The Effects of Health Belief Components on Women’s Intentions of Early Cancer Screenings

Early screenings for women’s cancer from a cross-cultural perspective

Early cancer screenings have been considered a highly effective way to reduce cancer mortality rates over the years (Choi et al., 2010; Smith et al., 2010). In case of the US, the American Cancer Society has published yearly reports on cancer screening guidelines and issues, hoping to increase people’s awareness of the benefits of regular cancer screenings (Smith et al., 2010). Researchers have reported that cancer mortality rates have dropped because of the remarkable progress in early detection and timely treatment (Juon, Seo & Kim, 2002; Tang et al., 1999). Early cancer screenings among women is particularly important to increase the likelihood of early detection and treatment of women’s cancer. For instance, breast cancer could be cured if it is detected during a medical check-up at an early stage (Aiken et al., 1994; Barg & Grier, 2008; Champion, 1985). Apantaku (2000) reported that early detection by regular mammograms can reduce breast cancer mortality by up to 30 percent.

Prior studies have used a cross-national approach to investigate cultural effects on people’s different attitudes and behaviors regarding cancer prevention (Han & Jo, 2012; Lee, Kim, & Han 2009; Lee et al., 2009; Tang et al., 1999). Indeed, the rates of incidence and mortality from women’s cancer are different across countries, as presented in Table 1 (Jemal et al., 2011; World Health Organization, 2008). For example, about half of breast cancer cases and approximately 60% of the deaths from breast cancer occurred in economically developing countries. However, the incidence rate of breast cancer was higher in Western Europe and North America, compared to Africa and East Asia (Jemal et al., 2011; Kagawa-Singer et al., 2010).¹ In light of this perspective, Kagawa-Singer and colleagues (2010) stated that cultural variables are linked to cancer outcomes through beliefs, attitudes, and behaviors related to prevention and screening. Accordingly, culture should be taken into account as a potential determinant of an individual’s attitudes and behaviors regarding early cancer screenings, and such a direction calls for more empirical research on this issue.

Table 1

Incidence and Mortality Rates of Women’s Cancer: A Cross-National Comparison

Several studies have addressed the role of cultural variables in women’s cancer examinations. They suggest that health professionals responsible for conducting cancer-screening promotion programs should consider variables related to indigenous cultural characteristics (Juon, Seo & Kim, 2002; Tang et al., 1999). Lee, Kim, and Han (2009) reported that women living in the US had early screenings at a higher rate than women with the same ethnic background who were living in a non-US territory. Even among Asian women, significant differences were found in self-examination screening behaviors based on where they lived (e.g., in the US or in their own Asian country) (Tang et al., 1999). These results indicate that cultural differences across countries are likely to influence women’s attitudes and behavioral intentions of early cancer screenings. Consequently, despite culture’s complicated nature, assessments of its influence could be facilitated if an observed factor is clearly distinguished by country, a physical area in which unique cultural aspects are shared among its citizens.

The health belief model and its application to women’s cancer

The health belief model (hereafter HBM) is perhaps the most widely used and highly influential among the social psychological theories that explain the willingness to engage in preventive health behaviors (Champion, 1985; Glanz, Rimer & Lewis, 2002; Tanner-Smith & Brown, 2010). The underlying concept of this model is that health behavior is determined by an individual’s beliefs or perceptions about a particular disease and the means available to decrease its occurrence (Becker, 1974). Since its proposal in the 1950s, the HBM has served as a robust theoretical framework to address the relationship between an individual’s health-related beliefs and behaviors. This model suggests that changes in preventive health behaviors are originally based on four components that reflect an individual’s perceptions of the disease: perceived susceptibility (perceived personal vulnerability to a health condition), perceived severity (perceived personal harm due to a health condition), perceived benefits (perceived positive outcome of adopting a recommended action), and perceived barriers (perceived obstacles to adopting a recommended action) (Rosenstock, 1974). While these original concepts of the HBM were proposed to explain people’s readiness to act, cues to action was later added to help explain what might activate readiness and stimulate overt behavior.

Although the HBM can be applied to preventive health behaviors in a myriad of ways, it has been particularly useful to researchers studying women’s health issues (Tanner-Smith & Brown, 2010). Therefore, the HBM can be naturally employed in the case of women’s cancer, which also has distinctive recommended actions such as early cancer screenings. When applying this model to women’s cancer, perceived susceptibility indicates a woman’s fear of contracting cancer or her sensitivity to the possibility, while perceived severity refers to a woman’s perception of the negative consequences of contracting cancer, such as psychological disorders or financial losses. On the other hand, perceived benefits and perceived barriers can be linked to one’s decision to undergo early cancer screenings (e.g., mammography or Pap smear test). Perceived benefits might include anxiety relief or a lower probability of cancer growth, and perceived barriers might include cumbrance or high financial cost. Finally, cues to action include mass media campaigns, published articles, advice from others, and acquaintances who have contracted cancer.

A great deal of previous studies have applied the HBM to women’s behaviors regarding early cancer screenings (e.g., Aiken et al., 1994; Champion, 1985; Tanner-Smith & Brown, 2010; Pirzadeh & Mazaheri, 2012). One noticeable finding in Tanner-Smith and Brown’s (2010) research, which reviewed studies using the HBM to examine mammographic and/or Pap screenings, revealed strong support for perceived benefits and perceived barriers and weak support for perceived susceptibility and perceived severity in explaining women’s preventive health behaviors. Noroozi, Jomand and Tahmasebi’s (2010) study also supported the direct and indirect effects of the HBM components on breast self-examinations among Iranian women. Meanwhile, Pirzadeh and Mazaheri (2012) found that designing and executing health education programs based on HBM would facilitate women’s willingness to have Pap smear tests for cervical cancer prevention. However, although researchers have often tried to explain women’s health-related cognition and behavior using the HBM, little is known about cross-national differences in the influence of factors associated with health beliefs on women’s perceptions of cancer and their preventive behaviors. Thus, the present study adopted the HBM and employed cross-national comparisons to investigate the following research questions:

RQ 1. How do women’s intentions to take early cancer screenings and their attitudes toward the relevant promotive message differ by culture?

RQ 2. How do women’s perceptions of health belief components regarding women’s cancer differ by culture?

RQ 3. How do the effects of each of the health belief components on women’s intentions to take early cancer screenings differ by culture?

RQ 4. How do the effects of each of the health belief components on women’s attitudes toward the message promoting early cancer screenings differ by culture?

Measures

Critical variables for this study included the four original health belief components-perceived susceptibility, perceived severity, perceived benefits, and perceived barriers-and, as dependent variables, a respondent’s intention to undergo early screening for women’s cancer² and attitudes toward messages promoting early cancer screenings that target women. Their causal relationships were empirically assessed using comparative analyses by country.

All four health belief components were operationally defined based on Rosenstock’s (1974) conceptual definitions and Champion and Scott’s (1997) measurement items on a five-point Likert scale. A series of reliability tests verified that the multiple evaluative items for each component were internally consistent. Detailed measures for the four health belief components and the results of the reliability tests are presented in Table 2. Further, the intention to receive an early cancer screening was assessed on a five-point Likert scale by the following question: “How strong is your intent to undergo early screening for women’s cancer?” All measurement items administered to Japanese and Korean respondents were translated into their respective languages.

The type of communication used as a cue to action, the fifth and external component of HBM, was a public service advertisement encouraging women to seek regular early screenings for cancer prevention. Three versions of this print advertisement were produced based on the primary language (English, Japanese, and Korean) of each of the selected countries. In addition to language, the advertising model (i.e., a Caucasian woman for the US and Asian women for Japan and South Korea) and the governmental institution sponsoring the message were customized to each country. All other creative components, such as copy, layout, and color tone, were equivalent to minimize any uncontrolled effects of intervening variables.

Attitudes toward the message, another dependent variable, were rated using seven evaluative items on a five-point Likert scale. These items, adopted from MacKenzie and Lutz’s (1989) attitude study, include “the message in the above ad is” appealing, truthful, reliable, useful, convincing, important, and relevant to me. These were also proven to be internally consistent through a reliability test (Cronbach’s α = .914).

Table 2

Evaluative Items for Each Variable and Reliability Test Results

Sampling and Data Collection

The current study’s sample consisted of women from three different countries: the US, Japan, and South Korea. North American and East Asian countries have often been chosen for cross-national comparative studies that investigate cultural effects (e.g., Han & Jo, 2012; Lee & Shinkai, 2005; Nomura et al., 2005; Paek, Yu & Bae, 2009). In terms of individualism-collectivism, uncertainty avoidance, long-term orientation, neutrality-affectivity, universalism-particularism, etc., these nations exemplify proper cases of cultural differences (Hofstede, 2001; Trompenaars & Hampden-Turner, 1998).

Respondents were females aged 30-59 because women in this age group are vulnerable to cancer on the whole and thus are typically the primary target of promotions to prevent cancer-related illness and death. Participants in the survey were randomly sampled from a worldwide online research agency’s registered lists of US, Japanese, and Korean consumer panelists. Gender, age, and nationality, according to the panelists’ personal profiles within the online database, were determining factors in sample selection.³ Sample sizes and respondent demographics were roughly equivalent across the three countries. An e-mail was sent to more than 4,000 consumer panelists inviting them to participate in the survey; 944 recipients agreed to complete the online survey. These respondents consisted of 330 US women, 297 Japanese women, and 317 Korean women. Table 3 shows the demographic profiles of the survey respondents. A chi-square test ensured that the respondents were equivalent in the distribution of age across the three nations (x² = 5.903, df = 4, p = ns).

Data Analysis

All research questions were tested using statistical analysis. Since cross-national comparisons were employed for this study, mean values among the three selected countries were compared and verified for any significant differences in respondents’ early screening intentions, message attitudes, and perceptions of health belief components. In the statistical analysis process, it was necessary to control the potential effects of some demographic variables since these confounding variables might influence a respondent’s health beliefs and consequences. Thus, MANCOVA (multivariate analysis of covariance), which treated a respondent’s age, marital status, and educational level as covariate, was used to test research question 1 and 2. For research question 3 and 4, multiple regression was used to cross-nationally assess individual effects of the four health belief components.

Table 3

Demographic Profiles of Survey Respondents

Early Screening Intention and Message Attitude (RQ 1)

Korean women showed the highest level of intention, followed by US women and Japanese women (F = 10.672, p < .001; Table 4). However, a subsequent post-hoc test revealed that this statistical significance of MANCOVA was caused by notable differences between the US and Japan (t = 3.163, p < .01) and between South Korea and Japan (t = 5.329, p < .001), whereas differences between the US and South Korea were insignificant (t = -1.613, p = ns).

Concerning women’s attitudes toward the message promoting early cancer screening, an identical pattern in cross-national differences was found. That is, Korean women showed the most positive attitude toward the message, followed by US women; Japanese women presented the least positive attitude (F = 7.258, p < .01; Table 4). A post-hoc test also revealed significant differences between the US and Japan (t = 3.180; p < .01) and between South Korea and Japan (t = 4.022, p < .001), but not between the US and South Korea (t = -.483, p = ns). For all three countries, however, results suggest that that both early screening intention and message attitude are relatively positive, as the mean values for the three countries were above 3.5 out of 5.

Table 4

Comparisons of Early Screening Intention and Message Attitude by Country

Levels of Health Belief Components (RQ 2)

The four original components of the HBM were individually assessed to verify any cross-national differences. First, for both perceived susceptibility and perceived severity, the two Asian countries were significantly higher than the US (South Korea was the highest, followed by Japan) (Table 5). On the other hand, the results for perceived benefits and perceived barriers revealed fairly different patterns. The benefits of early cancer screenings were perceived to be highest by the Korean respondents, followed by the US and then Japan, while barriers to early examinations were perceived to be the strongest by Japan, followed by South Korea and then the US. These results indicate that Japanese women are least likely to undergo early cancer screenings, as they ranked lowest in perceived benefits (a positive indicator) and highest in perceived barriers (a negative indicator) among the three nationalities.

Table 5

Comparisons of the Levels of Health Belief Components by Country

Effects of Health Belief Components (RQ 3 and RQ 4)

Statistical analysis results uncovered a very similar pattern across the nations. For all three countries, perceived susceptibility, perceived benefits, and perceived barriers had significant impacts on the intention to receive early cancer screenings; only the effect of perceived severity was not significant (Table 6). The predicted negative impact of perceived barriers was also verified. Regarding the effect strengths of individual components, cross-national differences were found. For US and Korean women, the impact of perceived susceptibility was weaker than that of perceived benefits and perceived barriers, whereas the effect of perceived susceptibility was strongest among Japanese women.

With regard to the effects of health belief components on attitude toward the message promoting early cancer screenings, the results did not much differ from those of the previous analysis. In general, across all three countries, only the impact of perceived severity was not statistically significant, whereas the other three components-perceived susceptibility, perceived benefits, and perceived barriers-had significant effects on message attitude (Table 7). The only exception was the insignificant effect of perceived barriers among Japanese women. As to the effect strengths, dissimilar patterns were found between the Asian countries and the US. Specifically, for Japanese and Korean women, the relative impacts of perceived susceptibility and perceived benefits were stronger than that of perceived barriers; on the other hand, for US women, the effects of perceived susceptibility and perceived barriers were much stronger than that of perceived benefits.

Table 6

Effects of Health Belief Components on Early Screening Intention

Table 7

Effects of Health Belief Components on Message Attitude

General Conclusions

Although the HBM is the most commonly utilized theory in health education and health promotion by far, little effort has been made to investigate its application to women’s cancer from a cross-national perspective. This study explored the relative magnitudes of health belief components and their individual impacts on intentions to undergo early screenings for women’s cancer by comparisons among three countries-the US, Japan, and South Korea-where intercultural similarities and dissimilarities coexist. All of these countries have relatively high rates of early cancer screening (American Cancer Society, 2008). A total of 944 females from the three countries participated in the survey, and their responses were statistically analyzed based on the four research questions drawn.

The statistical analysis of the survey data yielded some important findings about the effects of cultural value and identity on women’s cancer-related attitudes and behavioral intentions. Above all, the present study found cross-national discrepancies in several investigative items. For instance, the US and South Korea shared similar tendencies in early screening intention and message attitude, and their levels in these two barometers for early cancer examination were significantly higher than those of the Japanese respondents. On the other hand, for the cognitive components of the HBM excluding perceived barriers (i.e., perceived susceptibility, perceived severity, and perceived benefits), Japanese and Korean participants demonstrated similar levels, which were significantly higher than those of the US participants. The results include more meaningful findings that provide both theoretical and practical insights for effective health promotion, especially in culture-specific contexts.

Theoretical and Practical Implications

Cross-national differences were slight regarding the effects of each health belief component. For all three countries, perceived susceptibility, perceived benefits, and perceived barriers significantly influenced both early screening intentions and attitudes toward the message encouraging early screening; only the effect of perceived severity on the two dependent variables was not statistically significant, with the exception of message attitude for the Japanese respondents. It is difficult to infer why the perception of cancer’s seriousness did not affect these women’s preventive behavior and motivation. If the results were consistent with the prediction, perceived severity would have positively influenced both early screening intention and message attitude. One possible explanation is that the negative consequences of cancer might be so well established and so universally perceived that they simply cannot trigger an attitude change. However, because the true reason cannot be revealed without an in-depth empirical investigation, this finding could be an interesting topic for future research.

Among the three nations investigated, Japan appeared to have the lowest level of early screening intention and message attitude; this result is a red flag for cancer prevention among Japanese women. Not only was their level of perceived benefits lower than the US and South Korea, but their level of perceived barriers was also higher than the other two. Looking at the case of Japan exclusively, the influence of perceived benefits and barriers on women’s adoption of recommended health behaviors is apparent. In contrast to Japanese women, US and Korean women were similar in their levels of early screening intentions and message attitudes, as well as the individual effects of the four health belief components. These results indicate that, despite common predictions based on traditional cultural expectations, Eastern and Western cultures are not necessarily dissimilar in their perceptual and behavioral tendencies; on the other hand, even countries in the same geographical area (e.g., Japan and South Korea in East Asia) can differ considerably in various attitudinal aspects. In fact, these findings are not exceptional, considering a substantially high variation in the predictive powers of traditional cultural dimensions by country or research subjects, as Taras, Kirkman, and Steel’s (2010) meta-analytic review on Hofstede’s cultural framework revealed. Geographical distance may also become a less critical factor in categorizing cultural zones in current society where massive information is quickly transmitted and widely shared beyond national borders via online networks. In consequence, cultural segmentation by geographic area would require greater departmentalization (e.g., by country or other combination of criteria) than in the past.

Results from this study point to several practical implications. First of all, health promoters should consider the effects of health belief components when developing a message strategy to encourage a suggested health behavior, particularly when targeting women for cancer prevention. In many countries, for example, a promotional message that emphasizes the benefits or ease of early cancer examination is likely to be more persuasive to women than a message that underscores the threat of fatal cancers. Before application, however, the findings of this study should not be aggregated; differences between the countries are of paramount importance. The underlying premise is that health information tailored to a specific cultural group is likely to increase the possibility that the target audience will not only understand the information but also follow the recommended behavior. As any campaign for health promotion needs to consider the life experiences and cultural backgrounds of its target audience, cautiously tailored messages that account for cultural values and identity are likely to be more effective.

Study Limitations and Directions for Future Research

The limited scope of the present study suggests directions for future research. First, aside from the health belief components of the HBM, other factors-such as an individual’s medical interest, past experience, information source, and demographic status-could be tested as potential determinants of women’s health behavior and factors influencing communication effectiveness. Second, because of the various criteria required to distinguish one culture from another, further assessments on cultural effects need more specific investigation, accounting not only for country or nationality but also for race, language, immigrational status, health care infrastructure, and so on.⁴ Third, future research could expand the media type used for investigation (the current study only used print media), considering currently available media for health promotion and people’s changing patterns in media usage. Finally, although the study’s participants were randomly sampled from a global online consumer panel, this might be another source of bias since some exogenous variables, such as web accessibility and motivation for survey participation, might be involved in the sampling process. Therefore, researchers’ efforts to enhance data reliability would be required for further studies.

Today, cultural effects are becoming more difficult to predict than ever; thus, empirical investigations of cross-national phenomena are crucial to improving health-related practices and discussions. It is hoped that the current study’s contributions can provide a basis for these research initiatives.