ological studies, toxicological assays, animal studies, double-blind human trials,
meta-analyses, case reports and opinions. Within the field of epidemiology, a
variety of approaches can be used, including:
• Case-control studies. In this type of
study design, retrospective exposure factors are compared between persons with
the disease of interest (cases) and persons without the disease of interest (
controls). A variety of factors, including
environmental exposure, tobacco use,
nutrition and exercise level, are analyzed
to determine the potential effect on disease development. Isolating one factor’s
impact from another can be extremely
difficult, especially when it comes to
diet, since foods are eaten in combination. A major limitation of this study design is that participants are required to
recall their prior history of exposure.
• Cohort studies. Cohort studies can be
prospective, retrospective or ambispec-tive. Most cohort studies of diet and
cancer are prospective in design. In
these studies, a very large group of participants with and without the exposure
of interest (e.g., high vs. low consumers
of processed meat) are followed forward
in time, and the rate of disease development is compared between groups of
people with and without the exposure(s)
of interest. This type of study design theoretically overcomes many of the limitations of a case-control design.
• Meta-analysis. A meta-analysis is a
statistical technique that combines published results data across studies to produce an overall, or weighted, summary
of relative risk. This type of analysis may
provide a more accurate reflection of the
association across the available literature
with enhanced statistical power. The results of a meta-analysis, however, are
only as valid as the individual studies included in the assessment. Thus, the
“garbage in, garbage out” principle.
• Pooled meta-analysis. A pooled analysis is similar to a meta-analysis except
that in a pooled analysis, individual or
primary source data are combined across
studies to produce a de novo estimate of
relative risk. A pooled assessment may
allow for the standardization of exposure categories and analytical proce-
OCTOBER NOVEMBER 2008
dures. This is the approach that was used in Harvard’s upublished Pooling Project
paper on meat and colorectal cancer that found no relationship between the two.
Epidemiological Findings are Just a Piece of the Puzzle
Findings from these types of studies must be combined with experimental evidence
from animal bioassays and human trials and must fit together like an interlocking puzzle to feel confident in declaring cancer causation.
With respect to the conclusions on meat consumption, WCRF/AICR relied heavily on selected epidemiological studies. Just as in a courtroom, all the evidence must
be assessed and persuasive evidence must be presented beyond a reasonable doubt.
Progress through innovation
&RPSOHWH ,QVSHFWLRQ 6\VWHPV
How effective is your inspection program? To assure maximum
product safety and customer satisfaction, call Heat and Control.
We provide the latest technology for detection of foreign objects,
defects and missing items. At our demo centers, you can test
different metal detectors, checkweighers and X-ray inspection
equipment to get the best individual machine or combination
system for your products.
Contact Heat and Control today. We’ve got 57 years of experience,
and the complete line of packaging equipment and technical
support services ready to work for you.
,VKLGD ; UD\ LQVSHFWLRQ FKHFNZHLJKHUV
&(,$ PHWDO GHWHFWRUV
,VKLGD PXOWLKHDG ZHLJKHUV VQDFN EDJPDNHUV
)DVW%DFN FRQYH\RUV SURGXFW KDQGOLQJ V\VWHPV
7UD\ FDUWRQ ¿OOHUV
0RGXODU VXSSRUW SODWIRUPV
)RRG SURFHVVLQJ V\VWHPV
See us at Pack Expo
Booth N-3935
www.heatandcontrol.com/fso
OneCallDelivers8002275980
