Re: Ecological Studies

[John Goldsmith <gjohn@bgumail.bgu.ac.il>]

Radsafers:
     Another component of the ecological fallacy is fundamental to the 
distinction beween epidemiology and clinical observations.  In exposures
to a population group and/or effecs among such a group, it is fallacious 
to assume that valid information is presented for any single individual  or 
any non-random sample of these populations.
      Wade is quite correct in warning about the premise of all 
ecological studies being weak or false.
	           John Goldsmith, M.D., Professor of Epiudemiology 
On Fri, 14 Mar 1997, H.Wade 
Patterson wrote:

> POSTED TO RADSAFE
> 
> Group:
> It's a common misconception that so-called "ecological" studies are
> necessarily erroneous. If you will read what I've written below, perhaps
> you will take a different view.
> 
> Best wishes,
> -- 
> Wade
> 
> mailto:hwade@talltown.com
> 
> H.Wade Patterson
> 1116 Linda Lane
> Lakeview OR 97630
> ph 541 947-4974
> 
> Are All Ecological Studies Fallacious?; Not Necessarily.
> 
> In this forum and elsewhere, ecological studies have been deemed
> fallacious, simply because their authors have assumed that, since the
> majority of a group has a characteristic, the characteristic is related
> to a health state common in the group. Such wholesale judgments are
> incorrect because:
> 
> 1. The ecologic fallacy is committed only when it is MISTAKENLY assumed
> that, because the majority of a group has a characteristic, the
> characteristic is related to a health state common in the group.
> 
> 2. Thus, it is ONLY fallacious when the assumption is MISTAKEN.
> 
> 3.It is not, therefore, NECESSARILY fallacious to assume that because
> the majority of a group has a characteristic, the characteristic is
> related to a health state common in the group.
> 
> Here are examples of ecological studies that were not fallacious. i.e.,
> the assumption of an association was correct.
> a. During the latter years of the Roman Republic it was correctly
> observed that dwelling near the Pontine marshes ( group characteristic)
> was unhealthy (health state).
> b. A British physician * correctly observed that chimney sweeps (group
> characteristic) suffered a high incidence of cancer of the scrotum
> (health state).
> c. Another British physician, Edward Jenner, correctly observed that
> milkmaids (group characteristic)  had a decreased incidence of smallpox
> (health state).
> 
> Thus, it seems that unless the observation is MISTAKEN, the association
> may be correct.
> 
> Finally, therefore, the observations of a particular ecological study
> must be shown to be incorrect before the association is so judged.
> 
> Examples of ecological studies where it has not been shown that the
> observations are incorrect include:
> 
> 1. Craig, L.; Seidman, H. Leukemia and lymphoma mortality in relation to
> cosmic radiation. Blood 17 : 319, 1961.
> 
> 2. Frigerio, N.A.; Ekerman, K.F.; Stowe, R.S. The Argonne Radiological
> Impact Program (ARIP), Part I. Carcinogenic Hazard from Low-Level,
> Low-Rate Radiation; ANL/ES-26 Part I, Environmental and Earth Sciences,
> Sept. 1973.
> 
> 3. Frigerio, N.A.; Stowe, R.S.; Carcinogenic and genetic hazard from
> background radiation. IAEA Symposium, Biological and Environmental
> Effects of Low Level Radiation, vol. 2, pp 285-289, Vienna, 1976.
> 
> 4. Luckey, T. D., Physiological benefits from low levels of ionizing
> radiation. Health Physics, v43, 6, pp 771-789, (1982).
> 
> 5. Wei, L.X.; Zha, Y.R.; Tao , Z.F.; He, W.H.; Chen, D.Q.; Yuan, Y.L.
> Epidemiological investigation of radiological effects in high background
> radiation areas of Yangjiang, China. Journal of Radiation Research, 31,
> 1, pp 119-136, 1990.
> 
> 6. Nambi, K.S.V.; Soman, S.D. Further observations on environmental
> radiation and cancer in India. Health Physics, 59, 3, pp 339-344, 1990.
> 
> 7. Chen, D.; Wei, L. Chromosome aberration, cancer mortality and
> hormetic phenomena among inhabitants in areas of high background
> radiation in China. Journal of Radiation Research, 32 Suppl. 2, pp
> 46-53, 1991.
> 
> 8. Shihab-Eldin, A.; Shlyakhter, A.; Wilson, R. Is There a Large Risk of
> Radiation? A Critical Review of Pessimistic Claims. Environment
> International, 18, pp. 117-151, 1992.
> 
> 9. Latarjet, R. Radiation carcinogenesis and radiation protection.
> Cancer J., 5, pp 23-27, 1992.
> 
> 10. Biological effects of low level exposures : dose-response
> relationships. Edward J. Calabrese, editor. Boca Raton : Lewis
> Publishers, c1994.
> 
> 11. Hickey, R. J.; Bowers, E. J.; Spence, D. E.; Zemel, B. S.; Clelland,
> A. B.; Clelland, R. C. Low Level Ionizing Radiation And Human Mortality
> : Multi-Regional Epidemiological Studies. A Preliminary Report. Health
> Phys. 40(5) :625-641; May, 1981
> 
> 12. Hickey, R. J.; Bowers, E. J.; Clelland, R. C.; . Radiation hormesis,
> public health, and public policy: a commentary. Health Phys. 44(3)
> :207-219; March, 1983
> 
> * I wish to thank David Perry of the Rutherford Laboratoy for the
> complete citattion which is: Pott, Percival (1714-1788), English
> physician who is eponymically known for spinal deformity due to
> tuberculosis. He was also first to describe chimney sweep's cancer
> (carcinoma of the scrotum).
> 
>