To our knowledge, this is the first quantitative study in the Norwegian context that examined associations between psychological factors and avoidance of breast cancer screening, and intention to attend mammography screening. The main finding was that the factors predicting defensive avoidance of breast cancer screening and intention to attend mammography screening were not the same, indicating that intention to attend screening may not be just the opposite of screening avoidance. The findings showed that defensive avoidance of breast cancer screening was associated with lower perceived susceptibility to breast cancer, lower response efficacy of mammography screening, higher breast cancer fear and checking breasts for lumps, while intention to attend mammography within the next two years was associated with higher response efficacy of mammography screening, having a lower educational level, and regular previous mammography attendance compared to never attending.
Defensive avoidance of breast cancer screening
Following the propositions of the EPPM, we hypothesized that women with higher perceived threat and lower perceived efficacy would report higher defensive avoidance. Our findings, however, showed that women in our sample that perceived themselves as being less susceptible to breast cancer were more likely to hold defensive avoidance beliefs regarding breast cancer screening. While high perceived threat is reported as part of the fear control process within EPPM research [46], the majority of the EPPM studies use intervention designs that involve fear appeal messages and measure effects of fear appeals rather than existing perceptions. Thus, their results may differ from those obtained in cross-sectional studies. A study using cross-sectional design to examine the EPPM constructs got findings similar to ours, reporting the most defensive reactions among participants with low perceptions of threat and efficacy [47]. A potential explanation suggested by the researchers was that these low perceptions were a consequence of defensive reaction to some earlier fear appeal message that was highly threatening but conveyed too little efficacy and thus resulted in fear control response to minimize the perceptions of threat [47]. However, it is possible that our results indicate that women with low susceptibility to breast cancer in our sample think of breast cancer screening as not relevant. Stephenson and Witte [17] emphasized that defensive avoidance may be a hard construct to measure due to its ambiguity. They argued that after the participants of an intervention study were exposed to a fear appeal message, it was difficult to differentiate whether their responses to questions assessing defensive avoidance meant that they engaged in maladaptive coping as a result of a threatening message or simply ignored that message [17]. This corresponds with the previous studies on non-attendance of cervical cancer screening, which reported that women with low susceptibility to cervical cancer and women who felt healthy and had no symptoms were more likely to be non-attenders [48, 49], as were women who considered screening not relevant or not a priority [49].
We found that response efficacy of mammography screening was a significant negative predictor of defensive avoidance, indicating that women that do not consider mammography effective in averting breast cancer threat and making them feel safer may be less motivated to take part in it. This finding is consistent with prior fear appeal studies which showed that fear appeal messages with low efficacy generally led to greater fear control responses [46].
Our results showed that breast cancer fear was positively associated with defensive avoidance of breast cancer screening, meaning that women who worry a lot about getting breast cancer and are afraid to find a lump during screening may engage in defensive avoidance to rationalize why they do not need to attend screening, and thus reduce their anxiety. These findings are in line with previous research on breast cancer fear. Lagerlund, Hedin [8] found that those reporting that mammography would make them worry more about cancer were less likely to attend the screening. Similarly, Rippetoe and Rogers [50] reported that fear was directly related to defensive avoidance of performing breast self-examination.
Breast checks frequency was a significant negative predictor of defensive avoidance, meaning that those women who checked their breasts at least occasionally were less likely to have defensive avoidance beliefs regarding breast cancer screening. While this finding is contrary to Johansson and Berterö [51] and Lagerlund, Widmark [9] who reported that women practicing breast self-examination were more likely to not participate in the screening program, a possible explanation for our finding may be that our questions on defensive avoidance referred generally to breast cancer screening rather than just mammography, so women that check their breasts somewhat regularly may see it as protective against breast cancer.
Perceived severity of cancer treatment, self-efficacy despite lack of time, and mammography frequency were all significantly associated with defensive avoidance in the bivariate analysis, but not in the multivariate analysis. Perceived severity of cancer treatment positively correlated with breast cancer fear, while self-efficacy despite lack of time positively correlated with response efficacy of mammography screening. It is possible that a shared variance between the variables affected their ability to uniquely predict defensive avoidance. Mammography frequency was not a significant predictor of defensive avoidance even when entered into the model separately from breast checks frequency. This finding, therefore, could not be explained by the shared variance and needs to be examined further.
Finally, while it was observed in prior research that fewer immigrant women attend mammography screening in Norway [11], ethnicity was not a significant predictor of defensive avoidance of breast cancer screening in the multiple regression analysis. The bivariate results, however, showed that women with migration background reported higher mean defensive avoidance than did ethnic Norwegian women. Ethnic differences in defensive avoidance should be investigated further. A larger sample of women with migration background could have yielded different results and therefore should be recruited in future studies.
Intention to attend mammography screening within the next two years
We hypothesized that women with higher perceived threat and higher perceived efficacy would report higher intention to attend mammography. In line with this, our findings showed that women who felt more confident that they could attend mammography despite having little time were more likely to express intention to attend the screening within the next two years. Further, those women who considered mammography to be effective in reducing breast cancer threat and making them feel safer were also more likely to report intention to attend screening. These findings are consistent with prior research. Higher response efficacy and higher self-efficacy were consistently found to be associated with recommended response to health threat for a number of health behaviors in the fear appeals research [18, 19, 34, 46]. Further, when it comes to breast cancer screening specifically, self-efficacy and response efficacy were significant predictors of breast screening behaviors in the studies that utilized other social cognition theories [e.g. 23,25,46]. In the Norwegian context, participants in a qualitative study of mammography experiences stated that being busy with their daily life was one of the main reasons to postpone mammography attendance [14]. This further supports our finding that women would be more likely to report intention to attend mammography if they scored higher on self-efficacy to attend mammography despite lack of time. Finally, Solbjør, Skolbekken [15] stated that Norwegian women who regularly attended mammography screening saw it as the only option to protect against breast cancer, thus indicating high response efficacy perceptions.
Contrary to our hypothesis, perceived susceptibility to breast cancer was not associated with intention to attend mammography in neither bivariate nor multivariate analysis. In previous studies, perceived susceptibility was not consistently associated with breast cancer screening adherence, showing significant associations in some studies [23,24,25], but not other [52, 53]. Similarly, in Norway, qualitative studies reported women having generally low perceived susceptibility of breast cancer, which did not change over the years of repeated mammography attendance [14, 15]. This may indicate that perceived susceptibility is of a lesser importance for women’s intention to get a mammogram when mammography screening is publicly available.
Having high school education or less (compared to having some form of higher education) was a significant positive predictor of intention to attend mammography. A number of earlier studies have found that women with higher education were less likely to take part in the national screening programs, with researchers suggesting that women with higher education may have been using private mammography services [e.g. 54,55,56]. However, a study in Denmark reported that women who did not participate in organized mammography screening, did not seek private mammography services either [57]. Jensen, Pedersen [56] further found that in a Danish sample, higher levels of education were associated with non-attendance, possibly as a result of making an informed choice not to attend screening after evaluating pros and cons of mammography screening.
Finally, we found that women, who have reported attending mammography once a year, or once every two years, were more likely to express intention to attend mammography screening compared to those who reported that they never attended mammography. Irregular attendance compared to never attending was not a significant predictor of intention. Moreover, after mammography frequency was added into the regression model, self-efficacy to attend mammography despite lack of time stopped being a significant predictor of intention, indicating stronger influence of regular mammography attendance on intention than self-efficacy. One possible explanation for this is that regular attendance of the screening program creates a habit and, therefore, some other psychological factors may become less salient. Accordingly, Solbjør, Skolbekken [15] reported that those women who consistently attended mammography between 2003 and 2009 perceived screening as a routine procedure that they did not question.
The findings in the current study provide support for the use of EPPM as a theoretical framework for studying breast cancer screening attendance and non-attendance. Our findings suggest that when designing health promotion programs to increase mammography screening attendance it is important to focus not just on predictors of screening attendance but also on predictors of screening avoidance. Our results show that the mechanism of defensive avoidance of breast cancer screening is complex. While a threatening message may increase perceived susceptibility to breast cancer which should lower defensive avoidance response, it may also increase fear which is positively associated with defensive avoidance. EPPM studies with intervention designs show that high threat + high efficacy messages distributed in health promotion campaigns are most effective to move individuals from the fear control to the danger control group [17, 46]. This could be an appropriate strategy in the case of women that are high on defensive avoidance in our sample. Moreover, designing campaign messages that would emphasize the effectiveness of mammography screening may be especially useful, as response efficacy was the only psychological predictor that was significant for both defensive avoidance and intention. Finally, our findings suggest that emphasizing regular use of the mammography screening program, thus making it a routine procedure for women, would be an effective strategy to ensure future consistent attendance.
The strengths of this study include focus on both intention to attend mammography screening and breast cancer screening avoidance. Further, the project recruited participants from different parts of Norway including the capital city and its suburbs, provincial towns and rural areas and had a fairly large sample. Finally, only a few studies have applied the EPPM in breast cancer screening research, thus our findings contribute to a better understanding of how the EPPM constructs predict breast cancer screening-related behaviors. The study also has some limitations. The cross-sectional design of the study does not allow drawing any conclusions on causality and directionality of the relationships or measuring future attendance of the mammography screening. Further, the study used questions taken from previously validated scales, but the scales were not used in their full versions. While construct validity for each sub-scale was assessed and they showed satisfactory reliability, it is possible that some constructs may not have been fully captured. Convenience sampling and recruitment from mammography centers may have affected the reported mammography rates. Furthermore, due to being an addition to the larger project that was not related to mammography screening, the study recruited a younger sample than is typically studied in mammography screening research: almost half of the women in our sample were below the age of 50. The response rate in the study was about 30%; therefore, there may be a self-selection bias. Thus, the findings of the study may not be generalizable to the general population. Finally, although the measures used in this study included some factors that may influence defensive avoidance and attendance of mammography screening, such as pain and discomfort, there are other possible factors that are not included. Factors that may influence screening decisions are availability of information on the pros and cons of mammography screening, recommendations from general practitioners, and communication with health personnel at the mammography centers.
Future research may benefit by investigating contextual factors that may affect screening participation. For example, Solbjør, Skolbekken [15] reported in their qualitative study that trust in the Norwegian healthcare system was an important aspect of mammography attendance.
