Oral Presentation COSA 2015 ASM

Does stress increase risk of cancer? A 15 year prospective study of women at high risk of breast cancer (#46)

Phyllis N Butow 1 , Melanie Price 1 , Joseph Coll 2 , Kelly Phillips , Kathy Tucker 3 , Bettina Meiser 4 , Margaret Charles 5 , Chris Tennant 6 , the kConFab psychosocial and Clinical Follow-Up groups on behalf of the kConFab investigators kconfab
  1. CeMPED/ PoCoG, The University of Sydney, NSW, Australia
  2. PoCoG, Sydney, NSW, Australia
  3. Cancer Genetics, ancer Genetics South Eastern Sydney Area Health District Prince of Wales (and St George Hospitals), Sydney, NSW, Australia
  4. Psychosocial Research Group,, Prince of Wales Clinical School, University of NSW, Sydney, NSW, Australia
  5. School of Psychology, University of Sydney, Sydney, NSW, Australia
  6. Dept of Psychiatry, Faculty of Medicine, University of Sydney, Sydney, NSW, Australia

Aims: Stress is often cited by lay people as a risk factor for breast cancer but most studies to date have been methodologically flawed. This truly prospective study was specifically designed to overcome these limitations.

Methods: Adult, unaffected women from an Australian registry of high-risk breast cancer families (kConFab) completed the gold-standard interview-based Life Event and Difficulties Schedule (LEDS) and patient reported outcomes (PROs) assessing anxiety, depression (HADS), optimism (LOT), anti-emotionality (AES) and social support (Duke) three-yearly for up to 12 years.  At 15 years, breast cancer status was determined.

A Cox proportional hazards regression with the time scale anchored to age at interview was performed to examine the impact of acute and chronic stressors on risk of breast diagnosis in the following three years.  The analysis  adjusted for fourteen potential confounders including mutation status and family history and included a random family effect. Total number, total by severity, and at least one severe, acute event (AE) or chronic stressor (CS), were evaluated.

Results:  Data were available for 2741 women from one or more rounds. 103 women (3.75%) were diagnosed with breast cancer. There was no significant mean difference on any PRO between women diagnosed and not diagnosed. Univariately, the only stressor variable associated with time to diagnosis was having at least one severe CS; interactions between stressor variables and PROs were non-significant. In the fully adjusted analysis, which included having at least one severe CS and at least one severe AE, women with at least one severe CS were 1.7 times more likely to have a breast cancer diagnosis than women without a severe CS (HR=1.689 (95% CI= 1.016-2.807), p<0.043).

Conclusions: In the most rigorously designed study to date, severe chronic stressors increased the likelihood of a breast cancer diagnosis by 70% in women at high-risk.  Mood, social support and coping did not moderate this impact. Intervening to reduce severe ongoing stressors may reduce adverse physical and psychological health consequences.