With respect I think you should take care when trying to assess one paper
in an area when you are admittedly not an expert.
Both the infamous NEJM letter (it was not a peer reviewed paper) and the
EJC paper you just found are seriously flawed (see my published letter
regarding the latter).
I also question the simultaneous posting of emails to 3 lists.
That said, the epidemiological literature is full of such studies, where
one may certainly wonder what degree of selection went on - both selection
within the study (eg reporting only a subset of subgroup analyses) and
across studies (notably non-publication of 'uninteresting findings). the
fact that the authors find some plausible explanation is generally of low
value in my opinion.
The following recent paper is highly relevant in this context:
Ioannidis JP.
Why most published research findings are false.
PLoS Med. 2005 Aug;2(8):e124.
which is freely available at
http://medicine.plosjournals.org/archive/1549-1676/2/8/pdf/10.1371_journal.pmed.0020124-L.pdf
Doug
At 20:48 14/10/2005, =?iso-8859-8-i?B?9un06SDl4+XjIPnl5+g=?= wrote:
>I followed a suggestion to search for articles linking right-handeness to
>cancer. Look what I found:
> From Tzippy
>
>Research Papers
>
>Epidemiologic correlates of breast cancer laterality (Sweden)
>
>Anders Ekbom1, 2 [], Hans-Olov Adami1, 2, Dimitrios Trichopoulos2, Mats
>Lambe1, 3, Chung-cheng Hsieh2 and Jan Pontén4
>(1) Cancer Epidemiology Unit, University Hospital, S-751 85 Uppsala, Sweden
>(2) Department of Epidemiology, Harvard School of Public Health, Boston, USA
>(3) Department of Social Medicine, University Hospital, Uppsala, Sweden
>(4) Department of Pathology, University Hospital, Uppsala, Sweden
>
>Received: 2 May 1994 Accepted: 27 June 1994
>Abstract Breast cancer laterality was studied in relation to age in
>80,784 cases of invasive and 3,835 cases of pre-invasive breast cancer in
>women and 548 cases of invasive breast cancer in men reported to the
>Swedish Cancer Registry, 1970–89. In a subset of 11,274 women with
>invasive disease, data on parity were available through the Swedish
>Fertility Registry. Laterality also was evaluated in relation to age and
>reproductive variables in 3,986 cases from an international study from the
>1960s. The overall incidence of pre-invasive and invasive cancer was
>higher in the left than in the right breast among both women and men. The
>excess incidence of invasive cancer in the left breast was evident only
>after the age of 45 years in women; a similar phenomenon may exist with
>pre-invasive disease in women and in men. The age-dependent laterality
>pattern did not appear to be confounded by menopausal status. Among women
>younger than 45 years, nulliparity, right handedness, and late age at
>menarche was associated with a somewhat higher incidence of cancer in the
>right breast. The laterality findings are likely to be due to factors
>operating early in the carcinogenic process, perhaps at the pre-initiation
>stage.
>
>
_____________________________________________________
Doug Altman
Professor of Statistics in Medicine
Centre for Statistics in Medicine
Wolfson College Annexe
Linton Road
Oxford OX2 6UD
email: [log in to unmask]
Tel: 01865 284400 (direct line 01865 284401)
Fax: 01865 284424
Web: http://www.csm-oxford.org.uk/
|