question. Hickman et al. [42] also used this question as a proxy for functional impairment. To measure other climate-related worry, thinking, and distress. 5 x 5-point Likert scale questions capturing feelings of uncertainty in response to climate change. We selected and adapted questions from Reser et al.’s [43] 12-item, 6-point scale to address Wullenkord et al.’s [37] critiques of Clayton and Karazsia’s [33] climate anxiety scale. 5) Behavioural responses To assess whether climate change anxiety is associated with individual or collective climate actions. 3 x 5-point Likert scale questions capturing individual actions in response to climate change, and 3 x 5-point Likert scale questions capturing collective actions in response to climate change. We selected and adapted questions from Stanley et al.’s [44] 16-item, 100-point scale. One of these questions was similar in concept to Clayton and Karazsia’s [33] 5-item measure of sustainable behaviour. 6) Demographics To collect independent variables for use in subsequent analyses. 5 close-ended questions capturing age, gender identities, ethnic backgrounds, income, and postal code. Questions were based on the Alberta Human Rights Commission’s [45] description of gender identity, the Government of Canada’s Employment Equity Act’s definition of visible minority groups [46], and the Public Health Agency of Canada’s Foodbook Questionnaire (e.g. for income) [47].

https://doi.org/10.1371/journal.pone.0291303.t002 PLOS ONE Climate change and mental health survey protocol PLOS ONE | https://doi.org/10.1371/journal.pone.0291303 October 11, 2023 5 / 12