Mental health impacts of climate-related hazards among farmers: evidence from a community-based study

Abstract

Background

Climate-related hazards are intensifying worldwide, posing escalating risks to not only agricultural productivity but also mental health. Farmers in low- and middle-income countries are particularly vulnerable, yet the psychological impacts of climate disasters on these populations remain underexplored. This study aimed to examine the associations between climate-related hazard exposure and mental health outcomes among farmers in southern Thailand, a region heavily affected by climate disasters, including floods, droughts, and storms.

Methods

A community-based cross-sectional survey was conducted among 346 farmers in four districts of Chumphon and Surat Thani Provinces, classified as high- or low-climate hazard exposure areas, via five-year disaster records. Data on sociodemographic and behavioral characteristics, climate experiences, and coping strategies were collected via structured questionnaires. Mental health outcomes were assessed via the validated Thai version of the Depression, Anxiety, and Stress Scales (DASS-21). Multivariable logistic regression was applied to estimate adjusted odds ratios (aORs) to assess the associations between climate hazard exposure and adverse mental health outcomes.

Results

Depression was significantly more prevalent among farmers in high-exposure areas (49.7%) than among those in low-exposure areas (36.0%; p = 0.010). Increased flooding frequency in the previous years was strongly associated with depression (aOR = 6.10; 95% CI: 2.26–16.47), anxiety (aOR = 3.33; 95% CI: 1.18–9.38), and stress (aOR = 9.81; 95% CI: 3.74–25.74). Conversely, farmers in low-exposure areas reported significantly higher engagement in institutional supports (p < 0.001), including community resource management (20.5% vs. 2.2%), financial compensation (19.3% vs. 2.2%), and adoption of drought-resistant seeds (25.5% vs. 3.2%) whereas farmers in high-exposure areas largely relied on self-directed coping.

Conclusions

Climate-related hazards are significantly associated with increased risks of depression, anxiety, and stress among farmers. These findings underscore the importance of integrating mental health interventions into national climate adaptation frameworks, with a focus on early psychological screening, financial risk protection, and targeted psychosocial support for agricultural communities repeatedly affected by floods, droughts, and storms.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12995-026-00504-w.

Introduction

Climate change is one of the most urgent global health challenges of the 21st century, with wide-ranging effects on vulnerable populations [1–4]. Climate change has caused more frequent and intense extreme weather events, particularly droughts and floods, which are the primary hazards of climate change and affect not only global populations but also the global economy [5]. The understanding of flood and drought interconnections and their multi-hazard risk related to hazards, exposure and vulnerability, especially in the context of urbanized areas, has been intensively reviewed and reported by Pizzorni et al. [5]. However, the effects of intensifying floods, droughts, and storms on agricultural productivity destabilize rural livelihoods, particularly in low- and middle-income countries where adaptive capacity remains limited [6–7]. Farmers, who depend directly on natural resources for subsistence and income, face substantial risks when climate hazards disrupt crop cycles, damage infrastructure, and generate economic insecurity [8]. In addition to physical and financial hardship, these disruptions have profound and enduring effects on mental health, underscoring the need to consider climate change not only as an environmental crisis but also as a major public health threat with significant effects on mental health. Recent analyses have revealed that accelerating climatic shifts are intensifying pressures on environmentally dependent workers, markedly increasing their vulnerability to stress, anxiety, and other mental-health outcomes [9].

The health effects of climate change have received increasing attention worldwide; however, the mental health dimension remains underexplored [10–11]. Existing evidence from high-income countries suggests that exposure to floods, droughts, and other natural disasters is associated with elevated risks of psychological distress, post-traumatic stress disorder, depression, and long-term declines in mental well-being [12–14]. For instance, studies conducted in Australia and Europe have documented substantial increases in anxiety, depressive symptoms, and posttraumatic stress among farming populations and rural business owners exposed to prolonged droughts, cyclones and repeated flooding events [14–17]. However, these findings cannot be assumed to be generalizable directly to low- and middle-income countries, where vulnerabilities are compounded by agricultural dependence, fragile social safety nets, and limited access to mental health care [13, 18].

Thailand, one of the most climate-exposed countries in Southeast Asia according to the Global Climate Risk Index [19], experiences recurrent flooding, prolonged droughts, and seasonal storms. Previous studies in Thailand have largely examined the impacts of a single hazard, most commonly flooding, on mental health outcomes [20–22]. However, many farming households are exposed to multiple and interacting hazards, which together generate amplified risks that may exacerbate psychological distress. The scarcity of research on the cumulative mental health burden of multi-hazard exposures represents a critical gap in both the climate and health literature.

Given the above context, this study was conducted to (i) examine the impacts of multiple climate hazards including floods, droughts, and storms on the mental health of farmers, (ii) determine the prevalence of depression, anxiety, and stress among farmers, and (iii) determine the influences of multiple climate hazards exposure at the different risk levels. The study focused on the upper southern region of Thailand, which is highly vulnerable to climate variability and is central to the country’s agricultural economy through the production of major cash crops such as oil palm and rubber. By shedding light on this neglected intersection between climate change and mental health, this study provides insights into how climate change affects the psychological well-being of farmers and offers an evidence base to inform the integration of mental health considerations into climate-change adaptation strategies and strengthen agricultural resilience practices.

Methods

Background information, study design and setting

According to Thailand’s National disaster risk management plan (2021–2027), the country has been highly exposed and vulnerable to various types of hazards and disasters, including natural disasters, transportation disasters and contagious disease [23]. Among these hazards, floods, drought, and wind storms have been reported to negatively affect thousands of local people and cause millions Thai Baht of total economic damage and losses every year [23]. Over the past few decades, nation-wide floods, droughts, and storms have frequently and repeatedly occurred from June–September, January– May, and March–May, respectively [23]. However, floods in the southern part of the country normally occur from October–December [23].

A cross-sectional survey was conducted via a multi-stage cluster sampling approach in four districts of southern Thailand between June and August 2025. The classification of the study areas into high- and low-climate-hazard exposure areas was based on historical records of flooding, drought, and storm events over the past five years in conjunction with official disaster risk evaluations. Relevant data for district selection were obtained from the Provincial Department of Disaster Prevention and Mitigation and the Office of Social Development and Human Security, which oversees disaster statistics across 14 southern provinces. The Chumphon and Surat Thani Provinces were selected because they have the highest cumulative disaster statistics in the southern region. According to the Department of Disaster Prevention and Mitigation, Chumphon Province experienced 11 major floods, 2 droughts, and 252 tropical storms over the past five years, whereas Surat Thani Province recorded 10 floods, 1 drought, and 367 tropical storms during the same period. Recent reports on total number of households at risk of floods, droughts, and wind storms in Chumphon and Surat Thani Provinces are summarized in Table 1. Two districts from each province were selected via a combination of purposive and cluster random sampling to represent the study areas. On the basis of the systematic evaluation of five-year disaster records, Sawi district in Chumphon and Khiri Rat Nikhom district in Surat Thani were identified as high-climate-hazard exposure areas, whereas Thung Tako district in Chumphon and Ban Na Doem district in Surat Thani were classified as the low-climate-hazard exposure areas. Within these districts, subdistricts were randomly selected to ensure representativeness. Figure 1 illustrates the classification of study districts by climate-hazard risk level and geographic location within Thailand.

Table 1.

Number of households at risk from climate-related hazards in Chumphon and Surat Thani Provinces, 2021–2025 [24]

Risk level	Chumphon Province			Surat Thani Province
	Floods	Drought	Storms	Floods	Drought	Storms
Very high	104	15	29	156	20	29
High	190	51	124	429	45	204
Moderate	253	99	225	484	353	341
Low	80	66	70	138	185	124
No risk	171	567	350	111	715	620

Fig. 1.

Maps showing the locations of (a) Chumphon and Surat Thani provinces, and (b) high and low climate-hazard exposure areas in both provinces

Sample size calculation and participants in the study

The required sample size was estimated to be 240 on the basis of Woodward’s formula [25–26]. To ensure adequate representation across exposure areas and to account for an anticipated 50% nonresponse rate, this study applied a purposive oversampling strategy, resulting in a final recruitment of 600 farmers. Eligible participants included farmers aged 18 years and older who had been engaged in agricultural work in the selected study areas for at least one year. Four main agricultural types included crop cultivation, livestock production, fisheries, and mixed farming. Farmers who declined participation or withdrew during the survey were excluded. The participants who participated in the study were requested to provide information on sociodemographic background, work-related characteristics, perceptions and adaptive practices related to climate-related hazards, and mental health status following the questions on the questionnaire. The participants were informed about the study by local village health volunteers during monthly village meetings. Individuals who expressed interest were provided with detailed study information, gave written informed consent, and subsequently completed the study questionnaire. This study was conducted following the Declaration of Helsinki guidelines, and the protocol was approved by the Institutional Review Board of Faculty of Medicine, Chulalongkorn University (IRB No. 0157/68, COA No. 0595/2025). All participants provided written informed consent prior to data collection. The total questionnaire distributions and response rates are summarized in Supplementary Table S1.

Questionnaire design

The questionnaire was structured into three key domains, including (i) socioeconomic and work-related characteristics, (ii) perceptions of climate experience and adaptive practices and (iii) mental health status, as described below.

Socioeconomic and work-related characteristics

This demographic information included gender, age, and body mass index (BMI, calculated according to the Asian population criteria), along with education level, marital status, debt and economic status, smoking and alcohol consumption habits, underlying diseases, history of mental disorders, and frequency of physical exercise. The history of mental disorders in this study referred to participants’ self-reported prior mental health diagnoses. Notably, information on the family history of mental disorders was not collected. In addition, agricultural work-related factors included farm size, workplace type, and years of agricultural experience. The land size was reported in rai, a local Thai unit of area measurement, and converted to hectares (ha) for analysis (1 rai = 1,600 m² = 0.16 ha = 0.395 acres).

Perception of climate experience and adaptive practices

To explore perceptions and adaptive behaviors related to climate change, participants were asked about their experiences with the impacts of climate-related hazards on agriculture, personal health, and family well-being. The questionnaire also addressed challenges such as pest infestation and crop yield reduction. Adaptation strategies included increased use of organic or chemical fertilizers, improvement of irrigation or water systems, changes in crop type, adoption of faster-growing or drought-resistant varieties, engagement in nonagricultural work, access to loans or credit, participation in cooperatives, and involvement in community or governmental projects for natural resource management. Each adaptive practice was coded as a binary variable (No = 0, Yes = 1).

Mental health status

The participants’ mental health status was assessed via the Thai version of the Depression Anxiety Stress Scale-21 (DASS-21). This tool was developed by Lovibond and Lovibond [27] and has been translated and validated for Thai populations [28]. The DASS-21 comprises 21 items evaluating three subscales: depression, anxiety, and stress. Each contains seven items rated on a 4-point Likert scale (0 = “Did not apply to me at all” to 3 = “Applied to me very much or most of the time”). The Thai version demonstrated very good internal consistency, with Cronbach’s alpha values of 0.75 for the total scale, 0.82 for depression, 0.78 for anxiety, and 0.69 for stress [28]. Severity levels were categorized according to standardized DASS-21 criteria, as summarized in Supplementary Table S2.

Statistical analyses

Categorical variables are summarized as frequencies and percentages, whereas continuous variables are presented as the means with standard deviations or as medians with interquartile ranges, as appropriate. Differences between farmers residing in areas with high and low levels of climate hazard exposures were compared via the chi-square test or Fisher’s exact test for categorical variables and the independent t test or Wilcoxon rank-sum test for continuous variables.

Exploratory analyses were conducted to identify factors associated with adverse mental health outcomes (depression, anxiety, and stress) in relation to climate-related hazard exposure. The main determinants, which were the different levels of climate-related hazard exposure areas (high-, and low-risk areas) and the primary outcomes measured by the DASS-21 (depression, anxiety, and stress) were selected for analysis via multivariable logistic regression, with adjustments for potential confounders, including age, gender, exposure areas, debt, incidence of climate-related events, alcohol consumption, chronic disease status, marital status, and household income sufficiency. Coping and institutional support variables were additionally included in multivariable logistic regression models to examine their potential protective associations with depression, anxiety, and stress. The magnitude of association was reported as an adjusted odds ratio (aOR) with 95% confidence intervals (95% CIs), and statistical significance was defined as p < 0.05. All the analyses were performed via STATA program version 19 (Stata Corp, LP, USA).

Results

Socioeconomic characteristics

A total of 600 questionnaires were distributed, and 346 farmers returned completed forms, yielding a response rate of 57.7% (Supplementary Table S1). Among the 346 participating farmers, 185 (53.5%) resided in high-climate-hazard exposure areas and 161 (46.5%) resided in low-climate-hazard exposure areas. Table 2 summarizes the demographic characteristics of the participants who lived in both high- and low-climate hazard exposure areas. Overall, the two groups were comparable in terms of most baseline characteristics. As clearly shown in Table 2, the sex distribution, mean age, mean BMI, prevalence of chronic diseases and smoking behavior did not differ significantly among the participants living in the high- and low-climate hazard exposure areas. However, significant differences in sociodemographic and behavioral characteristics were observed. A higher proportion of divorced or widowed participants were observed in the high-climate-hazard exposure area than in the low-climate-hazard exposure area (p < 0.001) (Table 2). Educational attainment in the low-climate-hazard exposure area, of which approximately 30.6% had a bachelor’s degree or higher, was significantly greater than that of 17.9% of the total participants living in the high-climate-hazard exposure area (p < 0.001) (Table 2). Regular alcohol use was significantly and markedly more common among farmers living in areas with high climate hazard exposure than among farmers living in areas with low climate hazard exposure area, with 30.3% reported drinking regularly in areas with high climate hazard exposure area (Table 2). Perceived financial sufficiency showed a contrasting pattern in which a greater proportion of high-exposure farmers reported having sufficient income with savings, whereas farmers in low-exposure areas more frequently reported having enough income to meet only daily needs or insufficient income (p = 0.003).

Table 2.

Demographical characteristics of the participants (n = 346)†

		High climate-hazard exposure areas (n = 185)	Low climate-hazard exposure areas (n = 161)	p value ‡
Gender, n(%)				0.325
Male		96 (51.9%)	75 (46.6%)
Female		89 (48.1%)	86 (53.4%)
Age, mean ± SD		48.6 ± 14.19	49.39 ± 13.95	0.788
BMI, mean ± SD		25.19 ± 5.74	26.26 ± 6.15	0.204
Any chronic medical condition, n (%)				0.802
No		109 (58.9%)	97 (60.2%)
Yes		76 (41.4%)	64 (39.8%)
Specific chronic conditions, n (%)
Diabetes mellitus		35 (18.9%)	22 (13.7%)	0.189
Hypertension		37 (20.0%)	34 (21.1%)	0.797
Dyslipidemia		35 (18.9%)	29 (18.0%)	0.829
Asthma		7 (3.8%)	0 (0.0%)	0.013*
Ischemic heart disease		1 (0.5%)	0 (0.0%)	0.350
Cancer (any type)		1 (0.5%)	2 (1.2%)	0.483
History of mental health disorder, n (%)				0.588
No		180 (97.3%)	155 (96.3%)
Yes		5 (2.7%)	6 (3.7%)
Marital status, n (%)				< 0.001**
Single		44 (23.8%)	59 (36.6%)
Married		103 (55.7%)	92 (57.1%)
Divorced/Widowed		38 (20.5%)	10 (6.2%)
Education level, n (%)				< 0.001**
None		1 (0.5%)	11 (6.8%)
Primary school		45 (24.3%)	30 (18.6%)
Lower secondary school		23 (12.4%)	16 (9.9%)
Upper secondary school		55 (29.7%)	23 (14.3%)
Vocational certificate		15 (8.1%)	19 (11.8%)
Diploma in vocational education		13 (7%)	12 (7.5%)
Bachelor’s degree		19 (10.3%)	30 (18.6%)
Higher than bachelor’s degree		14 (7.6%)	20 (12.4%)
Smoking habits, n (%)				0.477
Non-smoker		117 (63.2%)	108 (67.1%)
Ex-smoker		48 (25.9%)	33 (20.5%)
Current smoker		20 (10.8%)	20 (12.4%)
Alcohol intake, n (%)				< 0.001*
Non-drinker		76 (41.1%)	91 (56.5%)
Occasionally		53 (28.6%)	62 (38.5%)
Regularly		56 (30.3%)	8 (5%)
Exercise (≥ 3 times/week), n(%)				< 0.001*
No		131 (70.8%)	72 (44.7%)
Yes		54 (29.2%)	89 (55.3%)
Household income insufficiency, n (%)				0.003*
Income is insufficient		42 (22.7%)	48 (30%)
Income is sufficient but without savings		71 (38.4%)	77 (48.1%)
Income is sufficient and with savings		72 (38.9%)	35 (21.9%)
Family relationship score, mean ± SD		8.19 ± 1.79	7.94 ± 2.03	0.359
Workplace				< 0.001*
Not a local resident		75 (40.5%)	36 (22.4%)
A local resident		110 (59.5%)	125 (77.6%)
Debt status, n (%)				0.688
Without debt		74 (40%)	61 (37.9%)
With debt		111 (60%)	100 (62.1%)

Remark: † Climate-hazard exposure was defined by combined occurrence of flood, drought, and wind storm events, which are well-documented manifestations of climate change in the study areas

‡ p values for continuous variables were calculated via t tests, and p values for categorical variables were calculated via Fisher’s exact test

* p < 0.05

** p < 0.001

Work-related characteristics

The work-related characteristics of all 346 participants are summarized in Table 3. The distribution of farming types differed significantly between the high- and low-climate hazard exposure areas (p = 0.002). Crop cultivation was the dominant agricultural type among farmers in areas with high climate hazard exposure, whereas mixed farming was more prevalent in areas with low climate hazard exposure. Livestock production and fisheries were relatively uncommonly not significantly different between high- and low-climate hazard exposure areas.

Table 3.

Work-related characteristics of participants (n = 346)†

	High climate-hazard exposure areas (n = 185)	Low climate-hazard exposure areas (n = 161)	p value‡
Agriculture type, n(%)			0.002*
Crop cultivation	134 (72.4%)	87 (54%)
Livestock production	4 (2.2%)	11 (6.8%)
Fisheries	2 (1.1%)	6 (3.7%)
Mixed farming	45 (24.3%)	57 (35.4%)
Farm size, n (%)			< 0.001**
Small scale (not exceeding 10 rai) §	43 (23.2%)	92 (57.1%)
Medium scale (11–50 rai)	115 (62.2%)	58 (36%)
Large scale (more than 50 rai)	27 (14.6%)	11 (6.8%)
Farming experience, n (%)			0.234
5–10 years	45 (24.3%)	52 (32.3%)
10–15 years	63 (34.1%)	46 (28.6%)
> 15 years	77 (41.6%)	63 (39.1%)

Remark: † Climate-hazard exposure was defined by combined occurrence of flood, drought, and wind storm events, which are well-documented manifestations of climate change in the study areas

‡ p values for continuous variables were calculated via t tests, and p values for categorical variables were calculated via Fisher’s exact test

* p < 0.05

** p < 0.001

§ One rai equals 1,600 m² (0.16 hectares (ha) or 0.395 acres)

Farm size varied markedly according to the level of hazard risk (p < 0.001). Small-scale farms predominated in areas with low climate hazard exposure, whereas approximately 1.7 to 2.4 times more medium-sized farms and large-scale farms were substantially more common in areas with high climate hazard exposure.

Farming experience did not differ significantly between the farmers living in high- and low-climate-hazard exposure areas (Table 3). In both areas, approximately two-fifths of the farmers reported more than 15 years of agricultural experience.

Awareness and experience of climate change and related disasters

The data on the awareness and experience of climate change and climate-related disasters are summarized in Table 4. Almost all the farmers (≥ 99%) had heard of climate change. Perception of climate change was high in both high- and low climate-hazard exposure areas, with only a minimal difference between groups (100% vs. 96.9%, respectively). The primary sources of climate information differed by exposure area (p = 0.001). The leading sources of climate change in the farmers living in the high- and low- level climate hazard exposure areas were the internet and television/newspaper, respectively. The experiences of recent disasters also varied by type of hazard; droughts were more common in areas with low climate hazard exposure, whereas storms and heavy rainfall were more prevalent in areas with high climate hazard exposure, as summarized in Table 4.

Table 4.

Data on awareness and experience of climate change and climate-related disasters (n = 346)†

	High climate-hazard exposure areas (n = 185)	Low climate-hazard exposure areas (n = 161)	p value‡
Heard about climate change, n(%)	185 (100%)	160 (99.4%)	0.283
Perceived weather changes, n (%)	185 (100%)	155 (96.9%)	0.015*
Information source, n (%)
TV/Newspaper	96 (51.9%)	111 (68.9%)	0.001*
Radio	18 (9.7%)	18 (11.2%)	0.659
Internet	134 (73.6%)	98 (60.9%)	0.012*
Local authorities	6 (3.2%)	12 (7.5%)	0.079
Friends	49 (26.5%)	46 (28.6%)	0.665
Family	22(11.9%)	18 (11.2%)	0.836
Disaster experience (past year), n (%)
Flood	100 (61.3%)	76 (58%)	0.562
Drought	42 (25.8%)	67 (51.1%)	< 0.001**
Storm	139 (85.3%)	90 (68.7%)	0.001*
Fire	0 (0%)	2 (1.5%)	0.113

Remark: † Climate-hazard exposure was defined by combined occurrence of flood, drought, and wind storm events, which are well-documented manifestations of climate change in the study areas

‡ p values for continuous variables were calculated via t tests, and p values for categorical variables were calculated via Fisher’s exact test

* p < 0.05

** p < 0.001

Coping and adaptation behaviors to climate hazards

Coping strategies

Data on coping strategies and support systems for climate-related hazards are summarized in Table 5. Farmers in areas with low levels of climate hazard exposure were significantly more likely to participate in community-based natural resource management programs (20.5% vs. 2.2%; p < 0.001), receive financial compensation from authorities (19.3% vs. 2.2%; p < 0.001), and adopt improved or drought-resistant seeds (25.5% vs. 3.2%; p < 0.001). These findings suggest stronger institutional engagement and resource access in areas with low climate hazard exposure area, whereas adaptation in areas with high climate hazard exposure has remained largely individual and self-reliant.

Table 5.

Data on coping strategies and support systems to climate-related hazards (n = 346)†

	High climate-hazard exposure areas (n = 185)	Low climate-hazard exposure areas (n = 161)	p value‡
Coping strategies, n(%)
Started non-agricultural activities	57 (30.8%)	64 (39.8%)	0.082
Changed to wage employment	48 (25.9%)	45 (28%)	0.675
Household migration/relocation	3 (1.6%)	6 (3.7%)	0.22
Participated in community natural resource management	4 (2.2%)	33 (20.5%)	< 0.001*
Changed farming / livestock practices	18 (9.7%)	26 (16.1%)	0.074
Invested in water storage	44 (23.8%)	34 (21.1%)	0.554
Increased fertilizer use	121 (65.4%)	90 (55.9%)	0.071
Use seeds that have been improved over the past year (better planting, drought resistance, etc.)	6 (3.2%)	41 (25.5%)	< 0.001*
Support systems, n (%)
Agricultural training/ skill development	6 (3.2%)	12 (7.5%)	0.079
Support from agri-livestock programs or banks	2 (1.1%)	7 (4.3%)	0.057
Visited agencies for climate information	3 (1.6%)	7(4.3%)	0.131
Participated in community resource management	3 (1.6%)	18 (11.2%)	< 0.001*
Received financial compensation	4 (2.2%)	31 (19.3%)	< 0.001*
Stored seeds locally (seed banks)	3 (1.6%)	4 (2.5%)	0.57

Remark: † Climate-hazard exposure was defined by combined occurrence of flood, drought, and wind storm events, which are well-documented manifestations of climate change in the study areas

‡ p values for continuous variables were calculated via t tests, and p values for categorical variables were calculated via Fisher’s exact test

* p < 0.05.

** p < 0.001

Support systems

Institutional and technical support mechanisms differed significantly across areas with different risk levels (Table 5). Farmers in areas with low levels of climate hazard exposure were more likely to participate in agricultural training or skill development and to receive assistance through agricultural extension or bank support initiatives. Access to institutional aid was also substantially greater in areas with low levels of climate hazard exposure, as reflected by greater participation in community-based natural resource management programs and higher receipt of financial compensation following climate-related hazards. These findings indicate that low-exposure districts benefit from more accessible and better coordinated adaptation and support infrastructures, whereas high-exposure areas demonstrated more limited and fragmented institutional response capacity (Table 5).

Mental health outcomes

Table 6 summarizes the mental health status of farmers living in areas with high and low levels of climate hazard exposure. The prevalence of depressive symptoms was significantly greater among farmers in areas with high climate hazard exposure than among those in areas with low climate hazard exposure (p = 0.010). The prevalence of anxiety and stress in farmers in both areas, however, were not statistically significant (Table 6).

Table 6.

Levels of depression, anxiety, and stress among farmers living in high- and low-climate hazard exposure areas (n = 346)†

Mental status	High climate-hazard exposure areas (n = 185)	Low climate-hazard exposure areas (n = 161)	p value‡
Depression, n (%)			0.010*
Normal	93 (50.3%)	103 (64%)
Abnormal	92 (49.7%)	58 (36%)
Anxiety, n (%)			0.513
Normal	74 (40%)	70 (43.5%)
Abnormal	111 (60%)	91 (56.5%)
Stress, n (%)			0.599
Normal	156 (84.3%)	139 (86.3%)
Abnormal	29 (15.7%)	22 (13.7%)

Remark: † Climate-hazard exposure was defined by combined occurrence of flood, drought, and wind storm events, which are well-documented manifestations of climate change in the study areas

‡ p values for continuous variables were calculated via t tests, and p values for categorical variables were calculated via Fisher’s exact test

* p < 0.05

** p < 0.001

Climate-related hazards and demographic factors associated with mental health outcomes in farmers

Table 7 presents the results of the multivariable logistic regression analysis examining factors associated with depression, anxiety, and stress in all the farmers. After adjustment for covariates, residing in high-climate-hazard exposure areas was significantly associated with increased odds of depression (aOR = 1.87; 95% CI: 1.12–3.12; p = 0.016), whereas associations with anxiety and stress were not statistically significant. Compared with those who experienced no change in flood frequency, those who reported increased flooding frequency during the past years presented markedly elevated odds of all three mental health outcomes, depression (aOR = 6.10; 95% CI: 2.26–16.47; p < 0.001), anxiety (aOR = 3.33; 95% CI: 1.18–9.38; p = 0.023), and stress (aOR = 9.81; 95% CI: 3.74–25.74; p < 0.001). Interestingly, an increased frequency of storms was significantly associated with lower odds of only depression (aOR = 0.38; 95% CI: 0.19–0.77; p = 0.007).

Table 7.

Multivariable logistic regression analysis of factors associated with depression, anxiety, and stress among farmers exposed to climate-related hazards (n = 346)†

Variable	Depression		Anxiety		Stress
	aOR (95%CI)	p value‡	aOR (95%CI)	p value‡	aOR (95%CI)	p value‡
Climate-hazard exposure
Low climate-hazard exposure area (Reference)	1.00		1.00		1.00
High climate-hazard exposure area	1.87 (1.12–3.12)	0.016*	1.15 (0.70–1.87)	0.583	1.66 (0.80–3.45)	0.172
Age group
≤ 40 years (Reference)	1.00		1.00		1.00
> 40 years	0.71 (0.40–1.29)	0.262	0.84 (0.47–1.49)	0.544	0.90 (0.42–1.90)	0.773
Household debt
No (Reference)	1.00		1.00		1.00
Yes	1.03 (0.64–1.66)	0.914	1.13 (0.71–1.80)	0.603	2.53 (1.20–5.33)	0.014*
Frequency of flooding events (past year)
Same frequency (Reference)	1.00		1.00		1.00
Increased frequency	6.10 (2.26– 16.47)	< 0.001**	3.33(1.18–9.38)	0.023*	9.81 (3.74–25.74)	< 0.001**
Frequency of storms events (past year)
Same frequency (Reference)	1.00		1.00		1.00
Increased frequency	0.38 (0.19– 0.77)	0.007*	0.90 (0.46–1.72)	0.739	0.45 (0.18–1.09)	0.080
Frequency of drought events (past year)
Same frequency (Reference)	1.00		1.00		1.00
Increased frequency	1.08 (0.59– 1.96)	0.805	1.34 (0.74–2.39)	0.327	1.97 (0.85–4.50)	0.110
Gender
Male (Reference)	1.00		1.00		1.00
Female	1.21 (0.74–1.99)	0.456	1.40 (0.86–2.28)	0.173	0.91 (0.46–1.81)	0.789
Chronic disease
No (Reference)	1.00		1.00		1.00
Yes	0.97 (0.57–1.65)	0.916	0.91 (0.55–1.52)	0.717	0.58 (0.27–1.26)	0.168
History of Mental Disorders
No (Reference)	1.00		1.00		1.00
Yes	0.99 (0.20–5.01)	0.991	2.07 (0.37–11.55)	0.408	1.93 (0.36–10.35)	0.444
Household income sufficiency
Sufficient with savings (Reference)	1.00		1.00		1.00
Sufficient without savings	1.49 (0.86–2.61)	0.157	1.41 (0.83–2.42)	0.205	0.70 (0.33–1.48)	0.349
Insufficient	1.32 (0.70–2.48)	0.394	1.42 (0.77–2.61)	0.260	0.59 (0.24–1.45)	0.251
Marital status
Single (Reference)	1.00		1.00		1.00
Married	0.73 (0.42–1.27)	0.270	0.91 (0.53–1.56)	0.724	0.51 (0.24–1.09)	0.083
Divorced/Widowed	0.97 (0.45–2.08)	0.927	1.10 (0.51–2.37)	0.816	0.69 (0.23–2.08)	0.511
Alcohol intake
None (Reference)	1.00		1.00		1.00
Alcohol intake (Occasionally)	2.03 (1.18–3.50)	0.011*	1.74 (1.03–2.96)	0.040*	1.19 (0.57–2.52)	0.643
Alcohol intake (Regularly)	2.76 (1.43–5.34)	0.003*	2.36 (1.20–4.64)	0.013*	0.77 (0.30–2.02)	0.599
Coping strategies
Started non-agricultural activities
No (Reference)	1.00		1.00		1.00
Yes	1.58 (0.88–2.87)	0.128	0.91 (0.52–1.59)	0.736	0.59 (0.24–1.48)	0.264
Changed to wage employment
No (Reference)	1.00		1.00		1.00
Yes	1.67 (0.89–3.13)	0.112	1.32 (0.71–2.45)	0.379	0.49 (0.19–1.27)	0.140
Household migration/relocation
No (Reference)	1.00		1.00		1.00
Yes	0.77 (0.08–7.65)	0.820	0.52 (0.07–4.18)	0.541	6.80 (0.50-92.71)	0.151
Participated in community natural resource management
No (Reference)	1.00		1.00		1.00
Yes	0.85 (0.21–3.45)	0.824	0.53 (0.16–1.75)	0.296	0.74 (0.11–4.88)	0.755
Changed farming/livestock practices
No (Reference)	1.00		1.00		1.00
Yes	0.61 (0.25–1.50)	0.284	0.83 (0.37–1.87)	0.657	0.29 (0.06–1.34)	0.112
Invested in water storage
No (Reference)	1.00		1.00		1.00
Yes	1.22 (0.63–2.34)	0.558	1.07 (0.57–2.01)	0.843	0.88 (0.35–2.19)	0.779
Increased fertilizer use
No (Reference)	1.00		1.00		1.00
Yes	0.55 (0.31–0.97)	0.040*	0.96 (0.56–1.63)	0.868	1.09 (0.50–2.40)	0.827
Use seeds that have been improved over the past year (better planting, drought resistance, etc.)
No (Reference)	1.00		1.00		1.00
Yes	0.55 (0.23–1.30)	0.173	1.27 (0.57–2.83)	0.557	0.97 (0.29–3.24)	0.960
Support systems
Agricultural training/skill development
No (Reference)	1.00		1.00		1.00
Yes	0.75 (0.15–3.81)	0.724	0.37 (0.09–1.59)	0.180	1.87 (0.22–15.96)	0.567
Support from agri-livestock programs or banks
No (Reference)	1.00		1.00		1.00
Yes	2.50 (0.19–3.36)	0.490	0.71 (0.06–7.70)	0.774	2.38 (0.05-114.88)	0.662
Visited agencies for climate information
No (Reference)	1.00		1.00		1.00
Yes	0.29 (0.03–3.31)	0.320	1.17 (0.15–9.43)	0.881	0.58 (0.04–9.27)	0.703
Participated in community resource management
No (Reference)	1.00		1.00		1.00
Yes	0.44 (0.08–2.24)	0.319	0.46 (0.11–1.83)	0.267	1.10 (0.12–10.43)	0.935
Received financial compensation
No (Reference)	1.00		1.00		1.00
Yes	0.74 (0.25–2.17)	0.577	0.62 (0.24–1.63)	0.331	0.26 (0.02–2.92)	0.272
Stored seeds locally (seed banks)
No (Reference)	1.00		1.00		1.00
Yes	0.11 (0.00-2.72)	0.176	0.46 (0.05–3.60)	0.462	0.14 (0.00-14.11)	0.400

Remark: † Climate-hazard exposure was defined by combined occurrence of flood, drought, and wind storm events, which are well-documented manifestations of climate change in the study areas

‡ p values for continuous variables were calculated via t tests, and p values for categorical variables were calculated via Fisher’s exact test

* p < 0.05

** p < 0.001

Alcohol intake, whether occasional or regular, was strongly associated with depression (aOR = 2.03; 95% CI: 1.18–3.50; p = 0.011 and aOR = 2.76; 95% CI: 1.43–5.34; p = 0.003, respectively) and anxiety (aOR = 1.74; 95% CI: 1.03–2.96; p = 0.040 and aOR = 2.36; 95% CI: 1.20–4.64; p = 0.013, respectively), but not stress. Sensitivity analyses comparing models with and without adjustment for alcohol consumption showed similar effect estimates and did not materially alter the main conclusions; therefore, the fully adjusted model is presented as the primary analysis, with alternative specifications provided in Supplementary Table S3. A strong association between household debt and stress was found (aOR = 2.53; 95% CI: 1.20–5.33; p = 0.014). No significant associations were observed for sex, age, marital status, chronic disease, history of mental disorders, or household income sufficiency with any of the mental-health outcomes. When coping strategies and support systems were examined by mental health outcomes, most individual-level coping measures showed no statistically significant associations with depression, anxiety, or stress after multivariable adjustment. Among adaptive coping measures, only increased fertilizer use was inversely associated with depression (aOR = 0.55; 95% CI: 0.31–0.97; p = 0.040). Other coping and institutional support strategies were not significantly associated with mental health outcomes.

Discussion

This study contributes novel evidence on the mental health impacts of exposure to multiple climate-related hazards in a low- to middle-income agricultural setting. The findings show that exposure to multiple hazards, including floods, droughts, and wind storms, is significantly associated with adverse mental health outcomes among farmers. Farmers residing in areas with high levels of climate hazard exposure presented significantly higher rates of depression than those in areas with low levels of climate hazard exposure, while anxiety and stress were prevalent in both groups. Notably, farmers who experienced increased flooding within the past year had markedly greater odds of all three psychological outcomes, suggesting that repeated recent exposure to climate hazards imposes a cumulative mental burden on agricultural populations. Regular and occasional alcohol consumption were strongly associated with depression and anxiety, whereas household debt was significantly associated with stress, indicating that financial strain may exacerbate psychological distress through heightened economic insecurity. These findings highlight the complex interplay between climate instability, economic vulnerability, and mental health in farming communities. Reliance on individual versus institutional coping strategies likely reflects structural and contextual constraints, rather than intentional choice.

Mental health impacts of climate-hazard exposure among farmers

Interestingly, farmers in areas with high climate hazard exposure reported greater perceived financial security than did those in areas with low exposure; however, this pattern likely reflects structural differences in agricultural systems rather than a protective effect of climate hazard exposure itself. High-climate-hazard exposure areas were characterized by predominantly crop-based production with a larger share of medium- and large-scale farms, whereas low-climate-hazard exposure areas consisted largely of smaller holdings with greater reliance on livestock and fisheries, the sectors that typically generate lower and less stable returns. These differences in farming type and scale likely contributed to the higher levels of perceived financial security among farmers in areas with high levels of climate hazard exposure despite their heightened environmental risks.

The association between high climate hazard exposure and depression is consistent with previous research in Bangladesh, India, Vietnam, and other low- and middle-income settings where repeated exposure to floods and droughts has been linked to elevated rates of psychological distress, posttraumatic stress disorder, and depression among farmers [29]. In China, floods, typhoons, tornadoes, and storms are events that can increase mental health problems, such as depression, anxiety, and posttraumatic stress disorder (PTSD), with children, teenagers, women, and elderly individuals being the most affected; PTSD rates after disasters range from approximately 10% to over 50% [30]. In Vietnam, a post-disaster study of survivors of Typhoon Xangsane reported a prevalence of depressive symptoms of 5.9%, along with panic disorder and anxiety prevalences of 9.3% and 2.2%, respectively, within several months following the event [31]. In Korea, people who lived through Typhoon Rusa experiences clear mental health effects, with approximately 12% having moderate to severe PTSD and many reporting fear of another typhoon [32]. A more recent cross-country systematic review also documented disabling health impacts among older adults following climate-induced floods and typhoons in China and the Philippines [33–34]. Another cross-sectional analysis of depression amongst Australian rural business owners following cyclone-related flooding also confirmed the impacts of flood exposure and flood-related financial factors on probable depression (25%) as they felt undersupported by flood-related services [5]. Farmers are shown to have high mental health problems, which are often equal to or worse than those of non-farmers. For example, a study conducted in Bangladesh reported that cyclone-affected smallholder farmers reported high rates of acute stress (42%), anxiety (38%), and depression (36%) [35]. Another study in Bangladesh reported 25% posttraumatic stress disorder, 18% major depressive disorder, and 15% mixed anxiety–depression following Cyclone Sidr [36]. These studies highlight a consistent pattern across low- and middle-income settings. Repeated environmental disruption undermines psychological stability over time. Interestingly, increased storm frequency was associated with lower odds of depression during the past year. This counterintuitive finding may reflect the relative perceived severity of agricultural hazards. Storms are typically acute and transient, whereas droughts and floods constitute chronic stressors that can fundamentally undermine agricultural viability by causing total crop failure or preventing planting altogether. Consequently, storm-related damage may be viewed as more manageable, in contrast to the prolonged uncertainty and psychological burden associated with the slow violence of water scarcity and crop loss.

Behavioral and socioeconomic factors of climate-related mental health risks

Behavioral and socioeconomic factors also contribute to the mental health status of farmers. This study found that regular alcohol consumption was strongly associated with depression and anxiety. Sensitivity analyses showed that including alcohol consumption slightly attenuated the association between high climate hazard exposure and depression, whereas flooding frequency remained significantly associated with all mental health outcomes. These findings suggest that alcohol use is unlikely to act as a confounder but rather lies on the causal pathway between climate-related hazard exposure and mental health, acting as a stress-related coping behavior rather than a preceding cause, as supported by previous studies [37–40]. Likewise, household debt was independently associated with stress, reflecting how financial strain amplifies vulnerability to mental distress. Previous Thai and international studies have linked indebtedness to feelings of hopelessness and loss of control, suggesting that economic hardship after climate shocks may act as both a trigger and a sustaining factor for psychological morbidity [41–42]. These patterns illustrate how mental health, behavior, and economic resilience are interdependent dimensions of climate vulnerability. In addition, participation in community resource management programs was notably higher in low-exposure areas. Together with the greater uptake of adaptive farming techniques, such as improved or drought-resistant seed adoption, this pattern indicates stronger institutional and community engagement in areas with low levels of climate hazard exposure. These findings highlight the potential value of strengthening institutional and community-based adaptation mechanisms rather than relying solely on individual-level coping strategies.

Institutional support gaps and the dominant role of recurrent flooding

Farmers in areas with high levels of climate hazard exposure rely primarily on individual-level coping measures such as increasing fertilizer use and water storage, relying on self-reliant adaptation without structural support. In contrast, farmers in areas with low levels of climate hazard exposure benefited from collective or institutional supports, including community resource management programs, improved seed adoption, and financial compensation. Differences in coping strategies between high- and low-hazard exposure areas are more likely attributable to disparities in access to institutional resources and community-level support, rather than to farmers’ preferences. Farmers in areas with high levels of climate hazard exposure appeared to rely more on individual, self-directed coping measures, likely reflecting the limited availability of institutional support mechanisms, whereas those in areas with low levels of climate hazard exposure had greater access to community-based programs and formal adaptation resources. These forms of institutional support may play a protective role by enhancing a sense of agency and social connectedness, both of which are linked to lower rates of psychological distress. Notably, farmers were not asked whether increased fertilizer use was undertaken in response to climate change; therefore, fertilizer use should not be interpreted as a climate-specific coping strategy but rather as a routine farm-management practice or a short-term response to production constraints. Overall, stronger community and institutional supports in areas with low levels of climate hazard exposure appear to play a more important protective role in mental health than individual, self-directed practices do. Prior studies have also shown that social support networks and organized adaptation initiatives strengthen resilience and mitigate psychological distress in climate-affected communities [43–44]. Limited access to cooperative mechanisms in high climate-hazard exposure areas may therefore contribute to isolation and helplessness, intensifying depressive symptoms. The frequency of climate-related hazards in recent years has emerged as the most powerful predictor of psychological distress among farmers. Those who reported increased recent flooding had more than sixfold higher odds of depression, over threefold higher odds of anxiety, and nearly tenfold higher odds of stress. These findings demonstrate that recent exposure to floods heightens psychological vulnerability. This pattern aligns with previous evidence that recent flooding exposure is linked to worsening mental health through repeated trauma, economic loss, and ongoing uncertainty, which together could undermine livelihood security and contribute to prolonged psychological strain, including hyperarousal, sleep disturbance, and emotional exhaustion [45].

Implications for climate-related risk management policy and resilience practices

All results obtained from this study deliver important implications for public health and agricultural resilience in the context of climate change. Thailand’s occupational safety and health framework under the Occupational Diseases and Environmental Diseases Control Act B.E. 2562 (2019) provides a platform to integrate only physical health into hazard assessment and worker health surveillance [46]. The act should also include climate-related mental health issues as environmental diseases that require a national health surveillance program. Integrating mental health surveillance within climate adaptation programs is essential to protect vulnerable farming populations. Routine screening for depression, anxiety, and stress in high-risk districts should be embedded into primary-care and agricultural extension services to enable early detection and timely referral for counseling or psychosocial support. Cross-sector collaboration among the Ministry of Public Health, the Ministry of Agriculture, and local administrative organizations can facilitate coordinated delivery of mental-health outreach, financial-risk protection, and community-based adaptation training. Establishing local cooperatives and peer-support networks can strengthen social cohesion and reduce isolation, which are both key determinants of psychological resilience. Embedding mental health considerations as a central pillar of national climate adaptation policy is crucial to sustaining both productivity and well-being among farmers exposed to recurrent floods, droughts, and storms.

Strengths and limitations of the study

This study has several strengths. This study is among the few studies to explore the intersection between farmers’ mental health and climate change, a population often overlooked in research. Second, regional classification served as a proxy for cumulative exposure to multiple climate hazards, capturing broader exposure contexts on the basis of historical disaster records. Therefore, this study provides a more comprehensive understanding of how cumulative climate risk affects farmers’ psychological well-being. Third, this study applied the standardized Thai version of the DASS-21, which is a validated and widely used tool to assess depression, anxiety, and stress, ensuring reliable measurement of mental health outcomes. Nonetheless, certain limitations should be acknowledged. The cross-sectional design precludes causal inference, and self-reported data may be subject to recall or social desirability bias. The response rate of 57.7% also raises the potential for nonresponse bias, although the final sample remained diverse and well distributed across exposure groups. Additionally, the binary, high- versus low-exposure, classification may have introduced some misclassification, particularly because the low-exposure areas reported higher-than-expected drought frequency. However, the broader exposure categories were grounded in five-year disaster records, which provide a robust basis for differentiation. Therefore, any misclassification is likely minimal and unlikely to meaningfully alter the overall patterns observed. Future longitudinal studies should explore how farmers’ mental health changes over time after exposure to climate hazards and evaluate whether targeted adaptation or psychosocial interventions can mitigate these adverse effects.

Conclusions

This study presents novel evidence on the mental health impacts of multiple climate-related hazard exposure among low- to middle-income country farmers, one of the most vulnerable and frequently ignored worker populations, especially in flood-prone areas. Exposure to multiple climate-related hazards, including floods, droughts, and storms, was significantly associated with adverse mental health outcomes among farmers, with those residing in areas with high climate hazard exposure showing significantly higher rates of depression than those in areas with low climate hazard exposure. Notably, farmers who experienced increased flooding within the past year had markedly greater odds of depression, anxiety, and stress. A critical disparity in institutional support was found, leaving high-exposure farmers to cope individually. These findings emphasize integrating mental health screening, financial protection, and structural psychosocial support into climate adaptation frameworks to protect vulnerable agricultural populations.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Ratanon Kaewvijit: Data collection, formal analysis, interpretation of data, writing - original draft. Penradee Chanpiwat: Conceptualization, interpretation of data, writing – review & editing. Pornchai Sithisarankul: Writing – review & editing, supervision. Patthrarawalai Sirinara: Conceptualization, methodology, data collection, validation, formal analysis, interpretation of data, writing - original draft, writing – review & editing, supervision. Thammasin Ingviya: Conceptualization, interpretation of data, writing – review & editing.

Funding

None.

Data availability

The datasets used and/or analyzed in this study can be obtained from the corresponding author upon request.

Declarations

Ethics approval and consent to participate

This study was conducted following the Declaration of Helsinki guidelines, and the protocol was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University (IRB No. 0157/68, COA No. 0595/2025).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.