DURHAM, N.C. -- Duke University Medical Center researchers
have found preliminary evidence suggesting a man's lifetime
risk of prostate cancer may be linked to the amount of male
hormone testosterone circulating in his body as early as
puberty or even in utero, although direct evidence of this link
remains to be shown.

The two possible risk factors they found -- high "free"
testosterone levels in adulthood and a small shoulder span in
relation to body size -- appear to be unrelated to one another.
However, they are both tied to hormone levels at various stages
of development, said Wendy Demark-Wahnefried, associate
professor of surgery at Duke and lead author of two parts of a
study that produced the findings.

While doctors won't be able to predict who will get the
disease based on these two factors alone, the results suggest
that "free" hormone levels and shoulder span could be
benchmarks for determining who is at greater risk for the
disease, Demark-Wahnefried said. Free testosterone refers to a
type of hormone that is not bound to a protein and thus can
freely enter cells throughout the body.

"We have to look at how hormone levels at different points
in time actually determine the risk of prostate cancer," she
said. "It is hypothesized that hormone levels throughout life
-- ranging from in utero to old age -- drive such events as
skeletal and muscle formation, fat deposition, baldness, and
that these events may provide the initial stimuli and promotion
for prostate cancer.

"By studying the tell-tale signs that hormones leave on the
body, our goal was to clearly separate those men at risk for
prostate cancer from those who are not."

Demark-Wahnefried's research, funded by the National Cancer
Institute and the Cancer Research Foundation of America, set
out to measure the link between prostate cancer and factors
such as height, weight, musculature and baldness -- all of
which are related to hormones. The two-year, blinded,
case-controlled study compared a group of 159 men with prostate
cancer to a control group of 156 men who had come to the
urology clinic for prostate screenings and other concerns such
as kidney stones. Subjects were aged 50 to 70 years.

In the first phase of the study, Demark-Wahnefried and her
colleagues at the Duke Comprehensive Cancer Center found nearly
a two-fold increase in the risk of prostate cancer among men
with high "free" testosterone levels, the form of testosterone
that can readily be used by cells throughout the body.

While the link between testosterone and prostate cancer has
been made before, previous studies have measured "total"
testosterone, a less active form of the hormone that is bound
to specific protein and thus cannot enter the cells.

The researchers also found a link between high testosterone
levels and vertex or "top of head" baldness. However, baldness
was not linked to prostate cancer in the study subjects,
probably because baldness is as much related to age as it is to
other factors like testosterone, she said.

Demark-Wahnefried theorizes that baldness at a younger age,
perhaps at 40, could be used to predict the later risk of
prostate cancer, a theory that she plans to study next.

The part of the study that looked at baldness, testosterone
levels and prostate cancer is published in the
September/October issue of the Journal of Andrology. Co-authors
of that study include Samuel M. Leski of Boston University
School of Medicine and Mark R. Conaway, Cary N. Robertson,
Richard V. Clark, Bruce Lobaugh, Barbara J. Mathias, Tara Smith
Strigo, and David F. Paulson of Duke.

In the next phase of the study, published in the September
issue of the Journal of Nutrition and Cancer, researchers found
that men with prostate cancer were more likely to have a
narrower shoulder span in proportion to their overall body
size, a trait that earlier studies have shown to be determined
during puberty.

Demark-Wahnefried said those earlier studies, although quite
limited in their sample size, showed that men who go through
puberty later have a broader shoulder span than men who go
through puberty early. She said this finding suggests that
hormone levels have a direct influence on shoulder span.

While the difference in shoulder span was less than a
centimeter, Demark-Wahnefried said it was the only physical
factor she studied that was significantly associated with
prostate cancer.

"Shoulder span may provide us with a benchmark of past
hormonal and/or nutritional status and help elucidate the
etiology of this disease," she said.

While researchers have long believed that prostate cancer is
linked to male hormone levels, Demark-Wahnefried said the
existing research has yielded conflicting results, similar to
the controversy surrounding estrogen as a risk factor for
breast cancer.

The current studies provide strong evidence that risk of
prostate cancer is, in fact, influenced by hormonal events that
occur much earlier in life, such as the formation of skeletal
frame, she said. However, Demark-Wahnefried said these events
may originate in utero and continue to manifest themselves at
developmental milestones throughout a man's life.