Abstract
Coronavirus disease is one of the major global public health challenges that has affected different groups of society. This study aimed to investigate how young people understand possible behavioral responses to prevent the spread of Covid-19. This descriptive-analytical study was conducted in 2022. The research population in this study was 2410 youths from the Kurdistan Province. The sampling method was simple and random in nature. The data collection tool was a researcher-made questionnaire based on the EPPM model structures. Data analysis was performed using descriptive and analytical statistics in SPSS software version 24. There was a relationship between demographic and background variables with EPPM constructs and protective behavior of Covid-19 (P ≤ 0.05). There was a positive and significant correlation between the constructs of the EPPM, fear, and protective behavior of Covid-19 in young people. The results of multiple linear regression showed that demographic variables, model constructs, and fear explained 53% of protective behaviors against COVID-19, among which the role of perceived efficacy was stronger than that of the other constructs. In the present study, the perceived efficacy dimension in the EPPM model was the main determinant of protective behaviors against Covid-19 in young people. Therefore, to promote protective behaviors, more emphasis can be placed on strategies to improve the mentioned structures; when the perceived efficacy of recommended health behaviors outweighs the perceived threat, the likelihood of compliance with protective health behaviors in the case of Covid-19 increases.
Keywords: Health behavior, COVID-19, Young adult, Models, Psychological
Subject terms: Diseases, Health care
Introduction
Corona virus disease is one of a major global public health challenge1. The World Health Organization declared this potentially fatal infectious disease a global pandemic on March 11, 20202. Covid-19 is transmitted from person to person through respiratory droplets or direct contact. Available findings show that avoiding exposure to risk is the key to preventing infection with Covid-19, for this reason, quarantine, physical, and social distance have become primary strategies to reduce incidence and mortality3. Compliance with hygiene principles, such as proper hand washing, wearing a mask in crowded places, and staying at home, are recommended to control the spread of the disease and break the chain of transmission4. Providing the necessary information about public understanding of Covid-19 is important for health policy makers to implement effective and appropriate strategies to prevent and control the disease. Several factors may affect response behaviors to Covid-195.
Multiple health models show that understanding the risk of Covid-19 is a critical component for any behavior change6. Among these, the developed parallel process model (EPPM) is important7,8. According to the EPPM, when people are exposed to a dangerous situation, they use two cognitive evaluations: one related to the effectiveness of the recommended strategies and the other related to the perceived threat9. Therefore, the EPPM shows that the perception of risk depends on efficiency, defensive response and perceived threat10,11. According to this pattern, people follow one of two paths in response to health message. Risk and fear control processes (Fig. 1). If the perceived threat content and perceived effectiveness are both high in a message, people follow the path of risk control; that is, they react to the threat with sufficient knowledge and use solutions to eliminate the threat. People who receive a message with a high level of threat and a low degree of efficacy go towards the fear control process, and the fear of danger acts as an inhibiting factor in adopting protective behavior8.
Fig. 1.
Extended parallel process model.
Therefore, to influence people’s behavior to follow the health recommendations of Covid-19, it is important to understand how people perceive the epidemic, how they evaluate these risks, and how such evaluations can inform them. Forcing a change in behavior seems important and necessary.
The results of the study by Shirahmadi et al. showed how the degree of perceived threat and perceived effectiveness affected people’s willingness to perform the recommended health behaviors regarding Covid-1912. Cheng Lin et al. also suggest that Developing health care policies and community interventions to improve preventive behaviors against Covid-19, health care managers should consider the main constructs of EPPM, and demographic characteristics (i.e. age and gender)13. The study of Jahangiri et al. entitled “Perception of the risk related to Covid-19 among the general public of Iran using the developed parallel process model”; the results showed a significant difference in the effectiveness and defense responses perceived among different population groups7.
Considering that previous studies have investigated different age groups and the youth age group has not been specifically studied, and considering the socio-cultural differences in different societies and the absence of a study in the target population (the youth of Kurdistan province), this study aims to investigate how people understand the spread of Covid-19 by using EPPM components in possible behavioral responses to help prevent and control the disease.
Methods
This descriptive-analytical study aimed to determine the extent of adoption of protective behavior regarding the covid-19 disease in the Kurdistan province. All methods were performed in accordance with the relevant guidelines and regulations. The study was conducted in compliance with the principles of the Declaration of Helsinki and was approved by the Ethics Committee of Kurdistan University of Medical Sciences (Ethics Code: IR.MUK.REC.1401.015).
Participant
The research community in this study comprised young people (19–29 years old). The sample volume was calculated using the formula N = Zα2P (1 – P)/d2. According to a similar study7 and considering the ratio (p) of 50, with a confidence interval of 95 and taking into account 10%, the non-response rate and 2% accuracy of the sample size was estimated to be 2410 people.
The total sample size (N = 2410) was proportionally distributed across ten cities based on the youth population (aged 19–29) covered by each city’s health system. The allocated number of participants in each city was: Baneh (264), Bijar (125), Sarvabad (79), Saqqez (348), Sanandaj (664), Divandareh (139), Dehgolan (104), Qorveh (199), Kamiyaran (160), and Marivan (328). In each city, the sample was further distributed proportionally among the comprehensive health service centers based on the number of eligible individuals (aged 19–29 years with an active health record) in each center.
Participants were selected from individuals with active health records at the comprehensive health service centers of each city. Within each center, the list of eligible individuals was reviewed based on the inclusion criteria (age 19–29 years and active health record). The required number of participants at each center was determined proportionally to the number of eligible cases. Individuals unwilling to participate or who provided incomplete responses were excluded from the study.
To ensure randomness, participants were selected using a simple random sampling method from the list of eligible individuals available through the health records of each comprehensive health service center. A random number table was used to select participants randomly.
The inclusion criteria were an age of 19–29 years and an active health record. Unwillingness to cooperate in the implementation of the plan and incomplete completion of the questionnaire were the exclusion criteria. The questionnaire was completed by interview method and in person from May 15 to the end of June 2022 in comprehensive health service centers and by the personnel of these centers. We confirm that informed consent was obtained from all human participants prior to their participation in the study.
Instrument
The questionnaire used in this study was developed by the researcher based on the Extended Parallel Process Model (EPPM) and consisted of two main sections: (1) Demographic/background information, and (2) constructs of the EPPM, including behavioral outcomes. Demographic and Background Section: This included age, education level, marital status, occupation, personal and family disease history, COVID-19 vaccination history, and family deaths due to COVID-19. EPPM Constructs and Sample Items: The EPPM-based part of the questionnaire assessed individuals’ understanding and protective behaviors regarding COVID-19 across the following domains: Perceived Susceptibility (8 items), Example: “If I do not wear a mask, I might get infected with COVID-19.” Perceived Severity (10 items), Example: “If I contract COVID-19, my family members will also get sick.” Perceived Response Efficacy (8 items), Example: “Wearing a mask is not useful for preventing COVID-19.” (reverse-coded), Perceived Self-Efficacy (6 items), Example: “I can avoid close contact with infected individuals.” Fear (measured implicitly within severity items such as ‘Thinking about COVID-19 scares me’). Protective Behavior (seven items). Items assessed actual behaviors (e.g., mask-wearing, handwashing, distancing), rated on a 5-point Likert scale ranging from “Never” to “Very often.” Example: “To what extent do you disinfect personal items to prevent COVID-19?” Behavioral Questions Sources: The behavioral section was developed through an extensive literature review and based on guidelines from the WHO and CDC regarding COVID-19 preventive measures, along with validated items adapted from prior studies7,14.
Questionnaire development, adaptation, and translation
The original items were developed in Persian by the research team, using the EPPM theoretical framework. The development involved a multi-step process: literature review of EPPM-based tools and COVID-19-related behaviors, drafting items aligned with the constructs, expert review for content validation, and pretesting with a small group (10 participants) for clarity and face validity. Since the questionnaire was developed in the native language of the participants (Persian), no translation was needed.
Content and face validity
To assess content validity, the draft questionnaire was submitted to a panel of ten experts in health education, epidemiology, and psychology. Based on their feedback, minor modifications were made to improve the item clarity and theoretical relevance. The Content Validity Index (CVI) and Content Validity Ratio (CVR) were calculated. All items met the acceptable thresholds (CVI > 0.79, CVR > 0.62). Face validity was qualitatively assessed through cognitive interviews with 10 individuals from the target population to ensure that the items were comprehensible and culturally appropriate.
Reliability testing
To assess reliability, a pilot study was conducted on 50 individuals selected through convenience sampling from health centers. Internal consistency was evaluated using Cronbach’s alpha: threat construct, α = 0.746; self-efficacy, α = 0.831; response efficacy, α = 0.712; behavior construct, α = 0.871.
Item modification after validation
Following expert and pilot reviews, several items were slightly reworded for better clarity. However, no items were removed, as all met the content validity and reliability criteria.
Data analysis
To describe the characteristics of the study subjects and determine the predictive factors, descriptive and analytical statistics, including mean, standard deviation, chi-square, t-test, correlation coefficient, and regression, were used. The normality of the data was assessed using the Kolmogorov–Smirnov test. Data analysis was performed using SPSS version 24 software at a significance level of less than 0.05. Data were analyzed using descriptive and inferential statistical methods. Descriptive statistics (mean and standard deviation) were used to summarize continuous variables, and chi-square and t-tests were used to compare categorical and continuous variables. Correlation coefficients were calculated to explore the associations between variables.
A two-step linear regression analysis was conducted to determine the predictive factors. First, univariate linear regression was performed for each potential predictor variable. Variables with statistically significant associations (p < 0.05) in the univariate models were entered into a multivariate linear regression model using the backward elimination method to identify the most relevant predictors.
Prior to conducting the regression analysis, we ensured that all essential assumptions were tested and met. The normality of the dependent variables was assessed using the Kolmogorov–Smirnov test. Multicollinearity among predictor variables was evaluated by calculating the Variance Inflation Factor (VIF), and all values were within acceptable limits (VIF < 10). Linearity, homoscedasticity, and independence of residuals were assessed through visual inspection of residual plots and statistical diagnostics. All assumptions were confirmed to be adequately met for the valid application of linear regression.
To determine the risk control process and fear control process of people, based on similar studies15,16, the average score of perceived threat (perceived sensitivity + perceived intensity) was subtracted from the average score of perceived efficiency (self-efficacy + response efficiency). If the result obtained was positive, it indicated the process of controlling the risk of corona disease, and if the result was negative, it indicated the process of controlling the fear of corona disease in the individual.
Results
In Table 1, the demographic and contextual variables of the participants by gender.
Table 1.
Demographic characteristics of the respondents (n = 2410).
| Variable | Men | Women | Total | |
|---|---|---|---|---|
| Frequency (percentage) | Frequency (percentage) | Frequency (percentage) | ||
| Job | Freelance job | 154 (6.4) | 296 (12.3) | 450 (18.7) |
| Government job | 18 (0.7) | 271 (11.3) | 289 (12) | |
| Unemployed | 601 (25) | 0 | 601 (25) | |
| Housekeeper | 0 | 1070 (44.3) | 1070 (44.3) | |
| Education status | Illiterate | 4 (0.17) | 20 (0.83) | 24 (1) |
| Elementary | 36 (1.5) | 143 (5.93) | 179 (7.42) | |
| Guidance | 177 (7.35) | 435 (18.05) | 612 (25.4) | |
| Diploma | 291 (12.07) | 579 (24.02) | 870 (36.09) | |
| Academic | 265 (11) | 460 (19.08) | 725 (30.08) | |
| Marital status | Single | 504 (20.91) | 619 (25.68) | 1123 (46.6) |
| Married | 269 (11.16) | 1018 (42.24) | 1287 (53.4) | |
| Residence | Urban | 349 (14.5) | 1008 (41.8) | 1357 (56.3) |
| Rural | 424 (17.6) | 629 (26.1) | 1053 (43.7) | |
| History of corona infection | Yes | 481 (19.95) | 976 (40.5) | 1457 (60.5) |
| No | 292 (12.1) | 661 (27.4) | 953 (39.5) | |
| History of hospitalization due to corona virus | Yes | 42 (1.74) | 39 (1.61) | 81 (3.4) |
| No | 731 (30.4) | 1596 (66.22) | 2327 (96.6) | |
| History of death in family members due to corona virus | Yes | 42 (1.74) | 79 (3.27) | 121 (5.02) |
| No | 731 (30.4) | 1558 (64.7) | 2289 (94.98) | |
| Corona vaccination history | Not vaccinated | 92 (3.81) | 146 (6.05) | 238 (9.87) |
| One dose | 64 (2.65) | 153 (6.34) | 217 (8.96) | |
| Two doses | 237 (9.83) | 660 (27.38) | 897 (37.21) | |
| Three doses | 381 (15.8) | 677 (28.09) | 1058 (43.9) | |
The results in Table 2 regarding the relationship between demographic and contextual variables with EPPM constructs and COVID-19 protective behavior show that there is a relationship between all demographic and contextual variables and perceived threat (P ≤ 0.05).
Table 2.
Association of demographic and contextual characteristics with EPPM constructs (n = 2410).
| Variable | Perceived threat | Perceived efficacy | Fear | behavior | |
|---|---|---|---|---|---|
| Mean ± standard deviation | Mean ± standard deviation | Mean ± standard deviation | Mean ± standard deviation | ||
| Gender | Women | 14.81 ± 4.28 | 18.47 ± 3.43 | 18.64 ± 5.02 | 23. 90 ± 6.01 |
| Men | 14.44 ± 4.10 | 13.17 ± 3.45 | 17.50 ± 4.75 | 21.13 ± 6.02 | |
| P-value* | 0.044 | < 0.000 | < 0.000 | < 0.000 | |
| Job | Freelance job | 15.77 ± 4.20 | 18.02 ± 3.45 | 19.25 ± 4.97 | 24.00 ± 5.63 |
| Government job | 14.12 ± 4.15 | 18.02 ± 3.40 | 18.36 ± 5.03 | 23.14 ± 6.04 | |
| Unemployed | 13.90 ± 4.05 | 12.95 ± 3.30 | 17.16 ± 4.82 | 21.08 ± 6.27 | |
| Housekeeper | 18.27 ± 4.96 | 19.68 ± 3.46 | 18.18 ± 4. 57 | 25. 17 ± 6.04 | |
| P-value** | < 0.000 | < 0.000 | < 0.000 | < 0.000 | |
| Education status | Illiterate | 11.45 ± 4.16 | 15.90 ± 2.90 | 17.50 ± 5.01 | 18.22 ± 7.32 |
| Elementary | 12.96 ± 4.58 | 17.41 ± 4.06 | 18.49 ± 5.05 | 21.08 ± 7.48 | |
| Guidance | 13.53 ± 4.00 | 17.45 ± 3.00 | 18.63 ± 5.07 | 22.25 ± 6.22 | |
| Diploma | 14.91 ± 4.11 | 18.21 ± 3.34 | 17.96 ± 5.05 | 23.54 ± 5.98 | |
| Academic | 15.91 ± 4.02 | 19.05 ± 4.15 | 18.36 ± 4.68 | 23.65 ± 5.68 | |
| P-value** | < 0.000 | < 0.000 | 0.101 | < 0.000 | |
| Marital status | Single | 14.93 ± 4.09 | 18.30 ± 3.89 | 17.92 ± 4.94 | 23.13 ± 6.02 |
| Married | 14.49 ± 4.33 | 17.90 ± 4.17 | 18.58 ± 4.95 | 22.91 ± 4.26 | |
| P-value* | 0.012 | 0.170 | 0.001 | 0.388 | |
| Residence | Urban | 15.15 ± 4.13 | 18.19 ± 3.97 | 18.29 ± 4.80 | 23.13 ± 5.93 |
| Rural | 14.09 ± 4.29 | 17.97 ± 4.15 | 18.26 ± 5.17 | 22.88 ± 5.43 | |
| P-value* | < 0.000 | 0.452 | 0.910 | 0.344 | |
| History of corona infection | Yes | 13.54 ± 4.08 | 35.71 ± 7.02 | 17.79 ± 5.14 | 22.80 ± 6.45 |
| No | 16.46 ± 3.81 | 36.68 ± 7.05 | 19.03 ± 4.58 | 23.34 ± 5.64 | |
| P-value* | < 0.000 | 0.001 | < 0.000 | 0.035 | |
| History of hospitalization due to corona virus | Yes | 14.61 ± 4.21 | 18.10 ± 4.05 | 18.27 ± 4.93 | 23.00 ± 6.12 |
| No | 17.57 ± 3.43 | 17.81 ± 3.84 | 18.66 ± 5.45 | 24.02 ± 6.05 | |
| P-value* | < 0.000 | 0.738 | 0.518 | 0.164 | |
| History of death in family members due to corona virus | Yes | 14.60 ± 4.20 | 18.10 ± 4.04 | 18.24 ± 4.93 | 23.04 ± 6.11 |
| No | 16.71 ± 4.09 | 16.06 ± 3.01 | 19.25 ± 5.16 | 23.27 ± 6.12 | |
| P-value* | 0.000 < | 0.920 | 0.036 | 0.688 | |
| Corona vaccination history | Not vaccinated | 12.91 ± 3.77 | 16.28 ± 4.35 | 16.87 ± 4.17 | 21.00 ± 5.78 |
| One dose | 14.40 ± 3.98 | 17.59 ± 3.36 | 17.88 ± 5.03 | 22.57 ± 5.85 | |
| Two doses | 15.76 ± 4.07 | 19.27 ± 3.64 | 18.04 ± 5.16 | 24.68 ± 6.00 | |
| Three doses | 12.57 ± 4.67 | 16.61 ± 4.11 | 18.87 ± 4.68 | 19.08 ± 5.58 | |
| P-value** | < 0.000 | < 0.000 | < 0.000 | < 0.000 | |
*T test, ** ANOVA.
In the final model (Model 2), the following variables showed statistically significant positive associations with protective behavior (p < 0.001):
Educational level (B = 0.421, β = 0.066)
Higher education was associated with increased protective behaviors.
Vaccination status (B = 0.417, β = 0.065)
Those who were vaccinated showed higher protective behaviors.
Self-efficacy (B = 0.497, β = 0.306) and response efficacy (B = 0.576, β = 0.424) were the strongest predictors, suggesting that individuals with greater confidence in their ability and the effectiveness of preventive behaviors were more likely to engage in such behaviors.
Fear (B = 0.204, β = 0.166)
Higher fear levels were significantly associated with increased protective behavior.
Conversely,
Sex (B = − 1.185, β = − 0.091)
Males (reference group assumed as females) were significantly associated with lower protective behavior.
History of COVID-19 infection (B = − 0.668, β = − 0.054)
Those who had previously contracted COVID-19 were less likely to engage in protective behavior.
All coefficients mentioned above were statistically significant (p < 0.001), and the directionality was consistent with theoretical expectations. For example, self-efficacy and response efficacy positively predicted behavior, whereas previous infection may have reduced perceived susceptibility, leading to lower behavior scores.
To examine the relationship between the EPPM constructs and protective behaviors against COVID-19 (Table 3), the relationship between variables was examined using the Pearson correlation test, and the results showed a positive and significant correlation between the EPPM constructs, fear, and protective behavior against COVID-19 in young people.
Table 3.
Mean scores of constructs, score range, and Pearson correlation between the variables.
| Constructs | Mean ± SD | score range | Perceived threat | Perceived efficacy | Fear | Behavior |
|---|---|---|---|---|---|---|
| Perceived threat | 14.49 ± 6.23 | 2–25 | 1 | |||
| Perceived efficacy | 18.70 ± 4.04 | 10–50 |
290 0.000 |
1 | ||
| Fear | 18.27 ± 4.96 | 1–35 |
277 0.000 |
699 0.000 |
1 | |
| behavior | 23.01 ± 6.15 | 1–35 |
307 0.000 |
179 0.000 |
298 0.000 |
1 |
*P ≥ 0.05.
**P ≥ 0.001.
Regression analysis was performed by entering the demographic and contextual variables, model constructs, and fear in separate steps (Table 4).
Table 4.
Regression findings for predicting COVID-19 protective behaviors.
| Dependent variable | Predictor variable | B | SE | Beta | T | P | R | R 2 |
|---|---|---|---|---|---|---|---|---|
| Behavior | Gender | − 2.767 | 0.263 | − 0.210 | − 10.315 | 0.000 | 0.210 | 0.044 |
| Education status | 0.875 | 0.129 | 0.137 | 6.763 | 0.000 | 0.137 | 0.019 | |
| Vaccination history | 1.875 | 0.127 | 0.289 | 14.79 | 0.000 | 0.289 | 0.084 | |
| History of corona infection | 0.540 | 0.257 | 0.43 | 2.105 | 0.035 | 0.043 | 0.002 | |
| Perceived threat | 0.401 | 0.028 | 0.277 | 14.137 | 0.000 | 0.277 | 0.077 | |
| Perceived efficacy | 0.855 | 0.022 | 0.630 | 39.714 | 0.000 | 0.630 | 0.396 | |
| Fear | 0.369 | 0.024 | 0.298 | 15.282 | 0.000 | 0.298 | 0.089 |
Overall, demographic variables and the constructs of the model and fear accounted for 53% of the total variance in protective behavior (Table 5). Although perceived threat was significant in the univariate analysis, it was excluded during the backward selection process because of its lack of significance in the multivariate context.
Table 5.
Final regression model of factors predicting COVID-19 protective behavior in youth.
| Model | R | R square | Adjusted R square | Std. error of the estimate | df | F | Sig |
|---|---|---|---|---|---|---|---|
| 1 | 0.730 | 0.533 | 531 | 4.169 | 8 | 5898.6 | 0.000b |
| 2 | 0.730 | 0.533 | 531 | 4.169 | 7 | 377.546 | 0.000c |
aDependent variable: COVID-19 protective behavior.
bPredictor variables include gender, education, vaccination history, history of COVID-19, perceived threat, perceived efficacy, and fear.
cPredictor variables include gender, education, vaccination history, history of COVID-19, perceived efficacy, and fear.
Discussion
The present study aimed to investigate the predictive factors of COVID-19 protective behavior among the youth of Kurdistan Province using the EPPM.
Regarding the relationship between demographic and contextual variables with EPPM constructs and COVID-19 protective behavior, there is a relationship between all demographic and contextual variables with perceived threat, perceived efficacy, fear, and adoption of COVID-19 protective behavior in youth. With increasing education level, perceived threat, perceived efficacy, and adoption of COVID-19 protective behavior in youth increased, which is consistent with other studies17–19. In the present study, the adoption of COVID-19 protective behaviors was more common in women than in men, which is similar to the findings of Motirzadeh et al.16, and Shirahmadi et al.12. Women perceive COVID-19 as a more serious health threat than men and are more likely to agree with and comply with public health measures, such as mask-wearing, social distancing, and isolation20,21. This higher compliance among women is partly attributed to greater health risk perception and their conventional roles as family caregivers, which motivate them to protect themselves and others20. Women demonstrate more knowledge of preventive measures and are more likely to engage in behaviors such as handwashing and wearing masks consistently21. This pattern has been observed across various countries and cultural contexts. Although Lin et al. (2021) reported no gender differences in the adoption of COVID-19 protective behaviors13. The reason for this discrepancy may be cultural differences and differences in the age groups studied. The low level of perceived threat components in young people indicates that this age group is different from older people17. They do not feel that they are at risk of contracting COVID-19 and its adverse effects.
In the present study, getting vaccinated against COVID-19 was one of the factors associated with adopting protective behavior against COVID-19. In this regard, in the study by Vali Aini et al. in 2022, the desire to protect others by getting vaccinated, general beliefs about vaccination, and people’s perception of the seriousness of the disease were among the factors associated with the acceptance of vaccination in society22. The findings of this study indicate that fear arousal in response to a health threat is significantly and positively associated with the adoption of protective behaviors against COVID-19, consistent with previous research identifying fear arousal as a strong predictor of individual health-related behaviors23–25. Serpas et al. reported that fear of COVID-19 fully mediated the relationship between perceived risk and preventive behaviors23. The results of the study showed that among the cognitive variables, perceived threat had the least power to predict protective behavior among young people, and perceived efficacy and fear had the most power to explain protective behaviors. In line with this finding, a study based on EPPM showed that perceived self-efficacy, response efficiency, and behavioral intention were determinants of performing COVID-19 preventive behaviors26. Molavi Vardanjani et al. found that among the EPPM constructs, self-efficacy and response efficacy had the strongest predictive power for self-care behaviors in pregnant women27. Similarly, Bashirian et al. reported that individuals with higher perceived efficacy were more likely to engage in health-promoting behaviors19. These findings, along with results from other related studies28, suggest that the efficacy components of the EPPM have a stronger influence on the adoption of protective health behaviors than do the perceived threat components. This highlights the importance of emphasizing efficacy beliefs in the design of health interventions and messaging, particularly during outbreaks.
Our results show that perceived threat did not significantly predict protective behavior in the final regression model. Although it was initially included in the univariate analysis, its non-significant role in the multivariate model implies that youth may not view COVID-19 as a serious personal threat. In line with this finding, a large-scale study in China found that perceived threat failed to significantly predict protective behaviors based on experiential (emotion-driven, intuitive) assessments, even though it did affect analytical (rational, information-driven) assessments and related actions. Information seeking and behavioral efficacy play critical roles in motivating action29.
A significant relationship was observed between demographic and background variables and the EPPM constructs, as well as protective behavior against COVID-19 (P ≤ 0.05). A positive and significant correlation was found between fear, EPPM constructs, and protective behavior in young people. Multiple linear regression analysis revealed that demographic variables, EPPM constructs, and fear collectively explained 53% of the variance in protective behavior against COVID-19. Among these, perceived efficacy emerged as the strongest predictor of intention. Similar to this finding, the model constructs predicted approximately 50% of protective behavior30. However, in the study by Molavi Vardanjani et al., the EPPM model constructs predicted 44% of the variance in protective behaviors in pregnant women27. In the study by Roberto et al., the EPPM constructs explained 64% of the variance in behavior31. The reason for the difference in the predictive power of the constructs may be the difference in the population group studied (pregnant mothers) or the expected behaviors (maintaining social distancing in Roberto et al. ’s study).Contrary to the lack of predictive power of the perceived threat construct (sensitivity + perceived severity) in the present study, in the study by Qian et al. (2020), the perceived severity construct was the strongest predictor of behavior change32. Given that the time of the study by Qian et al. was at the beginning of the COVID-19 pandemic, the difference with the present study can be considered the difference in the time of the study; it is natural for the perceived threat to be high at the beginning of the pandemic, with the passage of time and greater understanding of the causative agent of the disease and coping strategies, the perceived sensitivity and severity among people decreases.
While our data showed a trend of slightly lower protective behavior among those who received three doses compared to those who received two doses, this was not statistically significant. We interpret this cautiously, suggesting that individuals with full vaccination may feel a greater sense of security, potentially leading to reduced adherence to protective behaviors, a phenomenon known as risk compensation33. Risk compensation, also known as risk homeostasis or the Peltzman effect, is a theory that states that people are more likely to engage in risky behavior when they feel more protected, such as after vaccination34. This warrants further investigations.
In the present study, most of the young people participating in this study, despite the low perceived threat, had high perceived efficacy and were in the process of controlling the risk. In contrast to this finding, most of the young people participating in the study by Babafunke et al. (2022) were in the process of controlling fear17.
The reason for this difference may be due to the high perceived efficacy in the young people participating in our study, which caused that after about two years from the beginning of the epidemic, although young people do not consider themselves at risk of contracting COVID-19, protective behaviors have been institutionalized in them.
Limitations
This study had several limitations. First, the cross-sectional design limits our ability to infer a causal relationship. Second, reliance on self-reported data may introduce bias, particularly in the measurement of protective behaviors and perceived efficacy. Third, while the model explains 53% of the variance in protective behaviors, nearly half of the variance remains unexplained, suggesting the influence of other factors not included in this study. Finally, the sample was limited to young people, which may restrict the generalizability of the findings to other age cohorts.
Conclusion
The findings suggest that the EPPM framework, particularly perceived efficacy, plays a significant role in predicting protective behaviors against COVID-19 among young adults. However, while the model explains a substantial portion of the variance (53%), other unmeasured factors may also contribute to protective behavior. These results highlight the importance of addressing both perceived threat and efficacy in public health messaging to encourage compliance with protective measures during pandemics.
Acknowledgements
This project was financially supported by the Vice Chancellor for Research of Kurdistan University of Medical Sciences (IR.MUK.REC.1401.015), so the Vice Chancellor for Research of the university and all those who participated in this study are sincerely thanked and appreciated.
Author contributions
All authors of this article collaborated throughout the research and writing stages of the article.
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.
Declarations
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Contreras, G. W. Getting ready for the next pandemic COVID-19: Why we need to be more prepared and less scared. J. Emerg. Manag (Weston Mass). 18, 87–89. 10.5055/jem.2020.0461 (2020). [DOI] [PubMed] [Google Scholar]
- 2.World Health Organization. Coronavirus disease 2019 (COVID-19) Situation Report. (2020). https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200427-sitrep-98-covid-1.
- 3.Brooks, S. K. et al. The psychological impact of quarantine and how to reduce it: rapid review of the evidence. Lancet (London England). 395, 912–920. 10.1016/S0140-6736(20)30460-8 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Jernigan, D. B. & Update Public Health Response to the Coronavirus Disease 2019 Outbreak - United States, February 24, 2020. MMWR Morb Mortal. Wkly. Rep.69, 216–219. 10.15585/mmwr.mm6908e1 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pulcini, C. et al. Factors associated with adherence to infectious diseases advice in two intensive care units. J. Antimicrob. Chemother.57, 546–550. 10.1093/jac/dki483 (2006). [DOI] [PubMed] [Google Scholar]
- 6.Napper, L. E., Fisher, D. G. & Reynolds, G. L. Development of the perceived risk of HIV scale. AIDS Behav.16, 1075–1083. 10.1007/s10461-011-0003-2 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Jahangiry, L. et al. Risk perception related to COVID-19 among the Iranian general population: an application of the extended parallel process model. BMC Public. Health. 20, 1571. 10.1186/s12889-020-09681-7 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Popova, L. The extended parallel process model: illuminating the gaps in research. Heal Educ. Behav. Off Publ Soc. Public. Heal Educ.39, 455–473. 10.1177/1090198111418108 (2012). [DOI] [PubMed] [Google Scholar]
- 9.McMahan, S., Witte, K. & Meyer, J. The perception of risk messages regarding electromagnetic fields: extending the extended parallel process model to an unknown risk. Health Commun.10, 247–259. 10.1207/s15327027hc1003_4 (1998). [DOI] [PubMed] [Google Scholar]
- 10.Barati, M. et al. Effectiveness of Educational Intervention on Improving Preventive Behaviors in Fireworks Injuries: Applying the Extended Parallel Process Model. J. Educ. Community Heal. 8, 81–87. 10.52547/jech.8.2.81 (2021). [Google Scholar]
- 11.Lwin, M. O. Learning AIDS in Singapore: examining the effectiveness of HIV/AIDS efficacy messages for adolescents using ICTs. Knowl. Manag E-Learning Int. J.10.34105/j.kmel.2010.02.013 (2010). [Google Scholar]
- 12.Shirahmadi, S. et al. Fear control and danger control amid COVID-19 dental crisis: Application of the Extended Parallel Process Model. PLoS One. 15, e0237490. 10.1371/journal.pone.0237490 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lin, H. C. & Chen, C. C. Disease Prevention Behavior During the COVID-19 Pandemic and the Role of Self-Esteem: An Extended Parallel Process Model. Psychol. Res. Behav. Manag. 14, 123–135. 10.2147/PRBM.S291300 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Khazaee-Pool, M. et al. Development and initial validation of a scale for assessing affecting factors on preventive behaviors of COVID-19 (AFPB-CO): using the protection motivation theory. 10.21203/rs.3.rs-23051/v1 (2020).
- 15.Ghobadi, K., Rakhshanderou, S., Hosseini, F. & Ghaffari, M. COVID-19 protective behaviors of Iranian healthcare workers: a psychological theoretical design analysis. Heal Scope. 11, 8. 10.5812/jhealthscope-121318 (2022). [Google Scholar]
- 16.Motayerzadeh, S., Tahmasebi, R., Darabi, A. & Noroozi, A. Predictor factors of preventive behaviors based on extended parallel process model (EPPM) in residents of Bushehr province in 2021. Iran. J. Heal Educ. Heal Promot. 10, 399–411. 10.52547/ijhehp.10.4.399 (2023). [Google Scholar]
- 17.Babafunke, F. et al. Perceived risk and protective behaviors related to COVID-19 among the Nigerian general population: An application of the extended parallel process model. Sci. Afr.16, e01259. 10.1016/j.sciaf.2022.e01259 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Saqlain, M. et al. Knowledge, attitude, practice and perceived barriers among healthcare professionals regarding COVID-19: A Cross-sectional survey from Pakistan. J. Hosp. Infect.105, 419–423. 10.1016/j.jhin.2020.05.007 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Bashirian, S. et al. COVID-19 Prevention Behaviors among Health Staff: Data from a Large Survey in the West of Iran. J. Res. Health Sci.21, e00509. 10.34172/jrhs.2021.43 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Galasso, V. et al. Gender differences in COVID-19 attitudes and behavior: Panel evidence from eight countries. Proc. Natl. Acad. Sci. U S A. 117, 27285–27291. 10.1073/pnas.2012520117 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tan, J., Yoshida, Y., Ma, K. S. K., Mauvais-Jarvis, F. & Lee, C. C. Gender differences in health protective behaviours and its implications for COVID-19 pandemic in Taiwan: a population-based study. BMC Public. Health. 22, 1900. 10.1186/s12889-022-14288-1 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Valieiny, N., Poorcheraghi, H. & Pashaeypoor, S. Tendency to receive Covid-19 vaccine and its associated factors: An integrated review study. IJNR17, 58–73. 10.22034/IJNR.17.2.58 (2022). [Google Scholar]
- 23.Serpas, D. G. & Ignacio, D. A. COVID-19 fear mediates the relationship between perceived risk and preventive behaviors: the moderating role of perceived effectiveness. Psychol. Health. 37, 1457–1470. 10.1080/08870446.2021.1980566 (2022). [DOI] [PubMed] [Google Scholar]
- 24.Sand, G. & Bristle, J. Motivating Protective Behavior against COVID-19: Fear Versus Hope. J. Aging Health. 36, 350–366. 10.1177/08982643221089427 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Grano, C., Singh Solorzano, C. & Di Pucchio, A. Predictors of protective behaviours during the Italian Covid-19 pandemic: an application of protection motivation theory. Psychol. Health. 37, 1584–1604. 10.1080/08870446.2022.2062355 (2022). [DOI] [PubMed] [Google Scholar]
- 26.Kordi-Kalaki, F., Dehdari, T. & Abolghasemi, J. Determinants of COVID-19 preventive health behaviors in Iranian adults based on extended parallel process model. Sci. Rep.13, 14416. 10.1038/s41598-023-41643-y (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Molavi-Vardanjani, E., Azimi, M. & Fathian-Dastgerdi, Z. Determinants of self-care behavior among pregnant women during COVID-19 pandemic in Isfahan, Iran, According to extended parallel process model. mui-jhsr18, 307–313. 10.48305/jhsr.v18i4.1477 (2023).
- 28.Nazione, S., Perrault, E. & Pace, K. Impact of Information Exposure on Perceived Risk, Efficacy, and Preventative Behaviors at the Beginning of the COVID-19 Pandemic in the United States. Health Commun.36, 23–31. 10.1080/10410236.2020.1847446 (2021). [DOI] [PubMed] [Google Scholar]
- 29.Pang, Z., Zhao, C. & Xue, L. What drives people’s protective behaviors during the early stage of the COVID-19 pandemic in China. Front. Psychol.1310.3389/fpsyg.2022.781279 (2022). [DOI] [PMC free article] [PubMed]
- 30.Noroozi Masir, M., Tarrahi, M. J., Fathian Dastgerdi, Z. & Rahimi, M. Investigating the factors related to protective behaviors against COVID-19 in healthcare workers: Application of extended parallel process model. Heal Sci. Rep.6, e1778. 10.1002/hsr2.1778 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Roberto, A. J., Zhou, X. & Lu, A. H. The Effects of Perceived Threat and Efficacy on College Students’ Social Distancing Behavior during the COVID-19 Pandemic. J. Health Commun.26, 264–271. 10.1080/10810730.2021.1903628 (2021). [DOI] [PubMed] [Google Scholar]
- 32.Qian, M. et al. Anxiety levels, precautionary behaviours and public perceptions during the early phase of the COVID-19 outbreak in China: a population-based cross-sectional survey. BMJ Open.10, e040910. 10.1136/bmjopen-2020-040910 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.McColl, K., Martin-Lapoirie, D., Veltri, G. A., Arwidson, P. & Raude, J. Does vaccination elicit risk compensation? Insights from the COVID-19 pandemic in France. Heal Psychol. Behav. Med.12, 2287663. 10.1080/21642850.2023.2287663 (2024). [Google Scholar]
- 34.Hwang, J., Hwang, S., Kim, H. B., Lee, J. & Lee, J. Risk compensation after COVID-19 vaccination: Evidence from vaccine rollout by exact birth date in South Korea. Health Econ.33, 1811–1830. 10.1002/hec.4837 (2024). [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.