suggests the following sequence of events and
mechanisms.
1 An early (first day or two) decline in plasma
haptoglobin level and rise in serum ironlevel,
likely from exertional or ‘footstrike’ haemolysis;
along with an increase in plasma volume and fall
in haematocrit (i.e. dilutional pseudoanaemia)
that can not be prevented by iron supplementa-
tion (Dressendorfer et al. 1991).
2 A later (next few days) fall in serum iron and
increase in serum ferritin, likely from the acute
phase response.
3 A late (e.g. later stages of a 20-day foot race)
return of serum iron and ferritin toward baseline, as
the body seems to adapt to the stress of racing
and the acute phase response abates.
The above perturbations—some physiologi-
cal, some pathophysiological—make it difficult
to gauge iron balance in a given athlete who may
be resting, training or racing. For example, an
‘anaemia’ may reflect dilution or iron deficiency;
the serum iron may be normal, high, or low
depending on the stage of the race; and serum
ferritin falls with training yet rises if muscle
damage evokes the acute phase response.
Bearing these confounders in mind, practical
issues of iron balance in athletes are covered
next.
Iron status of fit athletes
Because training decreases ferritin level (iron
stores), one might expect highly trained athletes
to be iron deficient compared with non-athletes.
Indeed, beginning two decades ago, a spate of
cross-sectional surveys suggested that certain
athletes, especially distance runners, tended to
be iron deficient. This area of inquiry began in
part with concern that iron intake was insuffi-
cient in young Canadian women, and that
haemoglobin levels of Canada’s 1976 Olympic
athletes were ‘suboptimal’ (meaning lower than
those of the 1968 Australian Olympic Team). This
led to a survey of 52 collegiate distance runners
in Canada (Clement & Asmundson 1982), con-
cluding that 29% of male and 82% of female
330 nutrition and exercise
runners were ‘at risk for iron deficiency’ (had
serum ferritin of less than 25mg·l–1).
Corroborating surveys followed. Low bone
marrow iron stores were seen in competitive dis-
tance runners in Sweden and Israel (Ehn et al.
1980; Wishnitzer et al. 1983). In a second study
from Sweden (Magnusson et al. 1984), 43 elite
male distance runners had lower ferritin levels
and marrow iron scores than 100 non-athletic
controls. ‘Systemic iron deficiency’ (low satura-
tion of transferrin) was found in 56% of 113
joggers and runners in Denmark (Hunding et al.
1981). Competitive distance runners in Germany
had lower ferritin levels than elite rowers or pro-
fessional cyclists (Dufaux et al. 1981). From South
Africa came reports that 14% of male ultra-
marathoners (but just 2% of controls) had low
ferritin levels, as did 16% of female marathon
runners (Dickson et al. 1982; Matter et al. 1987).
From the USA came a report that one third of
women—and 7% of men—at a marathon fitness
exposition had low ferritins (Lampe et al. 1986b).
Other reports followed suit, as reviewed by Cook
(1994).
These surveys are limited, however, by small
sample size, no or few non-athletic controls, and
widely different definitions of ‘iron deficiency’
(Humpty Dumpty redux). The most reliable
studies are those based on ferritin assay, but
these used different ferritin ‘cut points’ for diag-
nosing iron deficiency (e.g. 12, 20, 25 or 40mg·l–1).
The median ferritin value for young women in
the USA is 25–30mg·l–1(Cooket al. 1986). A cut
point of 40mg·l–1(Matteret al. 1987) classifies
most young women—athletic or not—as iron
deficient. Even 25mg·l–1(Clement & Asmundson
1982) is too high; experts say that less than 12mg·
l–1is the proper ferritin cut point for diagnosing
iron deficiency (Cook 1994).
Other surveys question whether athletes differ
from non-athletes in iron balance, but these
surveys too have problems. In one study, 19 top-
level soccer players had ferritins similar to 20
controls (Resina et al. 1991). Likewise, ferritin
was similar in 72 elite runners vs. 48 non-runners
(Balabanet al. 1989). In this survey, however, one