Psychological distress in women at increased risk for breast cancer: the role of risk perception

Abstract

The magnetic resonance imaging screening (MRISC) study evaluates a surveillance programme for women with a hereditary risk for breast cancer. The psychological burden of surveillance in these women may depend on inaccurate risk perceptions. To examine differences in risk perceptions between three predefined risk categories and associations with psychological distress. BC-specific distress, general distress, and RP (cognitive and affective) were assessed, two months before a surveillance appointment. Cumulative lifetime risk (CLTR) of developing breast cancer was trichotomised into: (1) CLTR of 60–85% (mutation carriers), (2) CLTR of 30–50%, and (3) CLTR of 15–30%. In a total group of 351 women (mean age 40.5 years, range 21–63 years) the three risk categories significantly differed in their accuracy of assessing cognitive risk perceptions. In category 1, 60% had an accurate risk perceptions, in category 2, 43.7% and in category 3, 33.3%. Overestimators reported significantly more breast cancer-specific distress. After adding affective risk perception to the model, this effect disappeared. Affective risk perceptions showed significant associations with breast cancer-specific and general distress. Affective risk perception is a more important determinant for psychological distress than cognitive risk perception. This knowledge should be used during surveillance appointments in order to improve and individualise support for these women.

Introduction

One in every ten women in Western industrialised countries will develop breast cancer during her life-time. A genetic predisposition is suspected in approximately 5–10% of all breast cancer cases. In the mid-1990s, the BRCA1 and BRCA2 genes were identified [1], [2]. Carriers of a mutation in one of these genes have a significantly increased cumulative lifetime risk (CLTR) of developing breast cancer that has been reported to be between 60% and 85% [3], [4], [5], while Antoniou and colleagues [6] recently provided evidence for lower risk percentages in breast cancer patients unselected for family history. Other breast cancer susceptibility genes may also play a role, such as CHEK2 [7], but either their role still has to be elucidated or such genes are not yet identified. Women from families with a clear family history of breast cancer where a mutation has not (yet) been found are also at increased risk, which has mainly been estimated using the risk tables developed by Claus and colleagues [8]. Since the Claus model does not account for bilateral breast cancer, the occurrence of breast cancer in multiple family members as well as the occurrence of ovarian cancer, other models are also being developed, and may eventually provide more accurate risk estimations [9]. One of the management options for women at increased risk is regular surveillance mostly by use of an annual mammography and biannual clinical breast examination. A monthly breast self-examination is recommended.

Perceptions of the risk of developing breast cancer in `high-risk' women are frequently found to be inaccurate, but also show a wide variability. In the literature, between 9% and 57% of `high-risk' women are reported to have accurate risk perceptions [10], [11], [12], [13], [14], [15], [16]. This variability can partly be explained by which the risk perception is measured, either before or after counselling. Meiser and colleagues conducted a meta-analytical review in order to obtain an effect size of the impact of genetic counselling on the accuracy of risk perception. They found a significant medium effect size (r=0.56;P<0.01) which demonstrates the efficacy of genetic counselling in improving risk perception [17]. Despite improvements in accurate risk perception after genetic counselling, there are still women who continue to overestimate or underestimate their breast cancer risk [10], [12], [14], [18], [19], [20], [21]. Several reasons for sustained inaccurate risk perceptions can be given. Lacking sufficient numerical skills or overall education levels can cause inaccurate risk perceptions [22]. Processing information about heredity can lead to wrong assumptions about one's risk of developing the disease, for instance on the basis of physical or psychological identification with an affected relative [23]. Personal experience with breast cancer in the family may obstruct the adoption of realistic risk perceptions [24]. The format in which the risk information is given may also influence the accuracy of recall of the risk estimation. Watson and colleagues found that recall of risk is more accurate when risk information is given in Odds Ratios than in other formats [12].

Women with higher risk perceptions often display more psychological distress, both breast cancer-specific and general [11], [12], [15], [16], [20], [25]. Hopwood and colleagues [16] found more cancer worries in overestimators than in women who underestimated or who estimated their risk accurately. Meiser and colleagues [15] found that overestimators had both higher state anxiety, as well as breast cancer anxiety. These data result from studies that addressed the level of knowledge of risk, i.e., the cognitive dimension. Women had to indicate how they think about their own risk by ticking a number. Hopwood suggested that not only the objective risk information may be of importance, but also the way this information is processed by the individual [13]. This led us to hypothesise that, women may give an accurate or inaccurate estimation of their breast cancer risk, the way they feel about this risk may be very much lower or higher. Further, this felt or affective risk perception may have a more powerful association with psychological distress than cognitive risk perception.

In November 1999, the observational magnetic resonance imaging screening (MRISC) study started in the Netherlands evaluating a surveillance program for women at increased risk of breast cancer due to a genetic or familial predisposition (MRISC-part A). The programme consisted of an annual magnetic resonance imaging (MRI) scan and mammography, biannual physical examination and monthly breast self-examination. The participants were classified into one of three risk categories, corresponding to a CLTR of either more than 60%, a CLTR of 30–50%, and a CLTR of 15–30% [26]. A psychological follow-up study started in September 2000 (MRISC-part B). Herein, we describe the association between psychological distress and risk perception in women participating in the MRISC-part B. First, we differentiated between a cognitive and an affective component of risk perception in the three different levels of objective risk status. Next, we determined the association between general and breast cancer-specific distress, and cognitive and affective risk perception. We hypothesised that the women in the different risk categories differed in the perception of their risk; in a way that higher risk perceptions were associated with elevated levels of both types of psychological distress; and that affective risk perception was more prominently associated with psychological distress than cognitive risk perception.