The Relationship between Complementary and Alternative Medicine Use and Breast Cancer Early Detection: A Critical Review

Table 1

Summary of Studies.

Study^a  (Year)	Study design	Sample size	Selected participant characteristics	Data collection methods and CAM assessment	Period of CAM use assessed	Proportion of cancer-free participants using CAM	Breast screening measures^b	Relevant findings	Correlates of more CAM use
DiGianni, (2003 and 2006), [14, 15]	Cross-sectional (2003) Prospective cohort (2006)	104 without cancer history	>18 yrs, F; Enrolled in a breast/ovarian genetic testing clinic; USA	Mailed questionnaire 83% response rate at 1 year Y/N 8 CAMs^c	Ever use	Baseline  42%; 33% used 1-2 CAMs 1 yr followup  58%	Baseline  BSE (rarely/often) 1 year followup  CBE (# in past year) Mam BSE	Baseline—no associations  1yr followup  CBE negatively correlated with # of CAMs used at 1 year (𝑃<0.004); No association for Mam or BSE	Baseline   Perceived cancer risk, sunscreen use, fruit/vegetable consumption 1 yr followup  Higher anxiety, lower perceived breast cancer risk
Downey, (2009), [4]	Cross-sectional	71,083	52–64 yrs, F; enrolled in two washington state insurance companies; USA	Insurance claims data 4 kinds of insurance-paid CAM	Past year	Approximately 22–26% used CAM (depending on year); average 8 visits/yr <1% used only CAM therapies	Mam (past 2 years)	Complementary CAM use more likely to have Mam (OR 1.044; 𝑃=0.031); alternative CAM less likely to have Mam (OR 0.006; 𝑃=0.000); naturopathy negatively associated with Mam (OR 0.736; 𝑃=0.000); massage positively associated with Mam (OR 1.196; 𝑃=0.000)	Younger age, higher disease burden, enrolled in fee-for-service products; over the 3 measurement years; areas with lower education, income, and percentage of minority residents
Druss, (1999), [16]	Cross-sectional	10,675 overall (# answering breast screening items NR^f)	>18 yrs, M/F, age and sex-appropriate subset answered breast questions; national probability sample (medical expenditure panel survey); USA	Interview 77% response rate Y/N 13 CAMs that are practitioner-based	Past year	8.3% of overall sample; NR^f for women answering breast screening items	CBE Mam (past year)	More CBE in CAM users (58.7%–95% CI: 57%–60%) than non-users (69.7%–95% CI; 65%–74%) (𝑃<0.001)^d  No association with Mam	Female, caucasian, higher education, and residing in the west (USA) (only reported for overall sample)
Field, (2009), [17]	Cross-sectional	892	62% 40 yrs+, F; enrolled in the high breast cancer risk cohort; Australia and New Zealand	Mailed questionnaire— 73% response rate Y/N 35 CAMs	Ever use	General Use  55%; 80% >1 CAM therapy, 30% >4 CAM therapies; Intention to prevent cancer  6% of participants	Mam  (past 3 yrs)	No association	More education and physical activity, clinical anxiety, being a former smoker and lower perceived BC risk
Gollsche-wski, (2005), [18]	Cross-sectional	886	48–67 yrs, F; 61%, <55 yrs; random sample south-east Queensland, Australia	Mailed questionnaire— 59% response rate Y/N questions on herbal, phytoestrogen, nutrition and supplement CAMs	Ever use	82% 67% used nutritional approaches, 56% used phytoestrogens, 41% used herbal therapies	CBE, BSE (past 2 yrs)	More BSE in herbal therapy users (OR 1.69, 95% CI 1.34–2.52; 𝑃=0.01); and nutritional users (OR 1.68, 95% CI 1.13–2.50; 𝑃=0.01); no association with CBE	Younger, higher education, middle income, lower smoking, previous hormone therapy, good physical/general health
Gray, (2002), [7]	Cross-sectional	4404	>40 yrs, M/F; stratified sample (by chronic conditions) from health plan; Minnesota, USA	Mailed questionnaire— 86% response rate Y/N 17 CAMs	Past year	42% overall; 46% F	Mam (past yr)	CAM users significantly more likely to have had Mam (67% versus 62%)	Female, younger, higher education, single, employed, health limitations, improved health over past year. More exercise, vegetable intake, fast food consumption; less dietary fat and alcohol (only reported for overall sample)
Mueller, (2008), [19]	Cross-sectional	135 without cancer history, knew BRCA1/2 status	25–56 years of age, F; Enrolled in high genetic breast cancer risk clinic; USA	Telephone interview Y/N 13 CAMs	Past year	78% overall; 69% if spiritual healing/prayer are excluded^e; average 2.3 CAM therapies; 34% ≥3 CAM therapies (overall sample)	Mam (annual) BSE (Monthly)	BSE and CAM use inversely related (OR 0.3, 95% CI 0.1–0.8; 𝑃=0.017); no association with Mam	Older, higher education, ovarian cancer worry
Myers, (2008), [20]	Cross-sectional	2,198, varying risk based on family history	Average 63 yrs, F; family members of women enrolled in breast cancer family study; USA	Mailed questionnaire—70% response rate Y/N 8 CAMs	Ever use	Intention: Preventing Cancer  50%; 42% used 1 CAM, 32% used 2 CAMs, 15% used 3, and 12% used >3 CAM therapies.	BSE, CBE, Mam (ever)	In the univariate analysis, all 3 breast behaviors were associated with CAM use (OR 1.33, 95% CI 1.15–1.54; 𝑃=0.0002); In the multivariate analysis, associations did not remain significant.	Higher education, general health behaviors, optimism (multivariate analyses)
Robinson, (2002), [21]	Cross-sectional	1,593	>18 yrs, M/F, attendees at health fair USA	Questionnaire Y/N 8 CAMs, 13 herbs	Past 2 years	68%; 63% used herbs/supplements	CBE Mam (Past 2 years)	No association	Younger, female, higher education (high school completion), lower levels of health insurance (only reported for overall sample)

^aStudies listed by first author.
^bCBE: clinical breast examination; Mam: mammography; BSE: breast self-examination.
^cY/N refers to dichotomous responses to use of each CAM treatment.
^dThese data reflect the authors’ abstract, data section, and conclusions; the table in the paper presents opposite numbers and is assumed to be a typesetting error.
^eParticipants with cancer were included in this calculation because the authors state that overall patterns of the CAM therapies used didn’t differ between cancer survivors and women without cancer and data were not presented separately for each group.
^fNR: No response.