Natural hazards, such as floods, may have a profound and lasting impact on mental health, resulting in elevated levels of anxiety, depression, and posttraumatic stress disorder (PTSD) among affected populations.Reference Saeed and Gargano1, Reference Keya, Leela and Habib2 Recent meta-analyses reveal that the prevalence rates of these mental health conditions in the context of floods vary significantly—ranging from 2.2%-84% for anxiety, 3.23%-52.70% for depression, and 2.6%-52% for PTSD—with the most severe impacts observed in lower-income countries and among lower socioeconomic groups.Reference Mulchandani, Armstrong and Beck3
Factors such as relocation, service disruptions, pre-existing psychopathology, exposure and loss, socioeconomic status, sex, age, personality, biological vulnerability, and social support are significantly associated with trauma psychopathology.Reference Saeed and Gargano1, Reference Mulchandani, Armstrong and Beck3 However, the trajectory of early symptoms and the long-term progression of PTSD require further investigation. For instance, acute stress disorder (ASD) has not consistently predicted subsequent PTSD in longitudinal studies.Reference Bryant4 Contextual factors such as community resilience and social support also have to be taken into account.Reference Apel and Coenen5-Reference Diefenthaeler, Hartmann and Bassols7 Psychological effects can persist long after physical damage is repaired, underscoring the need for comprehensive studies to understand these impacts and support recovery.Reference Mulchandani, Armstrong and Beck3, Reference Apel and Coenen5, Reference Lamond, Joseph and Proverbs6 Hence, longitudinal, culturally sensitive approach is essential for examining how mental health symptoms evolve over time.
The major flood in May 2024 in Rio Grande do Sul (RS), Brazil (GLIDE number: FL-2024-000063-BRA), is considered the most significant catastrophe in the state’s history. It affected 2 398 255 individuals (22% of the population), displaced 581 643 people, sheltered 80 826, and resulted in the rescue of 82 666 people and 12 503 animals.8 The flood particularly impacted lower-income communities, with some areas remaining submerged for nearly a month. An ongoing web-based longitudinal study is tracking mental health outcomes during the flood and at subsequent intervals: 30-90 days later, 6 months, and 1 year. The findings aim to inform timely interventions. Preliminary analyses and interventions designed to mitigate the disaster’s impact are being shared with local media and stakeholders, focusing on psychoeducation and preventive measures to enhance community resilience.Reference Diefenthaeler, Hartmann and Bassols7, Reference Biolo, Hauck and Umpiere9
The preliminary results presented in this paper cover 64 individuals evaluated during the floods (May 10-18, 2024) and about 1 month later (June 10-22, 2024). The analyses examine sociodemographic factors, the extent of impact (relocation, material losses, disruption of services, and personal loss), and symptoms of depression, anxiety, acute stress, and PTSD, to elucidate how mental health impacts evolve over time.
Methods
Inclusion criteria were residing in RS at the onset of the flood, being over 18 years of age, and agreeing to participate in the follow-up by providing an email for contact. Participants provided informed consent online. The study was conducted in collaboration with the Ethics Committee of the Hospital de Clínicas de Porto Alegre (HCPA; CAAE 80017124.7.0000.5327), which rigorously reviewed and approved the research protocol. This preliminary study is part of an ongoing mixed-method study that will periodically evaluate the mental health impact of the flooding on the population of RS, using both longitudinal and repeated cross-sectional assessments.
Procedure
Due to the significant impact of the flood, which caused severe disruptions in essential services and transportation, an online strategy was the only feasible method for assessing those affected in the aftermath. A snowball sampling method was used, targeting residents of RS who were affected by the flood. The survey was administered through the SurveyMonkey platform and dedicated social media accounts were used to disseminate the online survey. Participants in this analysis were initially evaluated between May 10 and May 18, 2024 (Time 1, T1), and approximately 30-40 days later, between June 10 and June 18, 2024 (Time 2, T2). The initial data collection began about 5 days after the major flooding on May 6, 2024, in downtown Porto Alegre, the capital of RS. Although the downtown flooding subsided by May 16, many neighborhoods remained inundated for over 25 days.8
Measures
Demographic variables included sex, age, ethnicity/race, educational level, family income, and the city and neighborhood where participants resided at the onset of the flood. Participants were also asked whether they were directly affected by the flood, indirectly affected (through someone close to them), or not affected at all. Specific impacts assessed included urgent evacuation needs, partial or total loss of housing and belongings, loss of basic supplies, and loss of access to housing. Additionally, participants’ perceptions of their support network and their need for mental health support during and after the event were evaluated.
The Patient Health Questionnaire-9 (PHQ-9) and the Generalized Anxiety Disorder-7 (GAD-7), both concise self-report instruments with strong reliability and validity, were used to assess depressive and anxiety symptom severity, respectively. A cutoff score of 10 indicated the possible presence of clinically relevant symptoms for both measures. At T1, participants responded with yes (1) or no (0) to ASD symptoms listed in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5).10 At T2, the Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5) assessed PTSD symptoms, with a conservative cutoff score of 38 (sensitivity 70-80; specificity > 85) used to infer probable PTSD.
Data analysis
The data distribution was assessed using histograms, Q-Q plots, and evaluations of skewness and kurtosis. Descriptive statistics were reported as means and standard deviations (SD), medians and interquartile ranges (IQR), or absolute and relative frequencies. Inferential methods were selected based on the data distribution.
Differences in clinical scores between T1 and T2 were evaluated with paired t-tests or the Wilcoxon Signed-Rank Test. The McNemar test compared frequencies of categorical variables. Pearson’s correlation coefficient (r) and Spearman’s rank correlation coefficient (ρ) assessed the strength and direction of associations. For differences between independent groups, either the t test or the Mann-Whitney U test was used.
Moreover, we created standardized scores for common ASD and PTSD dimensions at both time points (including re-experiencing, avoidance, negative alterations in cognition and mood, and hyperarousal) by using items that were equivalent between the ASD checklist at T1 and the PCL-5 at T2. Delta scores for each dimension and for the total dimension scores (sum of dimensions’ scores for each time point) were then generated by subtracting T1 scores from T2 scores (T2-T1). These delta scores were used as the dependent variable to examine factors influencing symptom progression, accounting for the paired nature of the study. This approach allows for a precise assessment of within-subject changes and controls for initial symptom severity.
The significance level was set at 5% (P < 0.05), and all statistical tests were 2-tailed. Analyses were conducted using SPSS IBM software version 29.0.
Results
The sample included 64 participants. Data collection during the floods (T1) had a median date of May 11, 2024 (IQR: May 10-13, 2024), and follow-up (T2) had a median date of June 12, 2024 (IQR: June 11-14, 2024), with an average interval of about 32 days between the 2 time points.
Table 1 presents the main demographic, flood-related, and clinical variables for the sample.
Demographic, flood-related and clinical characteristics from the sample

PHQ-9 = Patient Health Questionnaire-9, used for screening depression symptoms; GAD-7 = Generalized Anxiety Disorder-7, used for screening anxiety symptoms; PCL-5 = Posttraumatic Stress Disorder Checklist for DSM-5; T1 = Time 1 (May 10-18, 2024); T2 = Time 2 (June 10-18, 2024); BRL: Brazilian Reais.
a On June 10, 2024, 1 dollar was equivalent to 5.37 reais.
Longitudinal Perspective
Paired samples t tests revealed no statistically significant difference in PHQ-9 scores from T1 to T2 (t(63) = 1.56, P = 0.124, 95% CI [−0.319, 2.569]), and a trend toward reduction in GAD-7 scores (t(63) = 1.71, P = 0.093, 95% CI [−0.209, 2.647]). The McNemar test showed no statistically significant change in the percentage of those scoring above the cutoff point for both measures. However, significant correlations were found between T1 and T2 PHQ-9 scores (r = 0.678, P < 0.001) as well as T1 and T2 GAD-7 scores (r = 0.537, P < 0.001).
The Wilcoxon Signed-Rank Test revealed significant increases in standardized posttraumatic dimensional and total scores from T1-T2, with moderate to large effect sizes (r): Re-experiencing (Z = −4.74, P < 0.001, r = 0.59); Avoidance (Z = −4.69, P < 0.001, r = 0.58); Cognitions and Mood (Z = −3.89, P < 0.001, r = 0.48); Hyperarousal (Z = −2.77, P = 0.006, r = 0.34); Total Score (Z = −5.45, P < 0.001, r = 0.68). At T1, hyperarousal symptoms were most prominent, followed by re-experiencing, with cognitive and mood alterations and avoidance being less prevalent. By T2, all symptoms increased, particularly re-experiencing and avoidance. For instance, avoidance symptoms rose from 2 individuals (3.1%) at T1 to 24 (37.5%) at T2 (McNemar test: P < 0.001). Correlations between total and dimensional delta scores with PCL-5 scores are shown in Table 2. Significant correlations between total and dimensional delta scores with total PCL-5 in T2 reinforce the validity of the standardized scores.
Pearson’s correlation coefficient between Deltas (T2 -T1) and PCL-5 scores at T2

PCL-5 = Posttraumatic Stress Disorder Checklist for DSM-5; T1 = Time 1 (May 10-18, 2024); T2 = Time 2 (June 10-18, 2024); D = Delta (score at Time 2 – score at Time 1); Delta Total (D-T) = Total PTSD score at T2 – Total ASD score at T1; Delta Re-experiencing (D-RE) = Re-experiencing score at T2 – Re-experiencing score at T1; Delta Avoidance (D-AV) = Avoidance score at T2 – Avoidance score at T1; Delta Arousal and Reactivity Symptoms (Hyperarousal) (D-H) = Hyperarousal score at T2 – Hyperarousal score at T1; Delta Negative Alterations in Cognitions and Mood (D-CM) = Cognitions and Mood score at T2 – Cognitions and Mood score at T1.
**P <0.001.
Specific ASD symptoms at T1 were associated with PCL-5 scores at T2, suggesting that certain ASD symptoms may predict later PTSD severity. The symptom “I have distressing dreams related to the event” exhibited the strongest correlation with PCL-5 scores (ρ = 0.430, P < 0.001), followed by “I have recurrent and involuntary memories of the event” (ρ = 0.364, P = 0.003), “I experience intense emotional distress when reminded of the event” (ρ = 0.326, P = 0.009), “I have difficulty concentrating” (ρ = 0.303, P = 0.015), and “I am unable to feel positive emotions, such as happiness or satisfaction” (ρ = 0.260, P = 0.038). Additionally, “I have recurrent and involuntary memories of the event” (ρ = 0.299, P = 0.017) and “I have distressing dreams related to the event” (ρ = 0.249, P = 0.047) were also significantly correlated with the delta total score.
The percentage of individuals receiving psychological or psychiatric support increased from 39.1%-51.6%, showing a near-significant trend towards greater uptake (McNemar test: P = 0.077). Standardized posttraumatic symptom severity significantly differed between T1 and T2 for those who urgently left their homes (U = 590.000, P = 0.017) and those who lost housing access (U = 578.000, P = 0.023). Variables such as “loss of basic supplies” and “road closures” approached significance. A cumulative impact score, representing the sum of these impacts, was significantly correlated with changes in posttraumatic symptoms (ρ = 0.394, P = 0.001) and the PCL-5 total score at T2 (ρ = 0.439, P < 0.001), indicating that greater cumulative impact is linked to increased PTSD severity over time. Additionally, participants with a prior psychiatric diagnosis showed a trend toward greater increases in posttraumatic symptoms (U = 647.000, P = 0.065) and higher PCL-5 scores (t(62) = −1.602, P = 0.057).
Discussion
This longitudinal analysis, which covers the immediate aftermath of the flood (T1) and approximately 1 month later (T2), provides important insights into trauma and stressor-related disorders symptom progression. Such timely sensitive data related to catastrophic weather events of this magnitude are scarce in the literature, as the impact of the event often hinders data collection.Reference Keya, Leela and Habib2
Previous studies indicate that an ASD diagnosis lacks sensitivity in predicting individuals who will later develop PTSD, suggesting that at least 1/3 of those who go on to develop PTSD do not meet ASD criteria.Reference Bryant4 This aligns with our findings, as ASD/PTSD symptoms showed a marked increase between T1 and T2, while anxiety and depression symptoms remained relatively stable. This shift suggests that the impact of trauma may follow a time-dependent trajectory that is more specifically linked to PTSD than to general mood disorders. These findings are consistent with previous research on heterogeneous trauma response trajectories,Reference Pietrzak, Van Ness and Fried11 including delayed-onset PTSD, as well as evidence indicating that PTSD’s biological mechanisms may develop over time and be influenced by subsequent stressors.Reference Bryant4
Moreover, the observed increase in posttraumatic symptoms from T1-T2 reveals a specific pattern that can guide the development of timely interventions. Although all symptom dimensions presented an increase, marked rises in re-experiencing and avoidance symptoms suggest a shift from immediate hyperarousal to more entrenched symptoms of reliving the trauma and avoiding reminders. These findings dialogue with a previous ASD/PTSD symptomatology network analysis,Reference Pietrzak, Van Ness and Fried11 which shows an important association between intrusive memories and associated reactivity as central to ASD, as well as a correlation between intrusive memories and avoidance of thoughts. This suggests that early psychoeducational interventions targeting re-experiencing symptoms and preventing avoidance behaviors among flood-affected populations could mitigate the development of more chronic PTSD. In this way, public policies that facilitate mental health care to affected individuals and disseminate information about “red-flag symptoms” that indicate the need for psychological evaluation should be part of any community recovery plan.
However, the study has several limitations. Reliance on self-reported measures may introduce response biases and, although a significant portion of participants experienced severe disruption, the predominantly white, female sample may not fully represent the broader population affected by the flood. Additionally, while the sample size was adequate for detecting trends, it limits the generalizability of the findings. The use of a web survey, necessary due to the circumstances, resulted in a convenience sample, further restricting the ability to generalize the findings. Despite these limitations, the observed increase in posttraumatic symptoms among those more exposed aligns with existing literature and supports the internal validity of the study.Reference Saeed and Gargano1–Reference Lamond, Joseph and Proverbs6, Reference Pietrzak, Van Ness and Fried11, Reference Bryant, Creamer and O’Donnell12 Furthermore, the stability of depressive and anxiety symptoms alongside rises in posttraumatic dimensions in the month following the event raises important questions that warrant further investigation. Future research should consider larger and more diverse samples to further explore these dynamics and refine strategies for supporting individuals affected by disasters.
Conclusion
The results suggest a time-dependent factor in the full development of PTSD symptoms following trauma. Initially, hyperarousal appears to be more prominent but gradually gives way to re-experiencing and avoidance symptoms, which are central to PTSD. Early psychoeducational interventions targeting re-experiencing symptoms and avoidant behavior may help mitigate PTSD severity in affected individuals. While these findings align with previous research, further investigation in larger and more diverse samples is needed to assess generalizability. Continued data collection in the affected regions, including planned follow-up assessments, is expected to generate additional evidence on the long-term trajectories of PTSD symptoms and contribute to a more refined understanding of their evolution over time.
Data availability statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
Funding statement
This project received financial support from the Cordenação de Aperdeiçoamento de Pessoal de Nível Superior (CAPES) and the Fundo de Incentivo à Pesquisa e Eventos do Hospital de Clínicas de Porto Alegre (FIPE-HCPA).
Competing interests
The authors declare no conflict of interest.
Ethical standard
The study was conducted in collaboration with the Ethics Committee of the Hospital de Clínicas de Porto Alegre (HCPA; CAAE 80017124.7.0000.5327), which rigorously reviewed and approved the research protocol. Additionally, the study received approval from the Ethics Committees of the Federal University of Rio Grande do Sul and Porto Alegre.
Participants provided informed consent online, and the study adhered to the latest version of the Declaration of Helsinki.
