Notes Washed Up in a Bottle

Don. W. MacCorquodale M.D. M.S.P.H.

Periodic Notes From the Field on Philosophy and Science.

COMPLEMENTARY MEDICINE

          The Netherlands is considering stricter laws on the practitioners of complimentary medicine after an actress died from breast cancer following treatment by 28 different practitioners of complimentary medicine (News.   Netherlands to crack down on complimentary medicine.   BMJ 2004; 328:   485).   Standard care was available for her, but she elected alternative treatments exclusively.   The Dutch Healthcare Inspectorate wants the law changed to provide greater supervision of alternative practitioners, and it also wants to make it illegal for anyone other than a physician to make a medical diagnosis.

BREAST CANCER RISK AND ANTIBIOTICS

          Velicer and her colleagues conducted a case-control study of breast cancer risk and antibiotics in a sample of slightly more than 10,000 women enrolled in the Group Health Cooperative, GHC (Velicer C M et al. Antibiotic use in relation to the risk of breast cancer.   JAMA 2004; 291:   827-835).   Cases were women older than 19 years, who had been diagnosed with invasive breast cancer, and controls were randomly selected from the GHC enrollment files.   There were slightly more than 2,200 cases and almost 8,000 controls.

          Data on antibiotic use were obtained from the GHC pharmacy records.   Between 97% and 99% of patients enrolled in the GHC get their prescriptions filled in the GHC pharmacy.   The investigators chose two measures of antibiotic use:   the cumulative number of days of antibiotic use and the total number of antibiotic prescriptions per participant. Logistic regression was used to estimate the relative risk of breast cancer in relation to antibiotic use.  

          A larger proportion of cases than controls were educated beyond high school, had ever used oral contraceptives, were more likely to have menarche before the age of 11 years, to have had a first birth after age 30 years, to have a family history of breast cancer, and to have a higher body mass index.   For all antibiotic classes considered together increasing cumulative days of use was associated with increased risk of developing breast cancer, and a dose-response relationship was apparent.   Results were much the same when the cumulative number of prescriptions was used as an indicator of use.   Adjustment for the usual risk factors (except alcohol use and lactation) did not appreciably change the results.   The increased risk was noted for all antibiotic classes.  

          COMMENT:   In an editorial in the same issue of JAMA, Roberta Ness and Jane Cauley pointed out that the frequency of missing data among controls limited the validity of adjusting for the common risk factors.   Mammography rates were quite low, 42%, for controls, which may have resulted in detection bias.   Undetected incidence of breast cancer among the controls would result in under estimating breast cancer risk.

          The editorialists suggested that a biological explanation of the mechanisms attributed to a given class of antibiotics is hard to reconcile with the authors finding that breast cancer risk was elevated across multiple classes of antibiotics.

          Ness and Cauley concluded their editorial with the observation that this study raises more questions than it answers.  

BREAST CANCER ON CAPE COD

          Several studies have found that age-adjusted breast cancer incidence in higher on Cape Cod than statewide levels.   Differences persisted in one case-control study following adjustment of established risk factors.   The authors of this article conducted a case-control study of invasive breast cancer occurring on Cape Cod during the 1988 and 1995 period (McKelvey W et al.   Association between residence on Cape Cod , Massachusetts , and breast cancer,   Ann Epidemiology 2004; 14:   89-94).   Cases, 1121, were women who had been residents on Cape Cod for at least 6 months prior to diagnosis.   Controls, 992, were frequency matched to cases on date of birth in decades and vital status.  

          These women were predominantly white, 98%, and ranged in age from 27 to 103 years.   The average age for cases was 66 years; for controls, 64 years.   Increased risk of breast cancer was associated with a family history of a first-degree relative, nulliparity, age 30 or more years at first birth, and having a high school diploma or a higher level of educational attainment.

          Breast cancer incidence was higher among women who had lived on Cape Cod for 5 or more years in both crude and adjusted analyses.   Adjusting for family history, parity, and age at first birth strengthened the association.   Risk was most increased for woman who had lived on Cape Cod for 25 to less than 30 years compared to women who had lived there less than five years.  

          The authors suggested that residence on Cape Cod might be a surrogate for unidentified environmental risk factors such as pesticides and water contaminated industrial and agricultural use.  

          COMMENT:   I am always uneasy about studies showing an excess of cancer in a given geographic area when the increased incidence is attributed to some unidentified environmental contaminant.   While some environmental agent could be responsible for the excess, such things must occur occasionally at random.   In other cases, the observed excess well might be due to residual confounding from unidentified factors.

CIGARETTE SMOKING         

          In his clinical review of the subject, Why people smoke, Marvin J Jarvis of the University College London wrote that “cigarette smoking is primarily a manifestation of nicotine addiction” (BMJ 2004; 328:   277-279).   Starting to smoke usually takes place in the early teenage years.   For the beginner, smoking a cigarette conveys as message such as “I am no longer my mother's child” or “I am tough.”   Children who start smoking tend to come from backgrounds that favor smoking such as high levels of smoking in parents, siblings, and peers; deprived neighborhoods; and schools where smoking is common.   These children commonly have low self esteem, are overweight, or are poor achievers at school.   “As the force from the psychosocial symbolism subsides, the pharmacological effect takes over to sustain the habit.”

GULF WAR VETERANS AND ANXIETY

          A group of investigators conducted telephone surveys of randomly selected Iowa veterans of the first Gulf War (Black D W et al.   Gulf War veterans with anxiety.   Epidemiology 2004; 14:   135-142).   They used a structured instrument called PPRIME-MD to determine “the prevalence of symptoms suggestive of specific conditions.”   Anxiety conditions, including generalized anxiety disorder, panic disorder, and posttraumatic stress disorder, were more common among Gulf War veterans than those not deployed to the Gulf, OR = 2.1, P5% CI:   1.3-3.1.  

          A number of conditions were more prevalent among Gulf War veterans with anxiety than those veterans who did not meet the criteria for anxiety conditions.   For example, 72% of the veterans with anxiety also had symptoms of major depression while 4% of those who did not meet the criteria for anxiety did so.   Alcohol abuse, symptoms of fibromyalgia and of chronic fatigue were also more common among the veterans with anxiety.   Risk factor significantly associated with anxiety included current cigarette smoking, exposure to combat on three or more occasions, and a variety of antisocial traits.   Having ever been seen by a mental health professional and a history of psychiatric hospitalization were strongly associated with anxiety with ORs respectively of 6.0, 95% CI:   3.9-9.2, and 6.5, 95% CI:   3.4-12.4.  

          COMMENT:   Army enlistees were screened psychiatrically before they were put on active duty during World War II.   I wonder if those who had seen a mental health professional or had a history of psychiatric hospitalization would have been allowed to serve in those days.