Flooding can have a serious effect on people’s health and well-being, both psychological and physical. In common with other types of major disasters, the effects on people’s health, relationships, and welfare can be affected by flooding. Floods are one of the most common types of global natural disasters responsible for almost 53 000 deaths in the last decade. Reference Alderman, Turner and Tong1 Flooding can cause substantial social and welfare problems that may linger over extended periods because of not only being flooded (the primary stressor), but also because of the secondary stressors (those stressors that are indirectly related to the initial extreme event, ie, economic stress associated with re-building) that arise as people try to recover their lives, property, and relationships. Flooding can challenge the psychosocial resilience of the hardiest of people who are affected. Reference Stanke, Murray and Amlot2 Psychosocial impacts of flood events suggest that they can have significant adverse effects on people’s well-being, relationships, and mental health.Reference Fewtrell and Kay 3
Climate change has contributed to an increase in the intensity and frequency of floods. It will also worsen hurricanes, heat waves, droughts, wildfires, dust storms, and tornadoes. Reference Stanke, Murray and Amlot2 This is because climate change entails changes in precipitation, along with temperature, wind patterns, atmospheric pressure, cloudiness, and humidity.
In their 2005 review on the global health impacts of flooding, Ahern et al. Reference Doocy, Daniels, Murray and Kirsch4 reported several epidemiological studies that assessed the effects of flooding on common mental disorders (including anxiety and depression), posttraumatic stress disorder (PTSD), and suicide. Most studies examining the effects of flooding on common mental disorders are from high or middle-income countries, and results revealed significant increases in depression, anxiety, and psychological distress among flood-affected adults. This means that experiencing a flood effect can lead to long-term negative psychological effects, long after the flooding itself has abated. Long-term health problems, in particular, adverse mental health outcomes, have been linked to the experience of flooding, including elevated levels of depression, anxiety, and PTSD in affected populations. Reference Fernandez, Black and Jones5–Reference Gray7 Several factors are known to be associated with the likelihood of experiencing adverse mental health outcomes post-flooding. These include sociodemographic characteristics such as age, gender, socioeconomic and education statuses, and factors related to the nature of flooding (eg, depth and duration). Reference Ahern, Kovats and Wilkinson8 Secondary stressors, including problems with relationships, loss of sentimental items, and difficulties with insurance or compensation, have also been associated with poor psychological outcomes.
Fort McMurray located in Alberta, Canada, has experienced several major natural disasters over the past 5 years. These include wildfires associated with an evacuation order in 2016, flooding in 2020, and also the COVID-19 pandemic in 2020.
These traumatizing experiences are believed to have adversely impacted the mental health and general well-being of survivors. Reference Stanke, Murray and Amlot6,Reference Brown, Agyapong and Greenshaw9 On April 27, 2020, a major ice jam and melting waters caused massive flooding in downtown Fort McMurray. This forced over 15 000 people to be evacuated according to media reports. Reference Snowdon10 At the individual level, experiencing an episode of flooding may trigger memories of previous flooding episodes, intensifying the distress experienced. Reference Brown, Agyapong and Greenshaw9 Multiple factors associated with flooding could trigger psychological distress, including fears and actions taken to protect family or belongings, experience of flooding, and long-term uncertainties concerning insurance. Reference Foudi, Osés-Eraso and Galarraga11–Reference Tapsell, Tunstall and Wilson13 Additionally, it is plausible that the act of evacuation itself would also trigger stressful memories. Those who have been previously affected by flooding may also experience anticipatory anxiety and stress about the prospect of future flooding. Reference Tunstall, Tapsell and Green14,Reference Mao and Agyapong15 This study assessed the clinical correlates of Major Depressive Disorder (MDD) following a flooding incident at Fort McMurray.
Methods
Study Setting
The study was conducted in Fort McMurray, Alberta, Canada. Fort McMurray is the urban service area of the Northern Alberta Regional Municipality of Wood Buffalo, with a diverse population of 111 687 as of the 2018 census, 16 mainly in temporary accommodations (hotels, motels, and boarding house) and in adjoining rural communities. Fort McMurray is a multicultural community, attracting people from all corners of Canada and the world. Generally, moves to Fort McMurray have increased in the last decade. The ethnic groups include 65.8% white, 22.1% Aboriginal, 3.8% South Asian, 2.5% Filipino, 1.9% black, and 1.2% Arab. The total number of dwelling units in the Municipality in 2020 is 28 281. This represents an 8.4% decrease from the total number of dwelling units in 2015. The decrease is largely attributed to dwelling units that were destroyed by the 2016 Horse River Wildfire. Reference Agyapong, Juhas and Omege17 The study site is considered the heart of one of Alberta’s (and Canada’s) hubs of petroleum production, located near the Athabasca oil sands. Besides the oil sands, the economy also relies on natural gas and pipeline transport, forestry, and tourism. People typically move to the Fort McMurray area to work in the oil sands industry. Reference Agyapong, Juhas and Omege17 The major employers of labor in Fort McMurray are oil companies, which made Fort McMurray a thriving industrial center that attracted an influx of oil-industry workers who took up both permanent and temporary residences.
Study Design
This study adopted a descriptive cross-sectional study design. The study was carried out between April 24 and June 2, 2021, through a self-administered questionnaire that was sent to residents of FMM via REDCap. Reference Harris, Taylor and Thielke18
Ethical Consideration
Study approval was granted by the Research Ethics Board at the University of Alberta (Pro00066054). This study was also conducted in accordance with the Declaration of Helsinki (Hong Kong Amendment).
Data Collection
Data collection tools for this study were developed based on published literature and questions from previously validated instruments. The general constructs of interest included relevant demographic information, flood-related questions, and the mental health status of respondents. A standardized measure for likely MDD included in the survey was the Patient Health Questionnaire-9. Reference Kroenke, Spitzer and Williams19
REDCap is a secure web application for building and managing online surveys and databases. It is also used to collect virtually any type of data in any environment. It is specifically geared to support online and offline data capture for research studies and operations. Reference Harris, Taylor and Thielke18 The questionnaire was distributed randomly via email using government, school, occupational, and community platforms. Consent was implied, by completing the survey. Links to self-administered questionnaires were distributed randomly via email using intermediary agencies such as government, schools, and occupational and community platforms.
Statistical Analysis
A data analysis was undertaken using SPSS Statistics for Windows (Version 26; IBM Corp, Armonk, NY). 20 Demographic characteristics, flooding-related variables, as well as responses to questions related to MDD were summarized by absolute numbers and percentages. Only completed responses were reported, with no data imputation. The chi-square test and Fisher’s exact analysis with 2-tailed significance (P ≤ 0.05) were performed to assess the association between the demographic characteristics and responses to questions related to flooding and MDD, as measured using the Patient Health Questionnaire (PHQ-9) index. The PHQ-9 (included in appendix) is a multipurpose instrument used for screening, diagnosing, monitoring, and measuring the severity of depression, and it is a reliable and valid measure of depression severity. Reference Kroenke, Spitzer and Williams19 the PHQ-9 scale is half the length of many other depression measures, has comparable sensitivity and specificity, and consists of 9 criteria upon which the diagnosis of Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), fourth edition, depressive disorders is based. The PHQ-9 scoring was done using the standard recommendation with the threshold for likely moderate to high depression being met if 5 of the 9 items were checked and the total score was ≥ 10.
Variables with statistically significant or near-significant association (P ≤ 0.1) for each of the sociodemographic and flooding domains and MDD were entered into a logistic regression model. Before performing the logistic regression analysis, correlational diagnostics were performed to identify any strong intercorrelations (Spearman’s correlation coefficient of 0.7 to 1.0 or -0.7 to -1.0) among predictor variables. Odds ratios from the binary logistic regression analysis were calculated to determine the association between the predictor variables and the presence of MDD controlling for the other variables in each model.
Results
Out of the total of 249 persons who accessed the survey, there were 186 responses received, giving an effective response rate of 74.7% (see Table 1). Of the 186 respondents who completed the survey, 85.5% (159) of the respondents were female while 14.5% (27) were male, 52.7% (98) were above 40 years of age, 40.3% (75) were between the ages of 26 and 40 years, and the lowest proportion (7.0%; 13) was less than 25 years of age. Also, 94.1% (175) were employed and 5.9% (11) were unemployed. The majority (71%; 132) of the respondents were married or partnered, 19.4% (36) were single, and 9.7% (18) were either divorced or separated at the time of this study.
Table 1. Demographic profile, clinical characteristics, and received support for the study population
A larger percentage of the respondents, 94.6% (176), resided in Fort McMurray during the year 2020 flood while 5.4% (10) of the respondents were temporary residents in Fort McMurray during the flood. The majority, 75.8% (144), own a home, 24.2% (45) were renting a home. Nearly half, 48.4% (90), of the respondents reported that they had not received a mental health diagnosis from a health professional following the flooding. The study also showed that 31.2% (58) of the respondents received a mental health diagnosis for depression, and 41.9% (78) of the participants received a diagnosis for anxiety. However, 1.1% (2) received a mental health diagnosis of bipolar disorder and personality disorder while a few, 9.1% (17), received other diagnoses from a health professional following the flooding.
The respondents reported that they received antidepressants, 31.7% (59); sleeping tablets, 11.3% (21); mood stabilizers, 6.5% (12); antipsychotics, 2.2% (4); benzodiazepines, 2.2% (4); and other medications, 1.6% (3), while 64.5% (120) were not on any medication due to mental health. Most of the respondents, 61.3% (114), did not receive mental health counseling in the past year, while 38.7% (72) received mental health counseling in the past year; 50.7% (98) reported that they would like to receive mental health counseling, while 47.3% (88) declined mental health counseling.
Seventy percent (123) reported that they were fearful of their life and family during the flooding, 29.3% (51) were not fearful; 66.7% (116) watched television images about the devastation caused by the flood on a daily basis; 21.3% (37) and 12.1% (21) did not watch daily and did not watch at all, respectively. Most of the respondents, 88.7% (156), did not suffer any loss of property during the flood, while 11.3% (21) reported that they suffered a property loss. Most, 85% (147), were living in the same household before the floods, 12.7% (22) were living in a different house even though their previous home wasn’t destroyed, and 2.3% (4) reported that they live in a different house because their previous home was destroyed by the flood.
Forty-four percent (74) received absolute support from family and friends, 15.6% (26) had some support, and 10.2% (17) only had limited support from family and friends (absolute support from family and friends implies full support without any restriction or condition while limited support is referred to not receiving full support from their family and friends); 6.4% (11) of the respondents claimed that they received absolute support from the Government of Alberta, 4.1% (7) had some support, 3.5% (6) got limited support, 7.0% (12) did not receive any support, and 78.9% (135) reported they were not impacted by the flood, hence there was no support from the Government of Alberta. The prevalence of mild to severe depression among the respondents was 53.7% (75).
Table 2 shows a summary of the results of univariate analysis of the association between respondents’ demographic profile, flood-related variables, and moderate to severe depression.
Table 2. Chi-square test of association between demographic profile, clinical, flooding-related variables, and likely depression
The chi-square test showed a significant (P ≤ 0.05) relationship between sociodemography, clinical variables, and moderate to severe depression. Nine variables were statistically significant, including (1) employment status, (2) received mental health diagnosis on depression from a health professional, (3) received mental health diagnosis on anxiety from a health professional, (4) received mental health diagnosis from a health professional, (5) would like to receive mental health counseling, (6) on antidepressant medication, (7) on benzodiazepine medication, (8) on sleeping tablet medication, and (9) on any mental health medication for health concerns.
Table 3 below shows the logistic regression model to determine the likelihood of moderate to severe depression among the study participants in Fort McMurray.
Table 3. Logistic regression model for respondents’ likelihood to present with likely MDD
*P < 0.05.
The multivariate binary logistics regression model was statistically significant; χ2(df = 11; n = 162) = 59.90, P < 0.001; this indicated that the multivariate binary logistic regression model could differentiate between the respondent who had a moderate to severe depression and respondents who had a mild depression.
The model accounted for 30.9% (Cox and Snell R2) to 41.3% (Nagelkerke R2) of the variance. According to the goodness-of-fit statistic using Hosmer-Lemeshow goodness-of-fit test, the model was adequately fit (χ2 = 5.34; P = 0.72) and correctly classified 75.9% of cases.
According to Table 3, only 3 variables—employment status, received a mental diagnosis of depression, and would like to receive mental health counseling—significantly predicted likely MDD. Study participants who reported that they are unemployed are 12 times more likely to present with MDD symptoms (OR = 12.16; 95% CI: 1.08–136.26).
Study participants who have received a mental health diagnosis of MDD are 5 times more likely to have moderate to severe depression (OR = 5.306; 95% CI: 1.84–15.27). Respondents who would like to receive mental health counseling are 6 times more likely to show MDD symptoms (OR = 6.12; 95% CI: 2.41–15.49).
Discussion
This study provides a strong indication of an association between flooding and probable MDD. Given projections that both the frequency and severity of flooding events will continue to increase in the future, Reference Alfieri, Bisselink and Dottori21–Reference Paul and Moser23 populations living in flood-susceptible areas may be exposed to multiple flooding events, which may cause repeated mental health issues.
In this study, unemployment showed a relationship with likely MDD among the study participants. Climate change and economic downturns have become a global problem that has been influencing many of the key social determinants of health, especially through unemployment and changes to living conditions. A natural disaster such as flooding is characterized by a rise in unemployment, reductions in income, as well as a reduction in total well-being. Unemployment is associated with an increase in distress and psychosomatic symptoms, anxiety, depression, melancholy, feeling of dependency, inability to solve routine problems, self-dissatisfaction, and lower self-esteem. Reference Warr, Drenth, Thierry and Willems24,Reference Cao, Wang and Cao25 While this study has shown an association between unemployment and depression in this vulnerable population exposed to flooding, it has not ascertained causality.
In our study, the prevalence of mild to severe depression among the respondents was 53.7% (75); this is in line with a prevalence of 54% among adults in a study on depression following natural disaster. Reference Tang, Liu and Liu26 Our study also showed that respondents who have received a mental health diagnosis for depression are more likely to have moderate to severe depression. Those who were already diagnosed with MDD seem to be having MDD symptoms, meaning that they were either not controlled on medications or therapies they received or the experienced trauma was out of control. In a community sample, Cao et al. Reference Cao, Wang and Cao25 identified 4 response patterns in the outcome of a natural disaster: (1) minimal symptoms, (2) predominant depression, (3) predominant PTSD, and (4) comorbid PTSD and depression. Of note, persons with a predominately depressed presentation represented approximately one-fifth of the community sample (and 40% of those presenting with symptoms) and reported more somatic complaints and a high level of interpersonal problems compared to those considered to be PTSD symptoms.
Experiences during the natural disaster (eg, flooding) may be more indicative of PTSD following a natural disaster compared to depression. Meanwhile, a collection of studies shows that further processes, specifically disaster-related stressors and individual-level factors, may lead to post-disaster depression. Reference Brewin, Andrews and Valentine27–Reference Stander, Thomsen and Highfill-McRoy29 Furthermore, past symptoms of depression may also serve as a strong indicator of post-disaster depression. Stander et al. Reference Stander, Thomsen and Highfill-McRoy29 proposed a causal relation between PTSD and depression, in which pre-military trauma PTSD forecasted depressive symptoms succeeding exposure to war, signifying a strong relation between lifetime PTSD and post-disaster depression can have important implications in disaster-preparedness protocols within mental health services. Reference Tracy, Norris and Galea30
The study population has experienced multiple natural disasters in the past 5 years, and in view of climate change, this may become the norm in other places, making it an ideal population. However, because they have experienced other natural disasters may act as a confounding factor either by increasing resilience or increasing vulnerability to developing mental health problems like depression.
MDD has a substantive negative impact on quality of life. Several studies suggest that depression is underdiagnosed and undertreated amongst older people in Western nations. Reference Mitchell, Rao and Vaze31–Reference Briggs, Kirchhoff and Hammes34
Respondents who would like to receive mental health counseling were more likely to express moderate to severe depression. Seeking counseling from a health care professional with a self-care focus contributes greatly to the treatment of any illness. Recent studies have shown that counseling encourages patients to actively contribute in treatment programs and promotes their positive outcomes. Reference Höglund, Winblad, Arnetz and Arnetz35,Reference Macdonald, Rogers, Blakeman and Bower36 Counseling helps the patients control their feelings, adhere to the therapy, clearly understand the treatment rationale, improve quality of life, decrease stress, reduce anxiety, feel more secured, and increase life satisfaction. Reference Macdonald, Rogers, Blakeman and Bower36
One of the factors that influences people’s perception of flood disaster preparedness is an early warning system. Early warning systems are an important part of community preparedness mechanism. Reference Marfai, Sekaranom and Ward37 Shelter becomes a disaster preparedness factor that cannot be excluded in flood disaster management. People affected by floods or vulnerable to disasters are usually advised to evacuate. The existence of shelter is also one of the factors influencing community preparedness.
Studies on disaster victims found that changes in substance abuse vary from mild to moderate increases among diverse populations. Reference David, Mellman and Mendoza38–Reference Knudsen, Roman, Johnson and Ducharme40 For example, a study following Hurricane Andrew of August 1992 found only 2% of area residents reporting alcohol dependence. Reference David, Mellman and Mendoza38
Study Limitations
Ethnicity and income may be potential confounders given that some studies have found correlations between these variables and mental health. Reference Phelan, Link and Diez-Roux41,Reference Link, Northridge, Phelan and Ganz42 Unfortunately, this study did not collect data on these. As part of an online survey limitation, the population to which gave responses cannot be described, and respondents with biases may select themselves into the sample.
Additionally, there may be a strong indication that the results of this study would have been different if the study were conducted prior to the flooding. Also, flooding at Fort McMurray occurred during the COVID-19 pandemic, so considering that some people may have been exposed to COVID-19 during that time, it is our assumption that this may have affected our results. We mainly focused on flooding variables and their correlation with Major Depression Disorder, through a cross-sectional study design that captured the data at the time.
Conclusions
This study makes a good contribution to the growing literature on the mental health effects of flooding, and its large scale and natural disaster also represent a unique opportunity to increase the understanding of the impacts of flooding on population health and well-being. It is important that the impact of flooding on people’s health and well-being is carefully understood. This will help inform public health actions to mitigate adverse health consequences of future flooding events.
Acknowledgments
Support with survey link distribution was received from the Fort McMurray Public and Catholic School Boards, Keyano College, and the Canadian Mental Health Association.
Author contributions
The study was conceived and designed by VIOA. FO drafted the initial manuscript, and RS conducted the data analysis. EE contributed to the data collection. All authors contributed to the study design, reviewing, and revising the initial draft manuscript and approved the final draft before its submission.
Funding statement
This study was supported by grants from the Mental Health Foundation and the Douglas Harden Trust Fund.
The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the results for publication.
Conflict(s) of interest
None.
Open access
