Day 035 - 12 Oct 94 - Page 58


     
     1   Q.   Yes.  "Strongly suggest that international comparisons are
     2        seriously confounded by factors related to
     3        industrialisation"?
     4        A.  Yes.
     5
     6   Q.   "Fat intake ranged from 5 to 47 per cent of energy but
     7        breast cancer rates did not vary with fat intake and were,
     8        at most, about 1/10th of those in the United States"?
     9        A.  Yes.  You are quite right referring to Dr. Campbell's
    10        characterisation of this sort of language.
    11
    12   Q.   You would agree with Professor Campbell.
    13        A.  When we are speaking of Dr. Campbell's own research, I
    14        think Dr. Campbell is probably the best person to speak
    15        for it, rather than Dr. Willett.
    16
    17   Q.   It may be.  We will have to look and see what is written
    18        and compare the two which we have not been able to yet.
    19        "Case-control studies and prospective cohort studies
    20        provide more opportunity to measure and adjust other
    21        factors that might be related to risk of cancer.  For
    22        breast cancer no material association with fat intake has
    23        been found in case control studies from New York, Hawaii,
    24        Australia, Greece and Japan.  Although a study from Canada
    25        was originally reported as showing a positive association
    26        it was not statistically significant" ----
    27        A.  Can you hold there for a moment?
    28
    29   Q.   Yes, sure.
    30        A.  Thank you.
    31
    32   Q.   That study from Canada was probably by Howard, but I will
    33        have to check it -- no, it is by Miller.  "It was
    34        not" -----
    35
    36   MR. JUSTICE BELL:  It was one of the ones referred to by
    37        Kinlen.
    38
    39   MR. RAMPTON:  Yes.  "It was not statistically significant and
    40        examination of the data indicates that the average fat
    41        intake reported by cases was virtually identical to that
    42        reported by control.  In other case-control studies only a
    43        limited list of foods was included, some have construed
    44        sporadic associations with foods containing fat as
    45        evidence that fat causes breast cancer.  These data are
    46        difficult to interpret because of the tendency to focus on
    47        foods within a study from which associations are seen, and
    48        because other aspects of the diet, including total energy
    49        intake, cannot be controlled for."  Do you agree with
    50        those criticisms of the defect in case-control studies? 
    51        A.  Overall, the majority of case-control studies have 
    52        found no relationship between fat and breast cancer. 
    53
    54   Q.   Exactly.
    55        A.  It is not universally true, but that has been the
    56        trend, yes.
    57
    58   Q.   Now we come to some work done by Dr. Willett himself: "The
    59        potential for distortion of associations because of
    60        differential recall" -----

Prev Next Index