The links of stress, substance use and socio-demographic factors with domestic violence during the Covid-19 pandemic in Portugal

Abstract

Background

Lockdown, as a measure implemented to combat the coronavirus disease 2019 (COVID-19) pandemic, left many domestic violence (DV) victims trapped with their abusers. This study intends to explore the links between perceived stress, substance use and socio-demographic factors with DV experiences during COVID-19 pandemic in Portugal.

Methods

A cross-sectional study was carried out on a sample of 1062 participants over 16 years old, residing in Portugal. Data were collected through an online survey conducted between April and October 2020. The associations between potential factors and DV were investigated using bivariable analysis and multivariable logistic regression.

Results

The prevalence of DV reported was 13.75% (n = 146), disaggregated into psychological violence (13%, n = 138), sexual violence (1.0%, n = 11) and physical violence (0.9%, n = 10). Multivariable analyses confirmed that perceived financial difficulties (OR = 1.608; P = 0.019), use of medications to sleep or calm down (OR = 1.851; P = 0.002) and perceived stress (OR = 2.443; P = 0.003) were responsible for DV exposure during COVID-19 pandemic. Younger age (<25 years old) and consumption of alcohol were associated with a higher risk of DV victimization.

Conclusions

Interventions aimed at preventing and confronting DV are necessary within the strategies to combat COVID-19 in Portugal, especially aimed at groups in vulnerable situations, during and after the pandemic.

Introduction

Worldwide, domestic violence (DV) represents a significant public health problem.^1,2 It refers to a range of abuse, including physical, sexual and psychological, in any relationship within households that involve intimate partners, relatives, other members, and includes children, adults or elders.³ The abuse can comprise any behaviour that frightens, intimidates, terrorizes, manipulates, hurts, humiliates, blames, injures or wounds someone.⁴ Incidents often escalate from threats and verbal assault to life-threatening violence or even homicide.^5,6 Given the nature of its consequences affecting personal integrity, DV constitutes a serious violation of international human rights law.⁷ The magnitude of DV impact, in the short and long term, whether in physical, psychological, financial or other terms, is felt in all areas of health and social life of the victims, their families and communities.⁸ It affects psychological and physical health, reduces the quality of life and results in decreased productivity.³ As a gendered phenomenon, DV disproportionately victimizes women,² but affects people of all ages, socio-economic statuses and cultural backgrounds.^9,10

Risk factors for DV are well documented in literature and include multiple and complex factors at individual, relationship, neighbourhood and community levels.^11,12 The association of individual-level characteristics, such as demographic (e.g. age, gender, socio-economic status), psychological (e.g. personality disorders, depression) and behavioural factors (e.g. drug use) with DV has also been examined.¹¹ Studies indicate that lower income levels are more likely correlated with DV.³ In a systematic review, stress (e.g. financial, work or parenting-related) was found to be a relevant factor associated with increased likelihood of DV.¹¹ Moreover, despite the disparate findings, some studies have found that alcohol and drug use are related to DV victimization. Substance use by the perpetrator or the victim has been referred in about 92% of reported episodes of DV.¹³

These risk factors for DV were particularly exacerbated during the coronavirus disease 2019 (COVID-19) pandemic. Recent research describes adverse economic impacts of the pandemic (e.g. unemployment, income loss), which along with the uncertainty related to the pandemic crisis, have driven a stressful environment for many households.¹⁴ Also, confinement orders, intended to protect the public and prevent widespread infection, left many DV victims trapped with their abusers.⁹ In addition, with personal movement limited and people confined to their homes, ideal conditions were created to fuel DV: social isolation of victims reflected on growing difficulties to access community services, increased control by formal confinement with related lack of opportunities to report occurrences and escaping from violence and potential impunity of the aggressor.¹⁵ Indeed, reports of an increase in DV during the pandemic have been emerging worldwide.^16,17 Accordingly, in a systematic review and meta-analysis of the effect of COVID-19-related restrictions on reported incidents of DV based on 18 empirical studies, it was found that most study estimates were indicative of an increase in DV post lockdowns.¹⁴

In Portugal, DV was set up as an autonomous crime punishable by criminal Law in 2007 and as a public crime in 2009. Despite the strong commitment from the Portuguese authorities to prevent this crime, in the last years DV has been the most reported crime in Portugal.^18,19 In 2020, despite a decrease of about 6% compared with 2019, a total of 27.637 DV occurrences were reported. About 75% of the victims were female and 81.4% of those reported a male aggressor. Close to three quarters of the victims were aged 25 years or over (74%), as well as 93.1% of the suspected aggressors. In almost half of the occurrences (48.6%) the victim was the spouse/partner, in 15% was the ex-spouse/partner, in 15.6% the victim was the son/daughter/stepson/stepdaughter and in 5.9% was the father/mother/stepmother/stepfather.²⁰ Most recently, data published by the Portuguese Association for Victim Support (APAV) revealed that during the first confinement decreed by the Government—from 22 March to 3 May 2020—86% of the reported cases to this NGO were of DV. A large percentage of the victims are women (83%), the age group with the highest prevalence of victimization is between 21 and 44 years old, and a substantial part of the victims did not carry out a professional activity.²¹

Monitoring the DV during the COVID-19 pandemic and documenting the impact of the pandemic on DV have been challenging due to the limited access to social service agencies by victims and strong difficulties of collecting self-report victimization data during the pandemic. As a result, this could end up in a rise in the under-reporting of these crimes.²² A better understanding of the factors and contexts underpinning DV can help in design and executing the appropriate restorative and preventing actions, especially aimed at groups at higher risk. In this study we examine DV during the COVID-19 pandemic and its links with perceived stress, substance use and socio-demographic factors.

Materials and methods

In the very early stage of the COVID-19 pandemic, the NOVA National School of Public Health in collaboration with Ghent University launched the research project ‘Violence in Intimate Relationships in Times of COVID-19: Gender Inequalities and (New) Contours of Domestic Violence?’. As such, a cross-sectional study was carried out in Portugal consisting of an online survey conducted between April and October 2020. Detailed methods and procedures are described elsewhere.²³ Briefly, the survey was disseminated through social media, partner networks and community institutions working within the scope of violence. Only respondents who gave informed consent, who were >15 years old and living in Portugal were able to participate in the survey.

Data were collected using a structured, closed-ended questionnaire, which was developed by Ghent University, based on the UN-MENAMAIS Study questionnaire,²⁴ a set of validated instruments and mental health and coping scales.^25,26,27 Minor adaptations were made to the Portuguese context related to language and response options to some questions (e.g. level of education). The questionnaire was pretested and then administered through the Qualtrics software program.

The dependent variable was experienced DV during the COVID-19 pandemic, consisting of all acts of psychological, physical or sexual violence that occurred within the domestic context, irrespective of biological or legal family ties, in the past 4 weeks. Specifically, participants were asked whether situations of physical, sexual and psychological violence had occurred, with response options being ‘Yes’ and ‘No’. A dichotomic variable ‘Experienced DV during the COVID-19 pandemic’ was created, grouping the affirmative responses to any situation of violence into ‘Yes’ and the negative response to all situations of violence into ‘No’. The independent variables were socio-demographic characteristics (gender, age, education), perceived ability to make ends meet with the household income compared with the period before the COVID-19 pandemic, substance use in the past 4 weeks (consumption of alcohol in a day and medications to sleep or calm down) and frequency of feeling nervous and stressed in the same period.

Data analysis

Descriptive analysis was performed to analyse independent variables and DV outcomes. Bivariable analysis associations between independent variables and reported DV were assessed using χ² and Fisher’s exact tests when appropriate. Multivariable logistic regression with backward selection was done to estimate adjusted odds ratios (OR) and 95% CIs of factors associated with reported DV. A significance level <0.05 was used. Data analysis was performed using IBM-SPSS Statistics version 26.

Results

In total, 1062 participants responded to the online survey. The sample consisted of 77.8% women, 48.0% were between 25 and 44 years old and 36.0% between 45 and 64 years, and 82.0% had attained a higher education (Table 1). In terms of socio-economic conditions, 55.9% perceived it was difficult to make ends meet with the household income during the pandemic.

Of the participants, 11.6% (n = 123) reported consuming three or more alcoholic drinks a day, and 33.3% (n = 288) said they took medication to sleep or calm down during the past 4 weeks.

Regarding the frequency of feeling nervous and stressed, 42.1% (n = 353) of the participants declared feeling this way often or very often in the past 4 weeks, and 36.0% (n = 302) sometimes.

Overall, 13.7% (n = 146) of the participants reported having experienced DV during the COVID-19 pandemic. Particularly, 13.0% (n = 138) reported experience of psychological violence, 1.0% (n = 11) sexual violence and 0.9% (n = 10) physical violence (Table 2).

Further analysis revealed that 55.8% (n = 593) of the participants had never experienced violence in their lifetime, 30.4% (n = 323) experienced violence only before the pandemic, 9.1% (n = 97) before and during the pandemic and 4.6% (n = 49) only during the pandemic (Table 3). A third of the participants who reported victimization in times of COVID-19 pandemic (n = 146) experienced DV for the first time during this period (33.6%, n = 49). Despite the low numbers, the proportion of new c of DV was higher among women (34.2%, n = 40 versus 31.0%, n = 9 among men), participants with higher education (33.9%, n = 38 versus 32.4%, n = 11 up to secondary education) and those who perceived being easy or easier to make ends meet (39.6%, n = 19 versus 30.6%, n = 30 among those who perceived to be difficult).

In the bivariable analysis (Table 4), it was found that there was a trend of lower age associated with more DV being reported. Specifically, in the group under 25 years of age, the prevalence of DV was higher (21.9%, n = 23). The older the participants, the less the reported DV. Also, the participants who perceived difficulties to make ends meet with the household income during the pandemic reported more DV (17.7%, n = 98). Regarding substance use, there was also a trend towards higher alcohol consumption associated with more reported cases of DV (21.1%, n = 26). A higher proportion of participants who use medication to sleep or calm down also reported DV (24.0%, n = 69). In addition, those who reported feeling nervous and stressed more often also reported the highest proportion of DV (26.6%, n = 94). The associations of sex and education with DV were not statistically significant.

All variables considered in the bivariable analysis were included in the multivariable regression model. Logistic regression analyses (Table 4) showed that reported DV was still more likely in those who reported difficulties to make ends meet with the household income (OR = 1.608; P = 0.019), those who used medication to sleep or calm down (OR = 1.851; P = 0.002) and those who reported frequent feelings of nervousness and stress during COVID-19 pandemic (OR = 2.443; P = 0.003). Sex (OR = 0.846; P = 0.491), age (OR = 0.685; P = 0.199/OR = 0.563; P = 0.067/OR = 0.410; P = 0.145) and alcohol consumption are not significant predictors of DV.

Discussion

Main findings of the study

Our study found that ~1 in 10 people experienced DV during the pandemic in Portugal, with psychological violence being predominant, followed by sexual and physical violence. Also, this study showed that around a third of the reported cases are first-time DV victims.

Multivariable analyses showed that perceived financial difficulties, perceived stress and use of medications to sleep or calm down were significantly associated with DV exposure during COVID-19 pandemic. Specifically, the findings indicate that perceived financial difficulties triggered by economic crisis during the pandemic are associated with greater vulnerability to different types of violence.^28,29 The data showed that those experiencing DV reported greater perceived stress. In addition, the findings indicate that the use of sedative was a factor that increased the probability of DV, which could be related to other factors in turn, such as psychological stress, financial difficulties and social isolation, in the context of a violent relationship.

A significant association was found between alcohol consumption and DV, even being out of the range of DV predictors. Individuals with higher alcohol consumption present an increased risk of being involved in DV situations. This could be a hint at the emergence of a subgroup of drinkers at risk of establishing potentially dangerous patterns of alcohol consumption during lockdown.³⁰

This research has confirmed that DV is a problem that mostly affects women. Nevertheless, the difference between women and men on reported DV was found to be not statistically significant, which was not expected in this study, given the extensive literature that describes this phenomenon as a gender-based crime where women are mostly victims.^9,21 The lack of association of lower educational level with experienced DV was also observed and may be due to the fact that the sample was constituted by 82.1% of people with a higher education level. The findings indicate that DV during the COVID-19 pandemic is transversal across different age groups. However, a higher prevalence was found in young adults, which is in individuals between 16 and 25 years of age (21.9%, n = 23). Thus, other important data were the lower reports of elder victims, which may be related to the fact that these behaviours are not understood as violent and abusive.³¹

What is already known on this topic

To our knowledge, there are few studies on DV during the COVID-19 pandemic in the Portuguese population, as in other countries, which make more difficult to place this dataset in context. However, we can say that the proportion of new DV cases found is consistent with the results of the Portuguese Association for Victim Support study on DV during confinement, which found that 34% of female victims stated they had no history of victimization.²¹ It is difficult to ascertain whether this represents an increase in new victims in the country, because data on prior victimization were often not yet collected. In any case, the emergence of new cases could be related to the consequences of social distancing measures and confinement orders.²²

Previous studies have shown that increasing unemployment and loss of income, in addition to isolation during the confinement of lockdown, have been drivers of DV.^32,9 A systematic review based on 11 papers (most from research in Asian countries) and six reports from international organizations regarding psychosocial consequences of the pandemic reported high prevalence of stress and anxiety, particularly related to the period of isolation and financial instability.³³ Thus, isolation and the socio-economic impact could also explain results showing association between perceived stress during COVID-19 pandemic and DV. In addition, the use of sedative could have adverse consequences for mental health when performed without the guidance of a professional. This requires special importance since Portugal has the highest prevalence rate of mental disorders among the European countries.³⁴

In terms of alcohol consumption, our findings are in line with evidence, suggesting that alcohol facilitates the occurrence and severity of DV.³⁵ An increase in alcohol consumption at home, in frequency and quantity, has been documented as a result of isolation.³⁶ In any case, it is known that frequent drinking can create an unhappy, stressful partnership that increases the risk of conflict and violence.³⁷ Also, our findings are consistent with studies showing that women^9,21 and those of younger age are at higher risk of DV victimization.^11,38

What this study adds

To our knowledge, this is one of the largest community-based surveys conducted so far on DV during the COVID-19 pandemic in Portugal. This study suggests that, apparently, confining people in their homes to protect them from spreading the COVID-19 virus may have increased the risk of experiencing DV. This constitutes a challenge for policymakers, as it seems to suggest that public health strategies to combat COVID-19 may have the unintentional consequence of putting people at greater risk of experiencing DV and associated mental and physical health problems.^39,40

Limitations of this study

The limitations of this study must be acknowledged. Due to the sampling procedure, results may not reflect the situation of the population in general. Volunteer bias cannot be excluded. Also, since it was an online survey, it might have included particularly those with digital literacy. This could explain the lower participation of older groups as well as the subrepresentation of individuals with lower educational qualification. Social desirability bias potentially led to under-reporting of experiences of violence. Nevertheless, the anonymous online nature of this survey might have contributed to minimizing those effects. Overall, the burden of DV could be under-represented. Given the relatively low numbers, which hindered further complex analyses, results must be treated with caution. Although the study design led to the inclusion of mainly highly educated groups, this can be considered a strength of the study as these groups are not frequently included in research on DV.^11,41

Conclusions

In conclusion, this study reaffirms that DV is a phenomenon that is present in many lives, being psychological violence the most reported during COVID-19 pandemic. A higher report of DV victimization was found among women and young people. Perceived financial difficulties, perceived stress and use of sedative were factors significantly associated with DV exposure during COVID-19 pandemic. This study shows that there is a need for increased attention to these factors and investment in support systems for victims of DV to be applied to pandemic contexts. It provides insights that can contribute to inform public policies regarding management mechanisms of DV within the national strategies to combat COVID-19, and to design effective responses for DV prevention and combat, especially aimed at groups socio-economically vulnerable, during and after the pandemic. Especially, responses targeting those potentially underserved and with limited digital literacy should be strengthened.

Statement of ethics

The study was approved by the Ethics Commission of the NOVA National School of Public Health (ref. CE/ENSP/CREE/1/2020). The research conforms with the guidelines for human studies and was conducted ethically in accordance with the World Medical Association Declaration of Helsinki. Anonymity of participants and confidentiality of data were guaranteed. Informed consent was obtained from all participants.

Acknowledgements

We thank all the participants who answered the questionnaire for their contribution to this study. We extend our thanks to Fundação Ciência e Tecnologia for funding GENDER RESEARCH for COVID-19. We also thank Dr. Sabine De Moor and Prof. C. Vandeviver (Ghent University) for their assistance with the Portuguese version of the questionnaire in Qualtrics. This project had the following partners: NOVA National School of Public Health (ENSP-NOVA), CICS.NOVA—Interdisciplinary Centre of Social Sciences from NOVA School of Social Sciences and Humanities (NOVA FCSH), University Institute of Maia (ISMAI), and School of Criminology, Faculty of Law, University of Porto (FDUP).

Conflict of interest statement

The authors have no conflicts of interest to declare.

Funding sources

The present publication was funded by Fundação Ciência e Tecnologia, IP national support through CHRC (UIDP/04923/2020).

Data availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Yilian M. Pérez,

Ana Gama,

Ana R. Pedro,

Maria J. L. Carvalho,

Ana E. Guerreiro,

Vera Duarte,

Jorge Quintas,

Pedro Aguiar,

Ines Keygnaert,

Sónia Dias,

References