Numerous observational and cohort studies have assessed and found inconclusive the relationship of a low-fat, high-fruit, and high-vegetable dietary pattern with breast and colon cancer risk. Principal results of the randomized controlled Women's Health Initiative Dietary Modification TrialArticleArticleArticle, which directly assessed the health benefits of a low-fat eating pattern, are reported in this issue of JAMA. Postmenopausal women in the intervention group were advised to reduce total fat intake to 20% of energy and to consume at least 5 servings of fruits and vegetables and 6 servings of grains daily; women in the control group continued their usual eating pattern. For an average of 8.1 years of follow-up, the investigators found the dietary intervention was associated with a modest but statistically nonsignificant reduction in invasive breast cancer and no reductions in the risks of colorectal cancer, cardiovascular disease, coronary heart disease, or stroke. In editorials, BuzdarArticle discusses the role of dietary intake in breast cancer risk and relapse prevention, and Anderson and AppelArticle discuss the implications of these results for future dietary interventions to reduce cardiovascular disease.
Prior studies of children and adolescents who were born with an extremely low birth-weight (ELBW) document that they are at increased risk of academic, social, and behavioral difficulties compared with their normal birth-weight peers. Among persons born ELBW, the transition to adulthood and achievement of typical goals for young adults have received limited attention. Saigal and colleaguesArticle examined markers of educational attainment, employment, and independence in a prospective, longitudinal study that enrolled ELBW infants and a comparison group of normal birth-weight children. The authors found no differences in the proportions of formerly ELBW young adults who achieved these life goals compared with those who were normal birth weight. In an editorial, Hack and KleinArticle discuss societal and environmental factors that influence the long-term outcomes of ELBW infants.
Meta-analyses yield a summary quantitative estimate of the results obtained in multiple studies of a specific research question. However, if the studies included in the analysis are not representative of all studies conducted, publication bias is said to have occurred and this can bias the results of the meta-analysis. Peters and colleagues conducted simulated meta-analyses to examine the performance of 8 statistical tests to detect publication bias. In an article in this issue of JAMA, the authors report the performance of the commonly used Egger's regression test vs an alternative regression test based on sample size (a modification of Macaskill's test). The authors found that neither test performed well under all circumstances; however, several characteristics of the alternative regression test support its use when odds ratios are the summary estimates in the studies analyzed.
Use of performance-enhancing substances is most visible among elite athletes, but a substantial number of adolescents and children also use such substances and are vulnerable to their health risks.
Factors to consider when choosing a palliative care setting for a terminally ill patient who cannot be cared for at home are reviewed by Hanson and Ersek in their discussion of Mr D, an 84-year-old man with metastatic melanoma. He was hospitalized because of a rapid decline in functional status and dehydration.
Public policy and legal issues related to automated external defibrillator programs.
Join Olga Jonasson, MD, in a teleconference on February 15, 2006, to discuss a randomized trial comparing watchful waiting vs standard open tension-free surgical repair for treating inguinal hernia.
For more information and to register for “Author in the Room,” please visit http://www.ihi.org/authorintheroom.
For your patients: Information about hospice care.
This Week in JAMA . JAMA. 2006;295(6):591. doi:10.1001/jama.295.6.591