The neuropsychiatric toll of rising temperatures on women’s health in low-income and middle-income countries: a scoping review Authors: Richard G Künzel, Carrie G Wade, Bizu Gelaye, Michelle A Williams Published: 2025 ABSTRACT AbstractBackgroundGlobal warming poses a significant threat to human mental health, with disproportionate effects on vulnerable and highly exposed populations, such as women from low-income and middle-income countries (LMICs). This scoping review mapped available evidence on associations between high ambient temperature and psychiatric, neurologic and neurocognitive outcomes in women from LMICs.MethodsA systematic literature search was conducted in the electronic databases MEDLINE, Embase, Web of Science, Global Health, CINAHL and Global Index Medicus in December 2024, and updated in September 2025, without date restriction. Controlled vocabulary terms were used to identify relevant studies in the domains of ambient heat exposure, mental disorders/mental health, emotional regulation, sleep, pain and LMICs. 70 studies were included based on a priori defined eligibility criteria. A qualitative, narrative synthesis was conducted.ResultsWe found evidence suggesting positive associations between exposure to high ambient temperature and adverse psychiatric, neurologic and neurocognitive outcomes among women from LMICs. Associations of heat with psychiatric outcomes seem to be non-specific. Studies were heavily concentrated in China, with notable underrepresentation from Latin America, Africa and Southeast Asia, and many used overlapping data sources. Most studies analysed female subgroups within mixed-sex samples, rather than focusing exclusively on female samples.ConclusionsA critical research gap remains concerning women from LMICs. Comprehensive, methodologically rigorous research is essential to understand regional differences and the temporal pathways through which heat exposure affects women, particularly those from Africa and Latin America. Understanding the severe health consequences of global warming is imperative as populations with the least resources are the most affected.Keywords: Mental Health & Psychiatry, Neurology, Suicide, Review WHAT IS ALREADY KNOWN ON THIS TOPIC WHAT THIS STUDY ADDS Our scoping review identified 70 studies that examined associations between high ambient temperature and psychiatric, neurologic and neurocognitive outcomes in women. Our scoping review identified 70 studies that examined associations between high ambient temperature and psychiatric, neurologic and neurocognitive outcomes in women. Our results suggest associations between heat exposure and adverse psychiatric outcomes. We uncovered a marked concentration of evidence in China with few studies focusing exclusively on women. Our results suggest associations between heat exposure and adverse psychiatric outcomes. We uncovered a marked concentration of evidence in China with few studies focusing exclusively on women. Our findings highlight major knowledge gaps, especially for disorder-specific outcomes and underrepresented regions such as Africa and Latin America. Our findings highlight major knowledge gaps, especially for disorder-specific outcomes and underrepresented regions such as Africa and Latin America. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY Further comprehensive and inclusive research is urgently needed to understand regional differences and temporal trajectories in the links between heat and neuropsychiatric health. Further comprehensive and inclusive research is urgently needed to understand regional differences and temporal trajectories in the links between heat and neuropsychiatric health. Our findings can inform policy, guide research priorities and support targeted prevention and intervention strategies for populations most vulnerable and least equipped to adapt to climate change. Our findings can inform policy, guide research priorities and support targeted prevention and intervention strategies for populations most vulnerable and least equipped to adapt to climate change. Background Global climate change is one of the most pressing human challenges in the 21st century,1 and global warming is its primary driver.2 Over the past decades, the earth’s surface temperature has steadily increased3 and is expected to increase further.4 Despite humans’ ability to adapt to changing environments, the pace and extent of this temperature increase and the increasing frequency of concomitant heat events may exceed this ability,5 6 yielding direct consequences for human health. High ambient temperatures directly cause between 300 0007 and 500 0008 deaths annually. Additionally, high ambient temperature poses a direct risk to cardiovascular9 10 and respiratory disorders,11 among other conditions.12 Such findings have motivated the WHO to declare that ‘climate change presents a fundamental threat to human health’.13 While several studies have documented the adverse impact of high ambient temperature on physical health14, the impact on mental health remains less well understood.15 This knowledge gap is particularly concerning given the substantial global burden of psychiatric, neurological and neurocognitive disorders, which affect millions worldwide and represent leading causes of disability.16 The reasons for this knowledge gap are multifaceted and include that mental health effects often occur with a time lag and within a complex biopsychosocial system.17 Nevertheless, recent findings indicate that individuals exposed to heat face higher risks of morbidity and mortality from psychiatric disorders, including schizophrenia18 and suicide.19 Additionally, neurologic and neurocognitive disorders—such as sleep disturbances20 and pain21—have been associated with climate change-related high ambient temperatures. These neurologic conditions are particularly concerning given their frequent comorbidity with psychiatric disorders, amplifying the potential population health impact as global temperatures continue to rise.22 High ambient temperature differentially affects certain populations. For instance, elderly people consistently show the highest risk for heat-related mortality.23,25 However, these disparities may not necessarily apply to other outcomes. For example, young people face the highest suicide risk during heatwaves, likely due to a greater sensitivity to environmental influences during brain development.6 These findings highlight the complexity of heat exposure and the need for population-specific research.14 Less is known about how high ambient temperature affects women’s mental health,26 though limited studies suggest elevated risk6 and particular vulnerability.27 28 This vulnerability may stem from reduced thermoregulation capacity during physical activity in hot environments due to lower sweating rates, though evidence remains inconclusive.29 Additionally, pre-existing mental disorders, which are tentatively more prevalent among women,30 may contribute to disproportionate adverse effects of high ambient temperature on mental health.31 This vulnerability is of particular importance for low-income and middle-income countries (LMICs), where women often face greater exposure to high ambient temperatures due to informal work and domestic responsibilities.32 33 Coping mechanisms are often limited due to scarce resources and high costs of preventive measures,34 further disadvantaging LMIC populations despite their minimal contribution to global warming.34 However, research specifically investigating the association between high ambient temperature and psychiatric, neurologic and neurocognitive conditions among women from LMICs is extremely limited. We identified only seven studies with this focus.3233 35,39 For example, Lin et al36 studied n=1931 pregnant women from prenatal care clinics in Shanghai and found that cumulative exposure to extremely high temperatures (31.2°C–34.1°C until lag-day 2) was associated with an almost threefold higher odds for high mental stress (≥66.6th percentile, Global-Severity-Index) compared with women exposed to normal temperatures (20°C–25°C). Qualitative studies from Kenya33 35 37 and Burkina Faso32 reported higher levels of anxiety, sleep disturbances and irritability,32 35 as well as exhaustion and restlessness33 37 in relation to extreme heat. Additional studies have investigated sex-mixed samples from LMICs, but did not examine sex-specific differences38,40, suggesting that substantial evidence about women’s experiences may be hidden within the existing literature. To address this gap, we sought to review the available evidence of the association between high ambient temperature and psychiatric, neurologic and neurocognitive conditions in women from LMICs. Methods Protocol registration The protocol for this scoping review was registered on the Open Science Framework. 40 The scoping review was conducted using the methodological guidance of the Joanna Briggs Institute Manual for Evidence41 and is reported in accordance with the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses–Extension for Scoping Reviews.42 Eligibility criteria and literature search We searched the electronic databases MEDLINE (Ovid Technologies), Embase (Elsevier), Web of Science Core Collection (Clarivate), Global Health, CINAHL and Global Index Medicus to identify studies eligible for this scoping review. The search strategy included terms for (1) ambient heat exposure, (2) sleep, (3) mental disorders/ mental health, (4) emotional regulation, (5) pain and (6) LMICs using controlled vocabulary terms, if available. No date limits were applied. The search was carried out on 12 December 2024 and updated on 22 September 2025. The search strategy is outlined in online supplemental Table 1. Identified records were retrieved from electronic databases, collected in EndNote (V.20, Clarivate, 2023) and uploaded into the web-based software Covidence43 by one coauthor (CGW) for deduplication and screening. Records were then screened, first, based on title and abstract, and second, based on full text, by one coauthor (RGK), using a priori-defined eligibility criteria. Eligibility criteria were defined by two coauthors (RGK and MAW) using the Population, Concept and Context framework.44 Inclusion criteria required that reports must: (1) include a primary analysis focused on individuals classified as biologically female (hereafter referred to as women) or a stratified analysis that examined women separately, (2) include environmental heat exposure, defined as elevated thermal conditions originating from natural or climatic sources (including high ambient temperature, heatwaves and seasonal temperature increases), (3) include objective and/or subjective measures of psychosocial stress, emotion regulation, psychoneurological pain and/or sleep disturbance outcomes (outcome variable definitions are provided in online supplemental Table 1), (4) report on a sample from an LMIC (low income, lower middle income, upper middle income), as defined by The World Bank between 2013 and 202345, (5) be a primary research report (including prospective and retrospective cohort studies and observational studies, such as case- control studies, cross-sectional studies and qualitative studies) and (6) be written in English. Exclusion criteria required that reports must not: (1) report solely on biological males or use sex-unstratified samples without separate analysis for women, (2) report on animal samples, (3) include artificial, occupational or industrial heat exposure, (4) report on a sample from a high-income country as defined by The World Bank between 2013 and 202345 and (5) be a controlled trial, a conference abstract or proceeding, an institutional report, a review or meta-analysis. Information extraction After the study eligibility decision was determined, information on pre-defined variables was extracted by hand by one coauthor (RGK) using a Microsoft Excel spreadsheet.46 Predefined variables included: first author name, year of publication, title of publication, digital object identifier, keywords, type of publication, study design, study aim, year of data collection, study population, sample country, sample country income classification, collection setting, sample ethnic distribution, further sociodemographic description, sample size, full sample size, sample size of female stratum, sample age, female stratum age, exposure (category, definition/tool, assessment time), outcome (category, definition/tool, assessment time), results, results classification for female stratum (positive association, null result, negative association, miscellaneous notes). Data synthesis A qualitative, narrative synthesis was conducted. Studies meeting the inclusion criteria were analysed with a focus on study characteristics, including outcome types, exposure types and geographic locations (continent and country). Results of this scoping review are presented in graphical and tabular formats. Deviations from the protocol Contrary to what was stated in the protocol, we included studies investigating seasonality rather than direct environmental heat exposure if the authors reported significant temperature differences between seasons. This approach resulted in four additional studies being included in the review.47,50 We also included the outcome cognitive function, as it represents an important subclinical condition for many mental disorders.51 Results As shown in figure 1, the electronic database search yielded 15 583 records, of which 6648 were unique and thus were carried forward for title and abstract screening. Of these, 259 reports underwent full-text screening. After full-text screening, 70 studies were included in the scoping review, all representing separate reports. In total, the 70 separate studies reported on 77 distinct effect measure sizes. Study characteristics are outlined in table 1 and online supplemental Tables 2 and 3. Numbers used as denominators for percentages vary between results sections, as some studies contributed multiple outcomes or were conducted in multiple countries. See online supplemental Material for a full explanation. Identification of studies. Table 1References of included studies, clustered by key variablesCharacteristicsNumber of studies (%)ReferencesExposureAmbient temperature14 (20.00%) 3960 71 81 84 86 99 102 111 Heat or heat wave47 (67.14%)32 33 35,38 53,58 61,68 72 73 75,78 82 83 85 87,91 93,98 103 104 115,117Apparent temperature5 (7.14%) 52 59 70 74 92 Seasonality4 (5.71%) 47 50 Outcome*Mental disorders15 (19.48%) 5053 55 59 61 63 72 Suicide14 (18.18%) 47 49101 Depression13 (16.88%) 3852 54 64 66 67 70 71 81 85 Schizophrenia10 (12.99%) 5666 71 86 92 Mental health5 (6.49%) 36 39 65 68 80 Dementia3 (3.90%) 57 62 93 Anxiety3 (3.90%) 38 66 71 Cognitive function3 (3.90%) 94 96 Cognitive function3 (3.90%) 94 96 Sleep disturbances3 (3.90%) 32 33 35 Mental/emotional well-being2 (2.60%) 33 37 Sleep duration2 (3.60%) 58 97 Bipolar disorder1 (1.30%) 71 Sleep restlessness1 (1.30%) 98 Sleep apnoea1 (1.30%) 99 Headache and migraine1 (1.30%) 100 Region†East Asia46 (63.01%) 3649 52 54 56 59 61 62 64 Latin America9 (12.33%) 5360 68 72 101 112 115 Africa7 (9.59%) 32 33 35 37 38 97 115 Middle East/South Asia6 (8.22%) 39 47 48 100 102 111 Southeast Asia5 (6.85%) 50 55 63 115 Income class†Upper middle income61 (83.56%) 3647 49 52 Lower middle income10 (13.70%) 3335 37 39 50 55 63 115 Low income2 (2.74%) 32 97 Study design†‡Retrospective time-series study30 (41.10%) 5053 55 56 59 61 63 70 72 Retrospective cohort study2 (2.74%) 36 67 Prospective cohort study14 (19.18%) 5254 58 64 65 68 81 84 94 Observational panel study1 (1.37%) 97 Qualitative and ecological study5 (6.85%) 32 33 35 37 101 Case-control3 (4.11%) 49 99 100 Case-crossover design13 (17.81%) 57 62 66 71 75 76 79 93 104 115 Cross sectional5 (6.85%) 38 39 47 48 80 *This overview includes n=77 references, as the studies by Scorgie et al,33 Brown et al,38 Wu et al71 and Zhong et al66 reported several outcomes. †This overview includes n=73 references, as the study by Kim et al115 was conducted in four suitable countries, namely Brazil, Philippines, Vietnam and South Africa. ‡Study design was clustered: Qualitative study=qualitative and ecological study design; Case-control=case-control and prospective case-series studies. References of included studies, clustered by key variables This overview includes n=77 references, as the studies by Scorgie et al,33 Brown et al,38 Wu et al71 and Zhong et al66 reported several outcomes. This overview includes n=73 references, as the study by Kim et al115 was conducted in four suitable countries, namely Brazil, Philippines, Vietnam and South Africa. Study design was clustered: Qualitative study=qualitative and ecological study design; Case-control=case-control and prospective case-series studies. Characteristics of the studies Country and country income classification As indicated in table 1 and figure 2, most studies were conducted in mainland China (n=46; 63.01%), followed by Türkiye (n=5; 6.85%), Brazil (n=5; 6.85%), Vietnam (n=4; 5.48%), Kenya (n=4; 5.48%), Mexico (n=3; 4.11%), Burkina Faso (n=2; 2.74%) and Columbia, Philippines, Pakistan and South Africa (each n=1; 1.34%). Consequently, most studies were conducted in upper-middle-income countries (n=61; 83.56%), followed by lower-middle-income countries (n=10; 13.70%) and low-income countries (n=2; 2.74%). Characteristics of included studies. This overview includes n=73 references, as the study by Kim et al115 was conducted in four suitable countries, namely Brazil, Philippines, Vietnam and South Africa. Duration is displayed in years based on the start and end year reported in the respective study. Income refers to the income classification according to the World Bank.45 Low income=2.74%. Design was clustered: Cohort study=prospective cohort, retrospective cohort and observational panel studies; Qualitative study=qualitative and ecological studies; Case-control=case-control and prospective case-series studies. Study designs and durations In total, 10 different study designs were used. The most common study designs used were retrospective time-series designs (n=30; 41.10 %), followed by prospective cohort designs (n=14; 19.18 %), case-crossover designs (n=13; 17.80 %), qualitative/ecological designs (n=5; 6.85 %), cross-sectional designs (n=5; 6.85 %), case-control/case-series design (n=3; 4.11 %), retrospective cohort designs (n=2; 2.74 %) and observational panel designs (n=1; 1.37 %). As shown in figure 2, the time durations investigated varied widely across studies, ranging from several days or months to up to 20 years. Definition of the exposures As shown in online supplemental Table 2, studies defined exposure differently. Most commonly, studies used ambient temperature, the outdoor temperature collected by local weather stations and provided by national meteorological services, either as a continuous exposure or as the basis for exposure categorisation. For studies investigating broader areas, such as regions, states or countries, temperature data was often averaged through interpolation, for example by 0.1° × 0.1°.52 In other cases, temperature data was averaged to address ‘irregular distributions of weather stations’.53 Methods included applying inverse distance weighting within a 150 km radius54 or calculating city mean values of daily maximum temperature.55 Other studies used alternative measures such as diurnal temperature range and ‘temperature change between neighbouring days’.56 Some studies additionally incorporated minimum and maximum daily temperatures to capture excess heat or nighttime temperatures when particularly relevant to the outcome, such as for sleep studies.57,59 14 studies (20.00%) used ambient temperature as a continuous exposure. For example, da Silva et al60 investigated the association between daily mean temperature (compared with 22.4°C as the reference) on hospital admissions for mental and behavioural disorders. More commonly, ambient temperature was categorised to depict different levels of heat exposure, such as heat waves of varying severity or duration (n=47 studies, 67.14 %). While no commonly accepted definition exists,58 most studies combined intensity and duration to define heat waves, often resulting in multiple definitions per study. For instance, Zhou et al58 defined heatwaves using ‘the 90th, 92.5th, 95th and 97.5th percentiles of the daily minimum temperature … (across) … ≥ 2 …, ≥3 …, and ≥4 (consecutive) days,’ yielding twelve unique exposure definitions. Other studies used threshold approaches, such as the 99th percentile relative to reference values like ‘daily mean temperatures during the warm months’,61 or ‘daily mean temperature of the entire study period’.62 Dang et al63 differentiated between the ‘independent effect of daily mean temperatures (intensity effect)’ and the ‘added effect of prolonged heat wave days (duration effect)’. Additional approaches included counting days with high temperatures within specific periods,54 64 65 or ‘compound heat days’, defined by an extreme hot day followed by an extreme hot night.66 67 Five studies (7.14%), including four qualitative studies, assessed heat or ‘higher-than-usual-temperature’ exposure through self-report, primarily via in-depth interviews.32 33 35 37 68 Five studies (7.14%) used apparent temperature to measure heat exposure, including one study that used both ambient and apparent temperature.59 Apparent temperature incorporates average temperature, relative humidity and wind speed into a single index, providing a more accurate indicator of how the body perceives heat, particularly in hot and humid environments.69 Two studies used the Humidex, a similar measure based solely on ambient temperature and humidity.59 70 Lastly, four studies (5.71%) used seasonality to measure exposure. All investigated seasonal patterns of outcome (three examined suicidality), with substantial temperature differences between defined seasonal months—highest in June–August and coldest in November–January.47,50 Definition of the outcomes As shown in figure 3, 15 different outcome types were investigated across the three outcome categories. Psychiatric outcomes were most commonly investigated, encompassing eight different outcome types. In the following section, 77 unique effect measure sizes are synthesised, as the studies of Scorgie et al,33 Brown et al,38 Zhong et al66 and Wu et al71 reported on several outcomes and were therefore counted in multiple categories. Distribution of outcomes used in the studies. The most common outcome was mental disorders without further specification (n=15, 19.48 %). Mental disorders were primarily measured through emergency department admissions or emergency ambulance dispatches for mental or behavioural disorders.5359 72,74 Mental and behavioural health disorders were also measured through hospitalisation records from various sources, death records,75 including hospitals or clinics for inpatient50 55 63 and outpatient visits,61 76 77 insurance systems78 and health systems.60 79 A conceptually similar but distinct outcome was mental health (n=5, 6.49 %). Unlike mental disorders, which were measured through clinical records, mental health was assessed using symptom-level instruments. These included the Short-Form Health Survey-3680, the Symptom Checklist 9036, the Life Events Scale for Pregnant Women36, the Depression Anxiety Stress Scale39, the Patient Health Questionnaire 9 (PHQ-9) and Generalised Anxiety Disorder 7 (GAD-7) for measuring Common Mental Disorders,68 and single items assessing mental health.65 The second most commonly investigated psychiatric outcome was suicide (n=14, 18.18 %), which was measured exclusively through case or death registers from participating hospitals, counties, regions or countries, often using the corresponding code of the International Classification of Diseases (ICD-8, ICD-9 or ICD-10). Depression (n=13, 16.88 %), schizophrenia (n=10, 12.99 %), anxiety (n=3, 3.90 %) and bipolar disorder (n=1, 1.30 %) were the mental disorders that were investigated as single outcomes. Depression was primarily measured at the symptom level using the Centre for Epidemiologic Studies Depression Scale Revised-10 as part of larger epidemiological surveys,5254 64 67 81,83 the PHQ-9,38 84 or through clinical diagnoses of major depressive disorder using ICD-10 codes from hospital records.66 70 71 85 Schizophrenia was measured through ICD-10 diagnoses obtained from hospital records,5671 86,91 whereas two studies used emergency department admissions for schizophrenia as the operationalisation.66 92 Anxiety was measured either through ICD-10 diagnoses obtained from hospital records71 or given at emergency department visit,66 the GAD-7.38 Bipolar disorder was only investigated in one study and measured via ICD-10 diagnoses obtained from hospital records.71 Mental/emotional well-being was investigated in two studies (2.60 %) and was assessed through open-ended self-reports in qualitative in-depth interviews.33 37 Neurologic and neurocognitive outcomes included dementia, headache/migraine, sleep disturbances and cognitive function. Dementia (n=3, 3.90 %) was measured through ICD-10 codes obtained from death registers in all three studies.57 62 93 Cognitive function (n=3, 3.90 %) was assessed using the Mini Mental Status Examination94 95 or a test measuring the cognitive skills ‘remembering, understanding and applying’.96 Sleep disturbance (n=7, 9.09 %) was examined using different approaches. Three studies used open-ended self-reports from qualitative interviews,32 33 37 two studies focused on sleep duration, measured objectively using wearable devices,97 or through self-report via surveys.58 Two additional studies examined ‘restlessness’ via survey self-reports98 and sleep apnoea occurrence using smart watch/wristband devices,99 respectively. Lastly, one study (n=1, 1.30%) investigated episodic tension-type headache and episodic migraine diagnosed according to the International Classification of Headache Disorders-3.100 Relationship between exposure to high ambient temperatures and psychiatric, neurologic and neurocognitive outcomes Seven studies (10.00 %) investigated female-only samples. These consisted of four qualitative studies,32 33 35 37 two cross-sectional studies and38 39 one retrospective cohort study of pregnant women.36 The remaining 63 studies (90.00 %) investigated sex-mixed samples with stratified analyses by sex. Of these, 23 studies (36.51%) reported stratified analyses with statistical significance testing for group differences, while 40 studies (63.49%) conducted only descriptive stratification without statistical significance testing. Due to multiple effect measure sizes reported in the studies of Scorgie et al,33 Brown et al,38 Zhong et al66 and Wu et al,71 the following section will synthesise a total of 77 effect measure sizes drawn from 70 studies. As shown in figure 4, 47 effect measure sizes (61.04 %) were reported as statistically significant positive associations between exposure and outcome among women. By ‘positive association’, we mean that increased exposure levels (eg, higher ambient temperature, more heat days) were associated with worse outcome levels (eg, higher suicide rates, elevated depression scores, fewer hours of sleep). ‘Statistically significant’ refers to p<0.05, which was the standard used by most studies when reported. Results of qualitative studies were considered ‘significant’ if authors reported adverse associations between exposure and outcome. 28 effect measure sizes (36.36 %) found no statistically significant association between exposure and outcome. Only two studies reported improved depressive symptoms84 or cognitive function96 in association with temperature increases. Summary of the results for female strata of the included studies. This overview includes n=77 effect measure sizes, as the studies by Scorgie et al,33 Brown et al,38 Wu et al.71 and Zhong et al66 reported several outcomes. Significance level for a positive association was set to p<0.05, as this was the one used by all included studies. a Positive association refers to the association: increase in exposure is associated with increased adverse levels of the outcome. b Negative association. Descriptively, effect measure sizes that found statistically significant positive associations more commonly used heat wave exposure (70.59% vs 29.41%) compared with those using ambient temperature exposure (29.41% vs 70.59%). In the following sections, we report study results by outcome group. Relationship between heat exposure and psychiatric outcomes Mental disorders and mental health 15 studies investigated the association between high ambient temperature and mental disorders. Most (73.33%) used heat or heat wave as the exposure, and the majority employed a retrospective time series design (80.00%), with only three studies choosing a case-crossover design (20.00%). Geographically, nine studies were conducted in China (60.00%), followed by three each in Brazil and Vietnam (20.00%, respectively). Most studies (73.33%) found statistically significant positive associations. Among studies investigating heat exposure, effect measure sizes varied considerably, mainly due to different exposure definitions. Relative risk (RR) estimates ranged from RR=0.91 (95% CI 0.80 to 1.03, at 22.2°C (the 90th percentile) compared with 18.02°C across lag days 0–1053;) to RR=2.53 (95% CI 1.57 to 5.16, at 24.5°C (the 99th percentile) compared with 18.01°C across lag days 0–1072). OR estimates ranged from OR=1.057 (95% CI 0.972 to 1.145) for heatwave exposure vs no exposure across lag days 0–579 to OR=1.24 (95% CI 1.03 to 1.49, at 30°C (the 97.5th percentile) compared with 18°C (the minimum risk temperature) across lag day 0–976;). Two studies investigated the association between apparent temperature and mental disorders.59 74 For example, Yin et al59 examined the association between high Humidex (incorporating ambient temperature, relative humidity and wind speed) and mental disorders among n=10 823 participants from Shenzhen, China. The authors found that exposure to high Humidex (at 41.9°C (90th percentile), lag day 0–1, compared with 12.0°C (minimal risk exposure)) was associated with a 19.4% increased risk for emergency department admission due to mental and behavioural disorders (95% CI 1.071 to 1.330). Two studies investigated the association between warm season and ambient temperature on mental disorder: Trang et al50 examined n=5979 women from Hanoi, Vietnam, and found an 18% increased risk of hospital admissions for mental disorders in summer compared with winter (95% CI 1.03 to 1.369). Similarly, da Silva et al60 found a 19% increased risk (95% CI 1.15 to 1.23) for mental and behavioural disorder admission on the index day compared with the reference temperature (22.4°C) among female participants in Curitiba, Brazil. All studies conducted in Brazil reported a statistically significant positive association between high ambient temperature and either emergency department visits53 72 or hospitalisations for mental or behavioural disorders.60 Notably, all Brazilian studies were conducted in Curitiba, and two of three studies likely used overlapping data sources.53 72 Studies from other countries showed more heterogeneous results: 77.78% of Chinese studies found statistically significant positive associations, while 66.66% of Vietnamese studies found no statistically significant associations. Five studies investigated the association between heat exposure and mental health. 40% used a prospective cohort study design or a cross-sectional design, respectively, while 20% used a retrospective time-series design. All studies found statistically significant positive associations between heatwave exposure and worse mental health outcomes. For example, Lin et al36 reported a 2.9-fold increased odds (95% CI 1.9 to 4.6) for high mental symptoms when exposed to extremely high temperatures (95th percentile, 31.2°C–34.1°C) compared with optimal temperatures (20°C–25°C; lag day 0–1) among pregnant women in Shanghai, China. Suicide 14 studies investigated the association between high ambient temperature and suicide risk. Of these, 42.86% used ambient temperature as the exposure, and most employed a retrospective time series design (57.14%). Fewer studies used cross-sectional or case-crossover (14.29%, each) or ecological or case-control designs (7.14%, each). The study countries were more heterogeneous than in other outcome categories. Five studies originated from China, four from Türkiye, two from Mexico and one each from Colombia, Brazil and a multicountry study (Brazil, the Philippines, Vietnam and South Africa). A slight majority of studies (57.14%) found statistically significant positive associations between high ambient temperature and suicide risk. Among studies investigating exposure as a continuous measure, association estimates were mostly positive but non-significant. For example, Fernández-Niño et al101 reported that daily temperature was non-significantly associated with the number of suicides among Colombian women (incidence rate ratio 1.06, 95% CI 0.98 to 1.16, p=0.15), while Akkaya-Kalayci et al102 reported a statistically significant correlation of r=0.213 (p<0.001) between ambient temperature on the index day and recorded suicide attempts. Associations were substantially clearer for heat exposure. One study of n=16 801 Chinese women reported a non-significant 1.51-fold increased risk (95% CI 0.91 to 2.25) for suicide death when exposed to high temperature (95th percentile, 18.7°C–31.9°C, city-specific) compared with the monthly minimum temperature (14.4°C–27.7°C, city-specific).103 Another China-based study of n=4042 women found a 1.98-fold increased odds for suicide mortality (95% CI 1.02 to 3.83) when exposed to heat (at 30.6°C, the 95th percentile) compared with 7.9°C (the lowest measured temperature).104 Finally, three studies investigated seasonality and suicide risk. Two studies, conducted in Türkiye48 and China,49 found statistically significantly higher suicide frequencies during northern hemisphere summer months compared with other seasons, while one study did not find this association.47 Depression and bipolar disorder 13 studies investigated the association between high ambient temperature and depression. More than half used heat or heat wave as the exposure (53.85 %), four used ambient temperature (30.77 %) and two apparent temperature (15.38 %). Most employed prospective cohort designs (53.85 %), with the remaining using retrospective time-series designs (15.38 %), case-crossover designs (15.38%), or cross-sectional designs (7.69 %). All but one study38 was conducted in China. Several studies used overlapping data sources: four used data from the China Health and Retirement Longitudinal Study52 64 67 83 and three the Chinese Family Panel Studies.54 81 82 The results from these studies were mixed, with 53.85% of studies reporting statistically significant positive associations and 38.46% finding no statistically significant associations. For example, Hua et al54 reported that each day of exposure to ≥30°C (compared with 20°C–25°C) increased depressive symptoms by 0.11 units among males (p<0.05) and an additional 0.11 units among females (p<0.01) in their nationally representative Chinese sample. Conversely, Jiang et al84 reported that every 5°C reduction across lag days 0–14 was associated with a 1.25-fold increased odds for mild depression (95% CI 1.054 to 1.483, p<0.01) as measured by the PHQ-9. Bipolar disorder was only investigated by one study.71 Wu et al found no association between admission to the hospital due to bipolar disorder and ambient temperature among women (OR=1.00 (n.s.)). Schizophrenia 10 studies investigated the association between high ambient temperature and hospitalisation or outpatient treatment for schizophrenia. Most (70.00%) used heat or heat wave as the exposure and used retrospective time-series designs (80.00%). All studies were conducted in China, eight in Hefei or the Anhui province (80.00%5666 71 86 88,90 92), one in Liuzhou (10.00%91) and one in Jining (10.00%87). Studies conducted in the same city used overlapping data sources. For example, six of eight studies conducted in Hefei/Anhui used data from the Anhui Mental Health Centre. Most studies (70.00 %) found no statistically significant associations between high ambient temperature and hospitalisation for or diagnosis of schizophrenia. Only three studies reported statistically significant associations: Pan et al88 found a 1.25-fold increased risk for hospitalisation for schizophrenia (RR=1.25, 95% CI only reported graphically) when exposed to high diurnal temperature range among n=26 882 women from Hefei. Zhou et al91 found a 1.075-fold increased risk for schizophrenia admissions (RR=1.075, 95% CI 1.006 to 1.148) when exposed to high mean temperature (max. 24.75°C) compared with 21.65°C (the 75th percentile) among n=5010 women from Liuzhou. Lastly, Zhong et al66 found a 15% increased odds of being diagnosed with schizophrenia (OR=1.15, 95% CI1.03 to 1.27) among women exposed to compound heat extremes (>percentile 90, 1 day, temperature not given) compared with exposure to median temperature (15.7°C). All other studies reported statistically non-significant effect measure sizes between RR=1.01 and RR=1.25 (online supplemental Table 2). Anxiety Three studies investigated the association between high ambient temperature and symptoms38 or diagnosis66 71 of an anxiety disorder. Most studies (66.66 %) used heat or heat wave as the exposure and were conducted in China (66.66 %). While two of three studies did not find a statistically significant association, Zhong et al66 found a 20% increased odds of being diagnosed with an anxiety disorder (not further specified; OR=1.20, 95% CI 1.08 to 1.35) among women exposed to compound heat extremes (> percentile 90, 1 day, temperature not given) compared with exposure to median temperature (15.7°C) in their case-crossover studies of N=450 633 women from the Anhui Province in China. Mental/emotional well-being Two studies investigated associations between high ambient temperature and emotional/mental well-being. Both were qualitative studies conducted in Kenya using the same sample but with different endpoints: Scorgie et al33 examined associations between heat exposure on maternal health, while Lusambili et al37 focused on infant health and maternal care. Both studies collected self-reported information on heat and well-being through in-depth interviews. Both studies reported that women’s well-being was substantially impacted by heat. Lusambili et al37 reported that ‘heat had significant effects on the postpartum women, including exhaustion, discomfort and the inability for mothers to take care of themselves’. Similarly, Scorgie et al33 reported that women became ‘irritable and moody’ and lacked ‘peace in (their) bod(ies)’ when exposed to heat. Relationship between heat exposure and neurologic and neurocognitive outcomes Sleep disturbances Three studies investigated the association between high ambient temperature and sleep disturbances. All but one used heat or heat wave as the exposure, and all used a qualitative study design. Among these studies, two were conducted in Kenya, likely using the same sample33 35 and one in Burkina Faso.32 All three studies reported that women’s sleep was substantially disturbed by heat exposure. For example, Balakrishnan et al35 found that “participants reported sleep disturbances and fatigue across multiple extreme weather events, including … higher-than-usual temperatures.” Similarly, Scorgie et al33 found that “high temperatures are associated with a range of symptoms and discomfort, with reported impacts by pregnant women including a lack of sleep.” Two studies investigated sleep duration specifically. Zhou et al58 used data from N=3631 female participants in the Chinese Longitudinal Healthy Longevity Survey and found that exposure to heat (threshold was 97.5th percentile, 3 days, temperature not given) was associated with approximately 1.05-fold increased odds for reduced sleep duration (results only presented graphically). In contrast, Koch et al97 conducted an observational panel study among N=71 women in Burkina Faso using wearable devices and found no association between overnight heat exposure and shorter sleep duration, though the authors did not present detailed results. Two additional studies investigated sleep apnoea and sleep restlessness, respectively. Wang et al98 examined the association between ambient temperature and sleep restlessness among Chinese Family Panel Study participants. The authors found that the number of days exposed to ≥28°C during the 7 days prior to the interview (compared with exposure to 16°C–20°C) was not significantly associated with self-reported sleep restlessness (β=0.0067, n.s.). A. Li et al99 estimated the association between ambient temperature and obstructive sleep apnoea (OSA) measured by wearable devices among participants from 313 Chinese cities. The authors found that for every 10°C increase in ambient temperature, the odds of OSA exacerbation increased by 8% at lag day 0. Dementia and cognitive function Three studies investigated the association between high ambient temperature and risk for death due to dementia, all using heat or heat wave as the exposure. All studies were conducted in China, employed a case-crossover study design and found statistically significant positive associations between heat and dementia-related death. Effect measure sizes ranged from a cumulative OR=1.77 (95% CI 1.49 to 2.09) for hot day exposure (threshold 97.5th percentile (34.2°C), 13 days, compared with temperature at the minimal risk (the first percentile, temperature not given))57 to a cumulative OR=2.55 (95% CI 2.20 to 2.96) for heat exposure with temperature ≥the 97.5th percentile (no temperature given, max. temperature 33.1°C) for 4 days.93 Three studies investigated the association between high ambient temperature and cognitive function, using heat or heat waves as the exposure. All studies in this group were conducted in China and used prospective cohort study designs. Among them, two studies used data from the Chinese Longitudinal Healthy Longevity Survey,94 95 while one used the Chinese Family Panel Study.96 Two studies found a significant positive association between heat exposure and worse cognitive function. Zhou et al95 analysed data from N=4712 elderly women across 23 Chinese provinces and found that heatwave exposure (≥97.5th percentile (36.7°C), for 4 days) was associated with a 2.5% increased hazard for cognitive impairment (HR=1.025, 95% CI 1.002 to 1.050). Hou and Xu94 analysed N=11 096 women from mainland China and reported a positive association between exposure to high temperatures (25°C–31°C) and the Mini Mental State Examination score of global cognitive function, though results were presented graphically only. In contrast, Chen and Yu,96 using data from the Chinese Family Panel Study, found that women exposed to temperatures 35°C or higher performed 0.376-SD better than men on cognitive tests compared with their performance at temperatures between 10 and 15 °C (β=0.376, p=0.03). Headache and Migraine One study investigated the association between ambient temperature and episodic tension-type headache and episodic migraine.100 This case-series study, conducted in Türkiye, found no evidence of associations between headache and migraine with ambient temperatures for male and female patients. Descriptive overview of methodological limitations In accordance with standard scoping review methodology,41 a formal risk of bias or quality appraisal was not conducted, due to the substantial heterogeneity in study designs (table 1). Instead, we provide here a brief descriptive overview of key methodological and reporting limitations found among the included studies. First, several studies relied on data from the same data source. Seven studies used data from the Anhui Mental Health Center (10.00%), five used data from the Chinese Family Panel Studies (7.14%), five from the China Health and Retirement Longitudinal Study (7.14%), and three from the Chinese Longitudinal Healthy Longevity Survey (4.29%). Additionally, two studies used data from the Hanoi Mental Hospital, the China Cause of Death Reporting System, the Curitiba Health Secretary and the Kenya Qualitative Study, respectively (1.89% each), suggesting that up to 40% of all studies in this research field may not be independent. Second, several study characteristics were underreported, including ethnic distribution, socioeconomic status, housing situation and age of the participants. For example, only 40 studies (57.14 %) reported any measure describing their sample’s age. Third, as shown in table 1, a large predominance exists of studies originating from China (63.01%), limiting the generalisability to other LMIC regions and potentially introducing regional or cultural bias. Given these limitations, the results should be interpreted as a descriptive map of available evidence, as intended in a scoping review. Discussion We reviewed available published evidence on associations between high ambient temperature and psychiatric, neurologic and neurocognitive outcomes in women from LMICs. Included studies provided evidence suggestive of positive associations between heat exposure and adverse outcomes across these domains. When statistically significant, most association estimates indicated approximately 10–20% increased risk or odds for heat-exposed, compared with unexposed individuals. Geographically, studies were heavily concentrated in China, with notable under-representation from Latin America, Africa and Southeast Asia. Additionally, most studies analysed female subgroups within mixed-sex samples rather than exclusively female samples. For psychiatric conditions, results indicate adverse associations between heat exposure and non-specific psychiatric symptoms and disorders. In contrast, associations with heat exposure and specific symptoms or disorders such as depression, suicide, anxiety or schizophrenia were less consistent, though effect measure sizes often trended towards adverse outcomes even when not statistically significant. This finding aligns with the broader literature. Lawrance et al105 found consistent evidence linking high and extreme temperatures to mental health outcomes in various samples, but noted that associations are complex and context-dependent. Similarly, Cianconi et al’s results suggest that links between global-warming-related environmental changes and specific disorders may arise from combinations of ‘various atmospheric changes and perturbations’ rather than single exposures.27 Although results are descriptive and must be interpreted in the backdrop of methodological limitations of the existing studies, several factors may explain why non-specific rather than disorder-specific associations seem to be more consistent. First, the pathways linking heat exposure to mental health are highly diverse, encompassing biological, social, behavioural and economic mechanisms.105 Second, psychiatric disorders commonly exhibit high comorbidity. Studies with a broader scope (eg, mental and behavioural disorders) naturally account for comorbidities, while symptom-specific investigations may miss them due to narrow outcome definitions. Finally, detecting disorder-specific associations may require considering exposure timing—differentiating between acute, subacute and long-term consequences—which most reviewed studies did not address. This temporal differentiation is well-established in climate change research and could improve future studies on heat and women’s psychiatric health. For neurologic and neurocognitive symptoms and disorders, most studies reported adverse associations with heat exposure, particularly for dementia. However, this finding requires caution given the limited research on many neurologic and neurocognitive conditions. For example, only one study examined heat exposure and migraine,100 and we found no studies investigating psychosomatic, somatoform or neuropathic pain disorders. The available evidence on sleep quality and quantity was primarily drawn from qualitative studies in the same region,32 33 35 37 limiting the ability to draw conclusive inferences. Nevertheless, research on sex-mixed populations yielded similar results.106 In their systematic review of 27 studies from all global continents, Chevance et al106 found that higher ambient temperature is associated with poorer sleep quantity and quality. Furthermore, they noted that climate change-related sleep loss is likely to be disproportionately greater among vulnerable populations, including women and individuals living in LMICs.106 107 Further research is essential to better understand the direct and indirect consequences of climate-related temperature increases for women in the global south, given that only seven included studies investigated exclusively female samples, over 60% were conducted in China, and up to 40% may use overlapping data sources. Our results suggest three main implications for future studies. First, future studies should prioritise understudied regions and countries, particularly those in Africa and Latin America. Current research is limited to Colombia, Mexico and Brazil, as well as Kenya, Burkina Faso and South Africa. Given substantial regional variation in climate conditions, research should investigate regional effect measure modification to identify and prevent global disparities. When using large databases, researchers should transparently report previous publications using the same data to avoid wasted resources and biased findings. Second, future studies should ensure close alignment between exposure definitions and research questions, as different exposure definitions carry distinct methodological and interpretational implications. We found that studies using categorical definitions of heat or heat waves—typically defined by combinations of duration and severity—more often reported statistically significant associations than those using continuous measures of ambient or apparent temperature. While categorising exposure can reduce statistical power due to the loss of information,108 109 it can create substantial mean differences between exposed and unexposed groups, potentially making observed associations seem more pronounced. In contrast, categorical approaches may better capture associations at exposure extremes that continuous analyses may miss. Thus, future research should consider both continuous and categorical exposure definitions to capture associations in a nuanced manner across the full exposure range. For the latter, transparent reporting of baseline temperatures and temperature distributions is essential, as the categorisation of extreme temperatures depends on these reference values. For example, 4 days at the 95th percentile differ substantially between regions with baseline temperatures of 5°C vs 25°C. Third, our reporting of study-specific effect measures, sample size and CIs provides a foundation for statistical power calculations in future studies. Achieving methodological rigour, including designing studies with sufficient statistical power, is essential for producing robust scientific evidence in this field. Our review has several limitations. First, study selection and data extraction were conducted by a single author due to resource limitations, which may increase the risk of bias. However, the inclusion criteria were thoroughly discussed, clearly defined and registered in advance. Furthermore, key study selection decisions were verified by another team member (MAW). Second, we covered only selected neurologic and neurocognitive outcomes, focusing on those most clinically relevant, particularly in relation to psychiatric comorbidities. However, we included a broad range of the clinically most important sleep, pain and cognitive conditions in our search. Third, because we included all studies that met inclusion criteria, a considerable number of included studies may not be independent due to the use of overlapping data sources. However, scoping such characteristics aligns with our review’s aim to map available evidence in this research field. Fourth, we did not conduct formal quality assessments of the reviewed studies, as this falls outside the scope of scoping reviews.41 However, we highlighted key methodological limitations of the included studies descriptively. Fifth, applying the World Bank classification110 for identifying LMICs, potential intracountry income disparities could not be accounted for. However, this classification is broadly accepted, and usage ensures comparability across studies. Conclusions Climate change poses a significant threat to human health, with disproportionate impacts on vulnerable and highly exposed populations. Women from LMICs represent a critical research gap—existing evidence is scarce and largely buried within subgroup analyses of mixed-sex studies. This scoping review reveals concerning evidence of adverse associations between high ambient temperatures and psychiatric, neurologic and neurocognitive health outcomes in women from LMICs. These findings demand urgent action: comprehensive research is essential to uncover regional differences and assess temporal trajectories of these associations. Understanding these severe health consequences is not only a scientific imperative but a moral one—the populations most affected by climate-related health impacts are those who have contributed least to the crisis itself. Supplementary material Acknowledgements The authors sincerely thank Stefania Papatheodorou for her valuable support and insightful feedback throughout this work. Footnotes Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors. Provenance and peer review: Not commissioned; externally peer reviewed. Handling editor: Barnabas Alayande Patient consent for publication: Not applicable. Ethics approval: Not applicable. Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research. Data availability statement Data sharing not applicable as no datasets generated and/or analysed for this study. References Associated Data Supplementary Materials Data Availability Statement Data sharing not applicable as no datasets generated and/or analysed for this study.