tive’. Remes (1979) found that 6 months of mili-
tary training increased red cell mass by 4%.
Young et al. (1993) found a 4% increase after 8
weeks of regular cycling, but the unusual proto-
col (cycling immersed to the neck in water) pre-
vented the expected increase in plasma volume.
Rayet al. (1990) found that regular, upright
cycling for 8 weeks increased red cell mass about
220 ml, but this increase was not significant.
Shoemakeret al. (1996) found no increase in red
cell mass when untrained men cycled regularly
for 11 weeks, but it seems possible that blood-
drawing for testing offset an increase in red cell
mass.
If training does increase the red cell mass, a
likely mechanism would be via erythropoietin.
But whether training—or competing—increases
serum erythropoietin level is also unclear;
studies finding an exercise-induced increase are
slightly exceeded by studies finding no change
(Weight et al. 1991; Klausen et al. 1993; Shoemaker
et al. 1996). All told, it seems likely that strenuous,
long-term athletic training (at sea level) can
increase red cell mass, but this is by no means
proven, and the mechanism for any increase
remains unclear. We need more research here.
Inadequate iron intake
If dietary iron is inadequate for physiological
needs, ferritin will decline. Insufficient dietary
iron can be a problem for female athletes who are
dieting, who have eating disorders, or who are
vegetarians. It is rarely a problem in men. No
good evidence exists for impaired absorption of
iron in athletes; the one such report (in male dis-
tance runners) seems flawed by abnormally high
iron-absorption in the non-athletic controls,
blood donors who may have been iron deficient
(Ehnet al. 1980).
Gastrointestinal bleeding
Gastrointestinal bleeding in athletes has been
widely studied and reviewed (Eichner 1989,
1996). It occurs most often in distance runners
and ranges from occult and trivial to overt and
332 nutrition and exercise
grave. If all reports are lumped, about 2% of
recreational marathoners and triathletes have
seen blood in their stool after running and about
20% have occult faecal blood after distance races.
In general, the longer the event and/or the
greater the effort, the greater the likelihood of
bleeding. In a recent study of 20 male triathletes
who provided stool samples during training,
taper and competition, 80% of the men had
occult faecal blood on one or more of the tests
(Rudzkiet al. 1995).
The source of gastrointestinal bleeding varies.
Anorectal disorders (e.g. fissures, haemorrhoids)
can be the source. Gastro-oesophageal reflux
commonly occurs in runners, but no reports have
identified an oesophageal source of bleeding. In
some athletes, no source is found depite endo-
scopies. The most common source is likely the
stomach, as verified by endoscopy studies of
runners after distance races (Schwartz et al. 1990).
Usually, this is a mild gastritis with superficial
erosions that heal quickly. Rarely, however, an
athlete bleeds massively from a peptic ulcer
during or just after running (Eichner 1996).
Aspirin or other analgesics can increase the risk
of gastrointestinal bleeding, as shown in a field
study of marathoners (Robertson et al. 1987). In a
recent study, aspirin sharply increased gastroin-
testinal permeability when volunteers ran 1 h on
a treadmill (Ryan et al. 1996).
The second most common source seems to be
the colon, usually from a segmental haemor-
rhagic colitis, presumably ischaemic. During
strenuous exercise, splanchnic blood flow may
decrease by as much as 80%, as blood diverts to
working muscles. Normally, the gut tolerates
this, but occasionally the colonic mucosa
becomes ischaemic, in line with level of effort,
unfitness, sympathetic response and dehydra-
tion. The result is superficial haemorrhage and
erosions. Most cases are mild and soon revers-
ible, but rare cases require subtotal colectomy
(Eichner 1996).
Repeated bouts of ischaemic colitis in female
runners could contribute to iron deficiency
anaemia. In a recent German study, seven of 45
elite male distance runners had ferritins of less