Table 1.

Recent studies categorized by type of climate change-related event.

Exposure Category	Study	Sample * N (% Female) Age	Study Location	Study Design and Period	Exposure (s) of Interest	Mental Health Outcome: Measure(s)	Major Findings
Acute
Dust storm	Lee et al. (2019) [34]	N = 30,704 (33.8%)	Republic of Korea: Seoul	Ecological 2002–2015	Dust storms Duration and intensity	Completed suicide	Suicide counts during Asian dust storm (ADS) days were significantly higher than those during non-dust storm days. Exposure to ADS was associated with a 13.1% (95% CI, 4.5–22.4, p = 0.02) increase in suicide risk on an ADS day. For every day increase in duration, an additional 7.7% (95% CI, 2.9–12.8, p = 0.02) risk in completed suicide.
Flood	Mulchandani et al. (2020) [35]	N = 819	UK: England	Longitudinal 2015–2018	Flooding Exposure and experience	Depression: PHQ-2 Anxiety; GAD-2 PTSD: PCL-6	Approximately 5.7% of participants reported symptoms of probable depression, 8.1% probable anxiety and 11.8% probable PTSD, with prevalence of all mental disorders higher in the flooded versus unaffected group. 9.4%of participants reported persistent damage to their homes. A significant reduction in prevalence for all probable mental health diagnoses was observed in flooded group over three years.
	Thomas et al. (2021) [36]	N = 171 (69%) Age: 18–65 y	India: Karnataka	Cross-sectional December 2018-January 2019	Flooding Exposure and experience	Presence of psychiatric symptoms (depression, anxiety, somatic disorders, sleep problems and substance misuse)	After experiencing a flood, 66.7% of the sample showed the presence of symptoms consistent with at least one psychiatric problem. This, however, was not statistically significant. Depression and anxiety were the most prevalent symptoms, 33% and 31%, respectively. A statistically significant correlation with property damage or destruction of houses and psychiatric symptoms was observed.
Fire	Bryant et al. (2014) [37]	N = 1017 (60.3%) Age: ≥18 y at time of fire (M = 53.1 y)	Australia: Victoria	Cross-sectional 2011–2013	Victorian Black Saturday Bushfires (February 2009–March 2009) Exposure and experience	Depression: PHQ-9 PTSD: PCL Non-specific psychological distress: K6	Rates of any disorder significantly differed in high-, medium- and low-affected areas at 43.1%. 33.3%, 27.5%, respectively (p < 0.001). Probable PTSD and depression in 15.6% and 12.9% of subjects in high-affected areas, respectively. Both significantly more than medium- and low-affected areas (p < 0.01) Death of someone, fear for one’s life and subsequent major life stressors were predictors for PTSD, depression and serious mental illness. Additional predictors for fire-related PTSD included female sex and lower educational level.
	Bryant et al. (2018) [38]	N = 735 (61.5%) Age: ≥18 y at time of fire (M = 53.5 y)	Australia: Victoria	Longitudinal 2011–2014 T1: December 2011–January 2013 T2: July 2014-November 2014 (5 y PD)	Victorian Black Saturday Bushfires (February 2009–March 2009) Exposure and experience	Depression: PHQ-9 PTSD: PCL Non-specific psychological distress: K6	At 5 years post-disaster, rates of symptoms decreased for PTSD (8.7% vs. 12.1%), depression (9.0% vs. 10.9%) and serious mental illness (5.4% vs. 7.8%) but remained higher than national levels. Extent of recent life stressors was the most robust predictor of later development of firerelated PTSD (OR 2.11; 95% CI [1.22, 3.65]) and depression (OR 2.86; 95% CI [1.74, 4.70]).
Hurricane	Bozick (2021) [39]	N = 5694 (57.4%) Age: ≥18	USA: Texas	Observational w/a probability sample 2017–2018 T1: Pre-Harvey T2: Post-Harvey	Hurricane Harvey (August 2017) Exposure and experience	Number of days in the past 30 days with poor mental health	Statistically significant differences between the pre- and post-Harvey samples in both poor physical and poor mental health. Rate of experiencing days with poor mental health is 32% higher in post-Harvey sample compared with the pre-Harvey sample (95% CI, p < 0.01). Following the storm, adults experienced an increase of 1.31 days a month of poor mental health (5.38 days total), equating to an additional 12,371 person-years of poor mental health in the span of one month.
	Garfin (2022) [40]	T1: N = 1637 (54.6%) Age: ≥18	USA: Florida	Longitudinal 2017–2018 T1: Pre-Irma T2: 1 month PD T3: Post-Michael	Hurricane Irma (September 2017); Hurricane Michael (October 2018 - November 2018) Effects of repeated exposures	Posttraumatic stress symptoms (PTSS) Global distress: BSI-18	Prior mental health conditions (b, 0.18; 95% CI, 0.07–0.28), prior hurricane-related loss and/or injury (b, 0.09; 95% CI, 0.02–0.17), hours of Irma-related media exposure (b, 0.03; 95% CI, 0.02–0.04), being in an evacuation zone and not evacuating (b, 0.14; 95% CI, 0.02–0.27) and loss and/or injury after Hurricane Irma (b, 0.35; 95% CI, 0.25–0.44) were positively associated with post-traumatic stress symptoms after Hurricane Irma. Similar results were observed with Hurricane Michael-related exposures. After Hurricane Michael, prior mental health ailments, posttraumatic stress symptoms related to hurricanes Irma and Michael were associated with overall functional impairment.
	Lai et al. (2013) [41]	N = 277 (52%) Age: Grades 2–4 (M = 8.7 y)	USA: Texas	Longitudinal May 2009–December 2009 T1: 8 months PD T2: 15 months PD	Hurricane Ike (September 2008) Exposure and experience	Depression: CDI PTSD: PTSD-RI-R	At 8 months post-disaster, 23% of children had clinically significant PTSD symptoms, 43% had clinically significant depressive symptoms and 10% had comorbid PTSD and depressive symptoms At 15 months post disaster, these decreased to 14%, 18% and 7% respectively. 33% of those with comorbid symptoms at 8 months continued to have comorbid symptoms at 15 months. Children with comorbid symptoms had poorer recovery, more severe symptoms and reported greater exposure and recovery stressors
	Orengo-Aguayo et al. (2019) [42]	N = 96,108 (50.3%) Age: Grades 3–12	Puerto Rico	Cross-sectional February 2018–June 2018 (5–9 months PD)	Hurricane Maria (September 2017) Exposure to hurricane-related stressors	Depression: NCTSN-HART PTSD: NCTSN-HART	Clinically significant symptoms of PTSD in 7.2% of children; more in girls (8.2%) than boys (6.1%; p < 0.001). Girls also displayed higher depression scores than boys (p < 0.001). Approximately 30% of subjects perceived their lives to be at risk during the disaster, with girls experiencing this more than boys (34.2% vs. 27.3%; p < 0.001). Exposure to hurricane-related stressors moderately correlated with PTSD symptoms. Ongoing loss and disruption, geographical location and SES were not.
	Schwartz et al. (2017) [43]	N = 130 (77.7%) Age: 18–92 y (M = 49.73 y)	USA: New York	Longitudinal T1: 11–28 months PD (M = 14.5 months) T2: 1 y after T1	Hurricane Sandy (October 2012–November 2012) Personal and/or property exposures	Anxiety: PHQ-4 Depression: PHQ-4 PTSD: PCL-S	Overall, symptom prevalence decreased from T1 to T2. This was significant for anxiety (50.0% to 41.5%; p < 0.01) and PTSD (29.2% to 24.8%; p = 0.001) but not depression (35.4% to 30.8%; p < 0.39). At T2, experiencing personal damage, property damage, or both was positively associated with PTSD symptoms, but not anxiety or depression symptoms. Strongest predictor of experiencing a mental health symptom at T2 was having symptoms at T1.
	Sullivan et al. (2013) [44]	N = 498 (0%) Male veterans only Age: 18–60 y	USA: Louisiana	Cross-sectional 2007–2008 (2.5 y PD)	Hurricane Katrina (August 2005) Stressful hurricane-related events	Depression: PHQ-2 Anxiety: GAD-7 Panic disorder: PHQ PTSD: SPRINT	Compared to veterans without (N = 250) preexisting mental disorders, those with preexisting disorders (N = 248) were almost 7 times more likely to screen positive for any new mental disorder. Preexisting PTSD had the highest risk for new mental illness, followed by schizophrenia then affective disorders Of veterans with preexisting mental disorders, 72% screened positive for new mental disorder. Almost a third of those without preexisting mental illness screened positive for at least one of mental disorder.
Multiple Exposures	Edwards et al. (2021) [45]	N = 4592 (unknown) Age: 10 y	Philippines	Longitudinal 2016–2017 (15 y PD)	Cumulative disaster exposure (tropical cyclones, extreme rainfall, drought, volcanic activity, storm surges, sea level rises, flooding, tsunamis, earthquakes, fire, marine pollution) Hazard occurred in last 3 years. (Y/N)	Food insecurity: FIES; Stunting: WHO Child Growth Standards; Caregiver stress: PSS; Caregiver depression CES-D; Family violence, witnessing violence, child physical abuse (Y/N)	On average, children were exposed to 2.87 disasters from 2014 to 2017. 41% of children experienced family violence and 33% of children reported witnessing violence in the past 12 months. 30% of children reported being physically hurt by a parent, with 20% being hurt by parents in a forceful manner. 35% of children were stunted and 35% of caregivers report severe food insecurity. Exposure to a greater number of disasters were associated with an increased chance of children witnessing violence, being hurt by an adult, hurt by a parent and hurt by a parent forcefully.
	Graham et al. (2019) [46]	N = 7525 (59.3%) Age: ≥16 y	UK	Cross-sectional 2013–2015	Severe weather (wind, rain, snow, flooding) Home damaged by severe weather event in last 6 months? (Y/N)	Common mental disorder: CIS-R PTSD: PLC-C Suicidal thoughts, non-fatal suicide attempts and self-harm by questionnaire	Subjects who lived in storm or flood-damaged homes in 6 months prior to interview had significantly increased risk of having a common mental disorder (p < 0.01), suicidal ideation (p < 0.01) and ever having attempted suicide (p < 0.01). Associations with PTSD (p = 0.06) and with recent use of mental health treatment/services was not significant (p = 0.06)
Subacute
Drought	Luong et al. (2021) [47]	Baseline: N = 2607 (59%) Pooled cross-sectional: N = 6519 (61%) Age: ≥18 y	Australia: New. South Wales	Ecological/longitudinal 2007–2013	Drought HDSI (measures agricultural drought); SPEI (measures meteorological drought)	Psychological distress: K10	Significant non-linear relationship between drought exposure and psychological distress (HDSI; p < 0.001; SPEI; p < 0.001). Results show an inverted Ushape, where drought initially correlates with increased psychological distress and the turning point (where distress begins to decrease) occurs around 3 years after drought.
Heat	Lee et al. (2018) [48]	N = 166,579 ED admissions for mental disease Age: –	Republic of Korea	Ecological 2003–2013	Heatwaves Temp (daily mean), humidity, total solar radiation	ED admissions related to mental health	Strongest association between mental health diagnoses (specifically anxiety, schizophrenia, dementia and depression) and high temp occurred within 0–4 days of high temp exposure. 14.6% of emergency admissions for mental disease were attributable to extreme hot temperatures, with anxiety the highest risk (31.6%) and elderly more susceptible (19.1%)
	Dang et al. (2022) [49]	N = 7780 Age: −18 y	Vietnam: Ho Chi Minh	Cross-sectional 2017–2019	Heatwaves Combination of intensity (≥97th percentile of the daily mean temp, 30.9 °C) and duration (≥2 continuous days)	Admissions for psychiatric illness to Ho Chi Minh Mental Health Hospital	Heatwaves increased allcause psychiatric hospitalization by 62% (95% CI, 36–93%) for the main effect and by 8% (95% CI, −3 to 19%) for added effect of heatwaves. The group age 18–60 was primarily affected by the main effect of the heatwave, while the group aged 61 and older was affected by added effects of the heatwave. Hospitalizations due to psychoactive substance use was significantly affected by the main effect of heatwaves (RR: 2.21; 95%CI:1.55–3.15) and psychotic disorders were highly vulnerable to the main and added effects of heatwaves with RR = 1.50 (95% CI, 1.20–1.86) and RR = 1.14 (95% CI, 1.01–1.30), respectively.
	Ngu et al. (2021) [50]	–	60 countries worldwide did	Ecological	Heatwaves Temperature and/or relative humidity	Completed suicide	In countries with a significant result, an increase in suicide of 3.5% is observed for every unit increase of heatwave counts. Around half the countries showed a significant increase in suicide with respect to relative humidity, with the per unit increase of relative humidity ranging from −6 to +4.9%.
	Yoo et al. (2021) [51]	N = 2.8 million (39.4%)	USA: New York State	Ecological 2009–2016	Heatwaves (27.07 °C and above)	ED visits for specific mental disorders	Acute exposure to high heat (27.07 °C and above) increased ED visits for mental disorders (substance use, mood and anxiety disorders, schizophrenia and dementia). No statistically significant difference in subgroups in the sample being more vulnerable to the effects of heatwaves.
Chronic
	Basu et al. (2018) [52]	N = 219,942 ED visits (44%)	USA: California	Ecological 2005–2013	Temp (daily mean, maximum, minimum)	ED visits related to mental health, external-cause injuries	During the warm season, a 5.6 °C (10 °F) increase in same-day mean apparent temperature was associated with 4.8%, 5.8% and 7.9% increases in the risk of visits for mental health disorders, self-injury/suicide and intentional injury/homicide, respectively. Greatest risk: Hispanics, whites, persons aged 6–18 years and females
	Bundo et al. (2021)[53]	N = 89,996 (50.1%) Age: ≥18 y	Switzerland: Bern	Ecological 1973–2017	Temp daily mean	Psychiatric hospitalizations	For every 10 °C increase in daily mean temperature, the risk of hospitalization due to mental disorders increased linearly by 4.0% (RR: 1.04; 95% CI 1.02, 1.07), even while controlling for air pollution and other meteorological factors. Greatest risk: diagnoses of schizophrenia or developmental disorders,
	Burke et al. (2018) [25]	N(USA) = 851,088 N(Mexico) = 611, 366 Age: –	USA, Mexico	Ecological US: 1968–2004 Mexico: 1990–2010	Temp monthly mean, monthly precipitation data	Completed suicide	For a 1 °C rise in monthly mean temp, 0.6% increase in suicide rates in US counties and 2.1% in Mexican municipalities. Effect similar in hotter vs. cooler regions. Unlike all-cause mortality, suicide increased at hotter temperatures, decreased at colder temperatures and the effect has not decreased over time or with increasing income or the adoption of air conditioning.
	Carleton et al. (2017) [54]	-	India	Ecological 1956–2000	Temp daily mean	Completed suicide	For days above 20 °C, a 1 °C increase in a single day’s temp during the growing season increased annual suicides by 0.008 per 100,000 people. Significant effect only seen during India’s growing season; no identifiable impact in non-growing season.
	Middleton et al. (2021) [55]	N = 5373 (unknown) Age: –	Canada: Newfoundland and Labrador	Ecological 2012–2018	Temp Critical threshold range; daily mean temperature	Mental health-related visits, including suicide related-visits	Daily mean temperature ranged between −31.4 °C to 26.8 °C and critical threshold range identified as −5 °C to 5 °C, with 17 days as the longest consecutive number of CTR days in a row. Incidence of mental health-related visits was significantly higher when the cumulative average of daily temperatures two weeks prior was warmer (above −5 °C) compared to cold temperatures (less than −5 °C).
	Mullins et al. (2019) [56]	N(ED) = 8294 N(Suicides) = 2,096,460 N(Self-reported mental health) = 4,120,514	USA: California	Ecological ED visits: 2005–2016 Suicide: 1960–2016 Self-reported mental health: 1993–2012	Temp daily mean, daily precipitation, humidity, daily sunlight	ED visits related to mental health Self-reported mental health status: BFRSS Suicide rates	Higher temp increased ED visits for mental illness, suicides and self-reported days of poor mental health No evidence of adaption: temp relationship stable across time and despite factors such as baseline climate, air conditioning penetration rates and accessibility of mental health services
	Page et al. (2018) [5]	N = 22,562 Age: ≥65; ≤65	United Kingdom	Cross-sectional 1998–2007	Temp daily mean	Mean daily count of deaths Diagnosis of psychosis, dementia or substance use	Individuals with a diagnosis of psychosis, dementia, or substance misuse had a 4.9% higher risk of death (95% CI [2.0–7.8]) per 1 °C increase in temperature above the 93rd percentile of the annual temperature. Younger individuals with a primary diagnosis of substance use disorder displayed the highest mortality risk.
	Tlihonen et al. (2017) [57]	N = 551,529 violent crimes	Finland	Ecological 1996–2013	Temp Monthly mean	Violent Crime (proxy for aggression)	A strong correlation was observed between the monthly violent crime rate and monthly mean ambient T (r = 0.51, p < 0.001). Ambient temp explained 10% of variance in violent crime, with 1.7% increase per 1 °C.
		N = 51 (0%) Males only - n(violent) = 33 - n(healthy) = 18 Age: M = 33.4 y	Finland	Cross-sectional 1996–1997	Temp Monthly mean	Peripheral serotonin transporter density	Among healthy group, temp correlated significantly with peripheral serotonin transporter density (r = −0.64, p = 0.0025), which itself correlated negatively with the incidence of violent crime (r = −0.56, p = 0.06). This was even stronger among violent offenders (r = −0.68, p = 0.02.), suggesting mediation by serotonergic system. No offenders suffered a DSM axis-1 mental disorder aside from substance use and none were on any serotonin-altering drugs
	Vida et al. (2012) [16]	N = 347,552 ED visits related to mental health Age: ≥15 y	Canada: Québec	Ecological 1995–2007	Temp daily mean, relative humidity	ED visits related to mental health	ED visits increased with increasing mean temp, with effects most notable in metropolitan and suburban areas. Incidence ratio risks increased with increased temp. At 22.5 °C (72.5 °F) and 25 °C (77.0 °F), the number was usually significantly higher than average. Visits increased with humidity in the younger age group.
	Xue et al. (2019) [58]	N = 21,543 adults Age: –	China	Difference-in-difference study 2010–2014	Temp Long-term level of temp, temp variability	Self-reported mental health scores: Center for Epidemiologic Studies Depression Scale test	A 1 °C increase in temp variability (e.g., SD of daily temp within a calendar year) was correlated with 15% risk of decreased mental health score. Effects were strongly linked to a higher probability of feeling nervous, upset, hopelessness and meaninglessness. No significant association between annual temperature mean and mental health score.

^*% Female and age data noted if available for sample population. Abbreviations: CI = confidence interval; ED = emergency department; M = mean; OR = odds ratio; PD = Post-disaster; PTSD = Post-traumatic stress disorder; RR = relative risk; SES = socioeconomic status; SD = standard deviation; UK = United Kingdom; US = United States of America. Abbreviations-Measures: AAQ = Anger Attacks Questionnaire; BRFSS = Behavioral Risk Factor Surveillance System; BSI-18 = Brief Symptom Inventory-18; CES-D = Centre for Epidemiological Studies Depression Scale; CDI = Children’s Depression Inventory; CIS-R = Clinical Interview Schedule-Revised; DISC-IV = Diagnostic Interview Schedule for Children IV; FIES = Food Insecurity Experience Scale; GAD-7 = Generalized Anxiety Disorder, 7 item; GAI = Geriatric Anxiety Inventory; HDSI = Hutchinson Drought Severity Index; ICD = International Classification of Diseases; IES-R = Impact of Scale-Revised; K10 = Kessler Psychological Distress Scale, 10 item; K6 = Kessler Psychological Distress Scale, 6 item; NCTSN-HART = National Child Traumatic Stress Network Hurricane Assessment and Referral Tool; PCL = PTSD Checklist; PCL-S = PTSD Checklist-Specific; PHQ = Patient Health Questionnaire; PSS = Perceived Stress Scale; PTSD-RI-R = Post-traumatic Stress Disorder-Reaction Index, Revision 1; PTSS = Post-traumatic Stress Symptoms; SPEI = Standardized Precipitation Evapotranspiration Index; SPRINT = Short PTSD Rating Interview.