Purpose of Study: Until recently, breast cancer (BC) was relatively rare in Mexico; although incidence is less than half of that in the US, it is now the most commonly diagnosed cancer in women. BC is most common in industrialized regions and in the northern parts of Mexico. This study provides validation results for a measure of Westernization, or more specifically, US social, lifestyle and cultural influences among Mexican women with invasive BC. Study participants are a sub-sample of the Ella Binational Breast Cancer Study. The creation and validation of a Westernization instrument is the first step in understanding US influence on lifestyle and reproductive factors that play a key role in BC risk and appear to influence tumor subtype.

Experimental Procedures: A 26-item Westernization survey was administered via interview to 341 Mexican women residing in Obregon and Hermosillo Sonora, as well as Guadalajara, Jalisco between 2009 and 2010. The Westernization instrument was developed through focus group and content expert review. The guiding themes included materialism, individualism, modernity, and reproductive and gender-related norms. The instrument contains three subsections with response formats to reflect various beliefs and behaviors consistent with the themes. The instrument was part of a larger BC risk factor questionnaire, which included questions on demographics, socioeconomic status, as well as all major BC risk factors (e.g., parity, breast feeding, family history, obesity, tobacco).

Data Summary: Among the respondents, the mean age was 54.4 years (+ 12.6) and 64% had less than the equivalent of a high school education. The mean number of full term pregnancies was 3.9 (+ 2.4) and the mean age at first full term birth was 22.7 years (+ 5.1); 86% reported ever breastfeeding. The mean body mass index (kg/m2) was 28.6 (+ 5.5); 2.0% were current smokers, 27.6% were former smokers, and 70.4% were never smokers; 8.9% reported a positive family history of BC. Psychometric analysis proceeded on each subsection of the instrument separately. The first subsection reflected 11 modernity related behavioral items. An exploratory principal components analysis showed two factors had eigen values significantly (p < .01) larger than those expected in simulation based parallel analysis, while the 3rd and further factors had eigen values significantly smaller than those expected in those simulation analysis (p > .01). Consequently a two-factor solution was determined to be robust and guided sub-scale development. The sub-scale reflected consumerism (e.g., eating fast food frequently, shopping to distract from one's feelings), was maximized at .84 when two lesser items were removed. The sub-scale reflected use of electronic devices, α = .67 when two lesser fitting items were removed. As traditional or gendered sex role responses (2nd subsection) were “yes/no” (I have/will do this), an alternative psychometric approach was done. The adjusted reliability coefficient was .79 for these seven items and removing no items raised this coefficient, suggesting all items should be used in this measure. Next, sub-scale scores were derived and associations were examined, total modernity was significantly negatively associated with traditional gender norms and age.

Conclusions: The results show that this new instrument has sound psychometrics in its first two subsections. The instrument may be used as a complimentary measure in sophisticated models of culture which are applied to Mexican and US Mexican immigrants. The next steps in this study will examine the components of Westernization and their relations to known protective and risk factors of BC and tumor sub-types.

Citation Information: Cancer Epidemiol Biomarkers Prev 2011;20(10 Suppl):A94.