Perceived Ambiguity, Fatalism, and Believing Cancer Is More Prevalent Than Heart Disease

Introduction

In most subpopulations, heart disease incidence is higher than cancer incidence,¹ yet laypeople may think the reverse given extensive cancer exposure² such as the “pink ribbon” campaign and celebrity deaths (e.g., Peter Jennings³). People often overestimate their own cancer risk⁴ and thus may also overestimate its relative prevalence. Such misperceptions could promote underuse of preventive interventions for cardiovascular disease and overuse of cancer risk reduction strategies.

Concomitantly, laypeople perceive cancer messages as highly ambiguous (defined in decision science as lacking in reliability, credibility, or adequacy⁵) and fatalistic. In one study,⁶ 72% of U.S. citizens believed there were too many cancer prevention recommendations and 47% agreed that everything causes cancer. These beliefs are associated with greater perceived risk and cancer worry.⁷ We used a nationally representative sample to assess whether such beliefs are related to misperceptions about relative prevalence. We also explored how these misperceptions might be related to demographics and risk-related behavior.

Methods

Data were collected from 3,376 U.S. adults in 2012–2013 via the National Cancer Institute Health Information National Trends Survey (HINTS) and analyzed in 2013. The sample included 51% women, 75% whites, 41% with high school education or less, 58% never smokers, and 75% with own or family cancer history.

Questionnaires were randomly mailed to residential addresses and completed by the adult with the next birthday. Relative prevalence perceptions were assessed with: How much do you agree or disagree with the following statement: In adults, cancer is more common than heart disease (1=strongly agree through 4=strongly disagree). Perceptions of ambiguity and fatalism were assessed on the same scale with How much do you agree or disagree with each of the following statements: It seems like everything causes cancer, There’s not much you can do to lower your chances of getting cancer, and There are so many recommendations about preventing cancer, it’s hard to know which ones to follow. Standard measures assessed sample demographics; personal cancer risk perceptions; cancer experience; health behaviors; cancer information seeking (including frustration when doing so); and numeracy (HINTS downloadable at hints.cancer.gov/docs/HINTS_4_Cycle_2_English.pdf).

Results

To control for bias in variance estimates due to the complex sampling design, jackknife replicate weights were employed. Univariate logistic regressions identified significant predictors of perceived prevalence. Significant univariate predictors (p<0.05) were entered simultaneously in a final multivariable regression model (Table 1).

Table 1.

Results from univariate and final logistic regressions predicting the belief that cancer is more common than heart disease

	Univariate logistic regression		Final logistic regression
Predictora	OR (95% CI OR)	p	OR (95% CI OR)	p
Gender (51% women)	1.09 (0.70, 1.69)	0.694
Age (M=46.64, SD=16.64)	1.00 (0.99, 1.01)	0.886
Less than HS (13%) versus college (29%)	1.82 (1.18, 2.80)	0.007	3.34 (1.61, 6.93)	0.002
HS (28%) versus college (29%)	1.96 (1.46, 2.61)	0.000	1.15 (0.69, 1.91)	0.584
Some college (30%) versus college (29%)	1.29 (0.92, 1.81)	0.134	1.26 (0.75, 2.12)	0.383
White (76%)	1.00 (0.95, 1.06)	0.991
Smoker (19%) versus never smoker (59%)	0.83 (0.55, 1.24)	0.348
Former smoker (23%) versus never smoker (59%)	0.81 (0.62, 1.05)	0.105
Cancer diagnosis (8%)	0.89 (0.69, 1.14)	0.337
Family members with cancer (67%)	0.97 (0.82, 1.16)	0.747
Numeracy (1–4, M=1.10, SD=0.31)	1.07 (0.75, 1.52)	0.716
Lifetime cancer risk (absolute risk) (1–5, M=2.73, SD=0.83)	1.09 (0.96, 1.24)	0.197
Lifetime cancer risk (comparative risk) (1–5, M=2.79, SD=0.88)	1.02 (0.84, 1.23)	0.852
It seems like everything causes cancer (1–4, M=2.27, SD=0.96)	0.80 (0.71, 0.90)	0.000	0.70 (0.58, 0.85)	0.001
There are so many different recommendations about preventing cancer it’s hard to know which ones to follow (1–4, M=2.07, SD=0.87)	0.72 (0.62, 0.84)	0.000	0.78 (0.59, 1.04)	0.091
There’s not much you can do to lower your chances of getting cancer (1–4, M=2.94, SD=0.89)	0.78 (0.68, 0.89)	0.001	0.92 (0.74, 1.15)	0.468
Cancer information seeking, frustration (1–4, M=2.99, SD=0.97)	0.82 (0.67, 1.00)	0.046	0.97 (0.73, 1.31)	0.864
Cancer information seeking, easy to understand (1–4, M=3.00, SD=0.91)	0.72 (0.60, 0.88)	0.002	0.80 (0.59, 1.07)	0.125
	Linear regressions
Outcome	β	t	p	d
Exercise intentions (1–3, M=1.78, SD=0.85)	0.04	2.85	0.006	0.10
Soda consumption intentions (1–3, M=2.03, SD=0.89)	0.05	0.33	0.746	0.01
Vegetable consumption intentions (1–3, M=1.99, SD=0.84)	0.04	0.53	0.596	0.02
Fruit consumption intentions (1–3, M=2.12, SD=0.86)	0.05	0.87	0.390	0.03
Weight management intentions (1–3, M=2.01, SD=1.17)	0.09	1.59	0.716	0.05
Sunscreen use (1–5, M=3.45, SD=1.41)	0.09	0.37	0.117	0.01
Mammography (1–6, M=3.29, SD=2.13)	0.04	−0.26	0.792	0.01

^aFor 1–4 scales, 1=strongly agree through 4=strongly disagree; for 1–5 scales, 1=very unlikely through 5=very likely; for exercise, vegetables, and fruit, 1=increase, 2=maintain, and 3=haven’t paid attention; for soda and weight, 1=decrease, 2=maintain, and 3=haven’t paid attention; for sunscreen, 1=never through 5=always; for mammography, 1=1 year ago or less, 2=more than 1, up to 2 years ago, 3=more than 2, up to 3 years ago, 4=more than 3, up to 5 years ago, 5=more than 5 years ago, and 6=never had a mammogram.

HS, high school

Approximately 44% agreed or strongly agreed that cancer was more common than heart disease. Higher perceived relative cancer prevalence was associated with lower education level but not age, race, smoking status, numeracy, personal risk perception, or cancer experience. The ambiguity and fatalism measures (and frustration in seeking cancer information) were significantly associated with believing that cancer is more prevalent. In multivariable analyses, the perception that everything causes cancer was the most predictive item. Misperceptions were related to exercise intentions but no other behaviors.

Discussion

Many U.S. citizens, including the highly educated, erroneously believe that cancer is more common than heart disease. Cancer experience and attitudes about personal risk appear unrelated to prevalence perceptions, as does age even though older individuals have more experience with both diseases. Individuals who perceived cancer as more prevalent were more likely to engage in exercise but not other behaviors related to both diseases such as healthy diet and physical activity.

Perceptions of cancer message ambiguity and fatalism may underlie misperceptions of cancer prevalence. Other work shows that people misunderstand the relative prevalence of different cancers, perhaps for similar reasons.⁸ Although more work is needed to establish causality, these findings suggest the need to design cancer messages not only to be less ambiguous but to explicitly present health risks in the context of other health risks—consistent with the finding that people are often more responsive to comparative than absolute information.⁹ Of course, such an approach would need to be tailored for subpopulations with higher cancer risk.

One limitation is that respondents may have focused on mortality rather than morbidity and may have used a narrow definition of heart disease. Although heart disease prevalence is higher across all adult age groups,¹ cancer accounts for more deaths in those aged <85 years.¹⁰ Nevertheless, if the misperception about prevalence is pervasive, further research should consider whether it is consequential and how it might be addressed.