Building a story: Word patterns in writing and quality of life in women with advanced breast cancer

Women with metastatic breast cancer experience diminished quality of life (QOL) as a consequence of disease, treatment or psychosocial issues. The role of the nurse includes providing interventions to promote QOL. QOL is a multidimensional concept, including affective and cognitive aspects, that has the potential to improve through use of effective psychosocial and behavioral interventions. Expressive writing (EW) is a psycho-behavioral intervention developed by Pennebaker to promote disclosure: narrative formation resulting in catharsis, reframing and integration. EW has been related to positive health outcomes in healthy and ill samples. Gaps in the literature were identified in exploring both use of EW and cognitive processes of disclosure in cancer patients. The purpose of this secondary analysis was to explore the relationship between patterns of language used in EW texts and QOL in women with metastatic breast cancer. Usage patterns of affect, causal and insight words in EW texts from a sample (n = 68) who completed baseline and three month QOL assessments were identified through use of the Linguistic Inquiry and Word Count (LIWC). Relationships between these patterns of language and QOL were explored. QOL was measured using the Functional Assessment of Cancer Therapy - Breast (FACT-B), assessing overall QOL status and subscales: physical well-being, social/family well-being, emotional well-being and functional well-being. The correlation between word pattern use and QOL was investigated using general linear regression. No relationship was observed between positive affect language and QOL. A statistically significant relationship was demonstrated between positive affect language use and emotional well-being subscale scores (β = 1.87, p = 0.02). No relationships were observed with QOL scores and progressive use of causal or insight language. Ten complete EW texts were manually scored to validate LIWC data. Evaluation of text containing grammatical errors, idiomatic language, or complex expression of ideas was inadequate with LIWC. A significant difference between LIWC and manual word counts for negative language was identified. Although no significant difference was noted in word counts for causal or insight language, differences in contextual meaning were noted. Further research including the use of qualitative methods to examine content and contextual use of language is warranted