Day 017 - 25 Jul 94 - Page 29


     
     1        know, it is not showing a relationship and it is a
              prospective study too which, of course, is a more reliable
     2        study.
 
     3        So I think what Dr. Kinlen is saying basically is that if
              you actually look beyond the earlier literature, look at
     4        some of the more up-to-date literature, studies that have
              been simulated to be conducted by the findings of earlier
     5        studies, that there is no consistent relationship between
              fat and the development of breast cancer.
     6
         Q.   Just turn to page 591.  I will not read the rest of 589 at
     7        the moment at least.  The last paragraph above the
               "Discussion".  "In most studies the reliance on recalled
     8        diet has obvious limitations and any objective measure of
              fat intake would be valuable. Serum cholesterol is
     9        influenced (among other factors) by fat intake, and so has
              relevance here; however, a prospective study of over
    10        90,000 women in California found no relation between serum
              cholesterol level and subsequent breast cancer".  That is
    11        Hiatt and others in 1982.
 
    12        "Discussion.  Among the studies of breast cancer based on
              individual women, whether of the case-control or
    13        prospective type, little support for the fat hypothesis
              has emerged.  In spite of this, the hypothesis finds
    14        widespread favour.  Indeed, in 1982 the Nutrition
              Committee of the United States National Research Council
    15        judged the evidence to be sufficiently strong to recommend
              that Americans reduce their consumption of fat on grounds
    16        of cancer alone.  More recently, the National Cancer
              Institute has embarked on an intervention trial among a
    17        large group of American women to determine if breast
              cancer incidence can be reduced by lowering their intake
    18        of fat.
 
    19        So many reasons are offered to defend the hypothesis in
              the face of the negative findings of individual-based
    20        studies, it would almost seem that the basis for the
              hypothesis was so secure that only confirmatory evidence
    21        is of interest."
 
    22        It may be, I do not know, Dr. Arnott, that that is a
              somewhat ironic or sarcastic -- what do you feel?
    23        A.  I would say it is an ironical remark.
 
    24   Q.    "International correlations re-examined.  In view of the
              largely negative nature of studies of individuals, it
    25        seems reasonable to re-examine the international
              correlational studies, since only these provide striking 
    26        support for, and indeed largely suggested, the 
              hypothesis.  In the most often quoted study, these 
    27        correlations implied that about 80% of the worldwide
              variation can be accounted for by variations in fat
    28        consumption (Armstrong and Doll).  Several observations
              appear to have relevance: First, the basic quality of the
    29        nutritional data used in these analyses is far from high,
              for they relate to crude 'food disappearance' within
    30        countries: Production plus imports for human consumption,
              minus exports, divided by the total population figure,

Prev Next Index