but no boys developed iron deficiency anaemia.
But at the outset, 29% of the female runners had
ferritins of less than 12mg·l–1, so this study,
without non-athletic controls, exaggerates the
contribution of training to the iron deficient pro-
files. Instead, it describes the typically low iron
stores of adolescent females—athletic or not—
with a small superimposed effect (fall in ferritin
level) from training.
Rowlandet al. (1987) found the same trend in
adolescent runners. At the start of a season, eight
of 20 females but only one of 30 males was iron
deficient (ferritin < 12 mg·l–1). By the end of the
season, one additional female and four addi-
tional males had become ‘iron deficient’ by the
same definition. No runner developed iron defi-
ciency anaemia.
Besides running, many other types of training
have been shown to decrease ferritin level. When
young men and women underwent a 7-week
(8 h · day–1) military-type basic training pro-
gramme, ferritin levels fell an average of 50% and
haemoglobin levels fell more than 5% (Maga-
zaniket al. 1988). When untrained men cycled
2 h · day–1four to five times a week for 11 weeks,
mean serum ferritin fell 73%, from 67 to 18mg·l–1
(Shoemakeret al. 1996). Rowland and Kelleher
(1989) found no significant fall in serum ferritin
during 10 weeks of swim training in adolescents,
but nearly half of the female swimmers studied
beganwith ferritin levels under 12mg·l–1(so there
was little room to fall). In contrast, Roberts and
Smith (1990) reported a decrease in ferritin over 2
years in female synchronized swimmers.
Even strength training decreases ferritin, as
shown by the 35% fall in ferritin in 12 untrained
men who underwent a 6-week strength-training
programme (Schobersberger et al. 1990). A
modest fall in ferritin was seen when young
women underwent a 13-week programme of
modest aerobic calisthenics (Blum et al. 1986).
Training also can decrease ferritin level in cross-
country skiers (Candau et al. 1992), female bas-
ketball players (Jacobsen et al. 1993) and in speed
skaters and field hockey players (Cook 1994).
So athletic training can decrease the ferritin
level. This decrease, however, is not necessarily
pathophysiologic; it may reflect only a shift of
iron from stores to functional compartment
(haemoglobin and myoglobin). Also, any
‘anaemia’ that develops in the same athlete may
be only pseudoanaemia, not necessarily iron
deficiency anaemia.Competition and iron profile
In contrast to prudent training, all-out competi-
tion, especially a prolonged or muscle-damaging
event, clouds interpretation of iron status by
evoking the acute phase response (Eichner 1986).
The acute phase response is an innate, general-
ized host defense against infection or inflam-
matory injury. In athletes, this may begin as
damaged muscle activates complement, which
recruits and activates neutrophils and monocytes
(and fibroblasts), which release cytokines (e.g.
interleukins 1 and 6, tumour necrosis factor). The
cytokines trigger muscle proteolysis and the
hepatic synthesis of proteins (e.g. C-reactive
protein, ceruloplasmin, haptoglobin, fibrinogen
andferritin) that may contribute to host defense.
The interleukins also activate lymphocytes,
cause mild fever and sleepiness, and decrease
serum iron level. So in the acute phase response,
as during a US Army Ranger training pro-
gramme (weeks of intense physical activity,
stress and sleep deprivation) serum iron falls
(and later rebounds), yet serum ferritin rises
(Moore et al. 1993).
In light of recent research on bench-stepping
exercise (Gleeson et al. 1995), it seems likely that
any exercise bout that evokes delayed onset
muscle soreness and damages muscle (sharply
increases serum creatine kinase level) can spur
an acute phase response that alters the markers
of iron balance. Indeed, an integration of diverse
field studies of athletes confirms this.
For example, during marathons and ultrama-
rathons (Dickson et al. 1982; Strachan et al. 1984;
Lampeet al. 1986a; Schmidt et al. 1989), multiday
foot races (Dressendorfer et al. 1982; Seiler et al.
1989), triathlons (Rogers et al. 1986; Taylor et al.
1987) and distance ski races (Pattini et al. 1990),
serial sampling and analysis of blood markers