Original Article | Published:

Iron supplementation maintains ventilatory threshold and improves energetic efficiency in iron-deficient nonanemic athletes

European Journal of Clinical Nutrition volume 61, pages 3039 (2007) | Download Citation

Guarantor: PS Hinton.

Contributors: PSH contributed to the design of the experiment, analysis of the data and assisted with writing and editing the manuscript. LMS collected and analyzed the data and contributed to the writing of the manuscript.




To determine the effect of iron supplementation on iron status and endurance capacity.


Randomized, double-blind iron supplementation.


University of Missouri-Columbia and surrounding community.


Twenty iron-deficient (serum ferritin, sFer<16 μg/l; serum transferrin receptor, sTfR>8.0 mg/l; or sTfR/log sFer index >4.5), nonanemic (hemoglobin, Hb>120 g/l, women; >130 g/l, men) men and women (18–41 years) were recruited via fliers and newspaper advertisements; 20 of 31 eligible subjects participated.


A 30 mg measure of elemental iron as ferrous sulfate or placebo daily for 6 weeks.


Dietary iron intake and physical activity did not differ between groups before or after supplementation. Iron supplementation significantly increased sFer compared to placebo (P=0.01), but did not affect Hb or hematocrit. Iron supplementation prevented the decline in ventilatory threshold (VT) observed in the placebo group from pre- to post-supplementation (P=0.01); this effect was greater in individuals with lower sFer before intervention (P<0.05). Changes in sFer from pre- to post-treatment were positively correlated with changes in VT (P=0.03), independent of supplementation. The iron group significantly increased gross energetic efficiency during the submaximal test (P=0.04). Changes in sFer were negatively correlated with changes in average respiratory exchange ratio during the submaximal test (P<0.05).


Iron supplementation significantly improves iron status and endurance capacity in iron-deficient, nonanemic trained male and female subjects.


Missouri University Alumni Association, by the Elizabeth Hegarty Foundation and by the Department of Nutritional Sciences.


Iron deficiency is a progressive condition that develops through three stages: iron depletion, iron deficiency and iron deficiency anemia (Food and Nutrition Board, Institute of Medicine, 2001). In the United States, the prevalence of iron deficiency anemia among women is 3–5%, whereas the prevalence of iron depletion and deficiency without anemia is much higher, approaching 16% (Looker et al., 2002). The prevalence of iron deficiency is lower among male subjects compared with female subjects, estimated to be 2% (Looker et al., 2002). Iron deficiency is more common among physically active individuals compared with their sedentary counterparts (Magnusson et al., 1984; Blum et al., 1986; Balaban et al., 1995; Chatard et al., 1999; Malczewska et al., 2000, 2002; Sinclair and Hinton, 2005), affecting 25–35% of adolescent and female athletes and 10–11% of male subjects (Constantini et al., 2000; Dubnov and Constantini, 2004). The negative functional and clinical consequences of anemia have been well documented, including fatigue and decreased maximal and submaximal work capacity (Gardner et al., 1977; Celsing and Ekblom, 1986; Celsing et al., 1988). More recently, mild to moderate iron deficiency without anemia has been found to adversely impact submaximal work capacity (Zhu and Haas, 1997, 1998a, 1998b; Hinton et al., 2000; Brownlie et al., 2001, 2004; Brutsaert et al., 2003).

Endurance is defined as the ability to maintain a submaximal workload until exhaustion. Accurately measuring endurance in humans is difficult because study participants must exercise for extended periods of time and performance is highly dependent on subject motivation. We and others have used time trials of sufficient length that subjects exercise at submaximal intensity for the majority of the test (Zhu and Haas, 1997, 1998a, 1998b; Hinton et al., 2000; Brownlie et al., 2001, 2004). We previously reported greater improvements in endurance, assessed by a 15 km time trial on a cycle ergometer, after 4 weeks of training in iron-deficient untrained women who received an iron supplement (20 mg elemental iron per day) compared with those who received a placebo (Hinton et al., 2000). Zhu and Haas (1998b) also tested endurance capacity of iron-deficient nonanemic women, measured by a 15-km time trial on a cycle ergometer before and after 8 weeks of iron treatment (45 mg elemental iron per day). The iron-supplemented female subjects had a 2.0 kJ/min lower energy expenditure (EE) and a 5.1% lower fractional utilization of peak oxygen consumption during the endurance trial compared with the placebo group.

A second alternative to exhaustive, submaximal exercise was employed by Brutsaert et al. (2003). These investigators used a progressive muscle fatigue protocol in order minimize the effects of motivation on endurance. After iron treatment, the rate of decrease in maximal voluntary contractions was attenuated in the iron group but not in the placebo group. Taken together, these studies suggest that repletion of iron stores in iron-deficient nonanemic women enhances favorable adaptations to aerobic exercise training, possibly by increasing energetic efficiency during submaximal exercise.

The purpose of this study was to extend our previous studies of iron deficiency without anemia and endurance capacity in sedentary individuals to chronically trained men and women. We hypothesized that iron versus placebo supplementation would significantly improve iron status over time; have a positive effect on ventilatory threshold (VT), but no effect on maximal oxygen consumption (VO2peak); significantly decrease respiratory exchange ratio (RER) and increase efficiency during the 60-min submaximal cycle test. We also hypothesized that the response to iron supplementation would be greater in subjects who were more iron-depleted before intervention, as they would have a greater potential to respond.

Subjects and methods


Ninety-four (35 male and 59 female) healthy, recreationally trained individuals (at least 60 min of aerobic exercise, 3 days/week for more than 6 continuous months), aged 18–41 years were recruited from the local community as potential subjects. Based on power calculations for changes in ferritin and serum transferrin receptor (sTfR) in response to iron supplementation from data obtained in an earlier supplementation study (Hinton et al., 2000), our target sample size was 10 subjects per group. After preliminary screening, 27 women and four men were identified as iron-deficient but not anemic on the basis of normal hemoglobin (Hb) concentrations (♀Hb>120 g/l, ♂Hb>130 g/l) and a subnormal result of at least one other indicator of iron status (serum ferritin (sFer)<16 μg/l, sTfR>8.0 mg/l or sTfR/log sFer index>4.5). As sFer levels reflect the iron stores and sTfR reflects the functional iron pool, the use of both indicators for classification of iron deficiency allows for evaluation of the full spectrum of iron status. These two indicators are inversely correlated, and their ratio, the sTfR/log sFer index, has recently been used as an indicator of iron deficiency (Baynes, 1996; Punnonen et al., 1997, 1998). This index is particularly useful for athletes whose sFer is low but above the cutoff for iron depletion (i.e. 16–30 ng/dl) and whose sTfR are elevated. Additionally, athletes may be identified as iron-deficient when classified using sFer, sTfR and the sTfR/log sFer index. Thus, indicators of iron stores and functional tissue iron were used to identify iron-deficient subjects in the current study.

All subjects were informed that the purpose of the screening was to identify iron-deficient individuals for a study comparing iron and placebo supplementation. Subjects were asked whether they would be willing to participate in this study if they were in fact found to be iron-deficient. Of the 31 subjects who qualified, 20 participated in the study (three men and 17 women). Subjects were made aware of the risks associated with this study and signed a consent form that had been approved by the University of Missouri Health Sciences Internal Review Board. All procedures involving human subjects were in accordance with the ethical standards of the University of Missouri Institutional Review Board and with the Helsinki Declaration of 1975 as revised in 1983.

The initial screening involved a blood collection of 7 ml and determination of Hb, hematocrit (Hct), sFer and sTfR to evaluate iron status. All subjects who participated in the screening received an iron status assessment; those who qualified for the study were told that they were iron-deficient and were asked to participate in the iron supplementation and physical performance study. Subjects completed a medical history questionnaire and none were found to possess the following exclusion criteria: current pregnancy or pregnancy within the previous year, recent infectious illness or fever, chronic inflammatory diseases, hemolytic anemia, musculoskeletal problems, history of eating disorders, smoking or consumption of medications that may interfere with dietary iron absorption or that have anticoagulant properties. In addition, subjects currently were not using iron supplements.

Study design

The experimental design of the study was a randomized, double-blind placebo controlled intervention trial. Eligible subjects were randomly assigned and gender-balanced into two groups, receiving either an iron supplement of 30 mg elemental iron as ferrous sulfate or an identical placebo capsule containing lactose filler (made by Investigational Drug Service at the University of Missouri) once daily for 6 weeks. This dose of iron has been shown to improve iron status after 3 weeks (Hinton et al., 2000). The subjects were provided a 6-week supply of capsules and were instructed to consume the capsules with meals to reduce side effects and with citrus juice to enhance iron absorption. They were also instructed not to consume any other multivitamin or mineral supplements containing iron during the entire study period and were asked to continue their regular diet for the duration of the study. Subjects recorded capsule ingestion, consumption of medications, illness, menstrual status, gastrointestinal symptoms, physical activity and musculoskeletal problems in a daily log. Subjects were asked to return the supplement containers with any unconsumed supplements at the end of the 6 weeks to monitor compliance with the supplementation.

Physical activity

Pre-study habitual physical activity levels were assessed by using a questionnaire that was based on the Compendium of Physical Activities (Ainsworth et al., 1993) and asked subjects to report duration, intensity and frequency of each activity they participated in during the previous 7 days, as previously described (Hinton et al., 2000). Energy expended during exercise was calculated using self-reported habitual physical activity (activity type, intensity, duration and frequency), measured body weight and published metabolic equivalents (METS, 1 MET=4.186 kJ/kg body weight/h) for various activities (Ainsworth et al., 1993). This was carried out to confirm similar physical activity levels between treatment groups after randomization and to eliminate subjects who did not meet the minimum required activity level (60 min aerobic exercise, 3 days/week for more than 6 continuous months).

Nutrient intake and EE

Participants completed written 3-day diet records (including 2 weekdays and 1 weekend day) during the first and last weeks of the study to monitor dietary iron intake. Dietary records were analyzed using the Food Processor II, Nutrition and Diet Analysis software, version 8.0 (esha, Salem, OR, USA). The results of these dietary analyses were used to characterize the diets of the participants and to assess the amount of iron consumed in the daily diet. Subjects also completed daily physical activity logs, and weekly physical activity was quantified (kJ/week) using the Compendium of Physical Activities (Ainsworth et al., 1993) based on activity type, frequency, duration and intensity.


Body size and composition were assessed using standard procedures (Lohman et al., 1988). Body composition was determined by gender-specific three-site skin fold measures, that is, chest, abdomen and thigh for men and triceps, suprailiac and thigh for women. Equations by Jackson and Pollock (1978, 1980) were used to calculate body density and the Siri equation was to calculate percent body fat (Siri, 1956).

Performance measures

Each subject completed a maximal oxygen consumption test and a 60-min submaximal test (60% of VO2peak) on a cycle ergometer pre- and post-supplementation. Subjects performed the VO2peak test first to determine the appropriate workload for the submaximal test, which was performed 24–48 h later. Subjects were asked not to perform any strenuous physical activities 48 h before either exercise test. To control for the effects of dietary intake before exercise testing, subjects were instructed to consume the same diet for pretreatment and post-treatment exercise tests. Subjects were instructed not to consume food or caffeinated beverages 3 h before exercise testing. Exercise testing was not standardized by menstrual cycle phase because whole body substrate oxidation during exercise does not vary across the menstrual cycle in active women (Horton et al., 2002).

Exercise tests were conducted on a mechanically braked, calibrated cycle ergometer, equipped with a digital readout of cadence (r.p.m.) and distance (km) pedaled (model 818E, Monark, Varberg, Sweden). Concentrations of oxygen and carbon dioxide in expired air and respiratory volume were measured using an Ametek S3A/I oxygen analyzer, a Parvo Medics CO2-100 gas analyzer and a Hewlett Packard 47303A Digital Pneumotach Vertek Series computer and interface system. A Hans Rudolph breathing valve was used for both tests (Kansas City, MO, USA). Data output from the instruments was directed to the computer for breath-by-breath calculation of oxygen consumption (VO2), CO2 production (VCO2), RER (VCO2/VO2) and minute ventilation (VE), collected and averaged over 15 s intervals. Heart rate (HR) was monitored throughout the tests with an electrocardiograph (Quinton Cardiology, Bothell, WA, USA). ECG leads were connected at site V1, V6 and between intercostals three and four on the right side of the rib cage.

Peak oxygen consumption (VO2peak) was measured following a modification of the protocol described by McArdle and Magel (1970) for the cycle ergometer. Subjects warmed up for 5 min at a self-selected pace and resistance. Subjects then chose their preferred cadence, which was kept constant throughout the test; subjects pedaled at this cadence during the post-supplementation VO2peak test. Men began the test at a resistance of 1.5 kp and women began at 1.0 kp for the first 2 min of the test. After the initial 2 min, the resistance was increased 0.25 kp every 2 min until the subjects were unable to maintain their self-selected cadence. The test was considered a valid measurement of VO2peak if the subject met two or more of the following criteria: RER>1.1, plateau in VO2 readings despite increase in workload, exercise HR no more than 15 bpm below age-predicted max HR. The highest VO2 obtained was considered the subject's VO2peak. VT was determined using the ventilatory equivalent method (VEQ method), defined as ‘that intensity of activity which causes the first rise in the ventilatory equivalent of oxygen (VE/O2) without a concurrent rise in the ventilatory equivalent of carbon dioxide (VE/CO2)’ (Reinhard et al., 1979; Powers et al., 1984). VT was evaluated independently by both investigators and the final determination was made by consensus.

Submaximal exercise performance was assessed during a 60-min exercise bout on a cycle ergometer set at 60% of each subject's individual VO2peak. Subjects pedaled at the cadence that was used during the VO2peak test, and the resistance was set and adjusted accordingly during the test to maintain a workload equivalent to 60% of VO2peak. Oxygen consumption (VO2), CO2 production (VCO2), RER (VCO2/VO2) and minute ventilation (VE) were averaged over 15 s intervals for the duration of the test. HR was recorded every minute using Polar HR monitors (Polar CIC Inc., Port Washington, NY, USA). Gross energetic efficiency during the submaximal test, defined as the ratio of energy expended to external work performed, was calculated by dividing the average EE in watts by the average work rate in watts and multiplying the ratio by 100%. Average EE and average work rate were determined using data averaged over 15 s intervals.

Blood collection and preparation

Iron status was assessed at baseline (week 0) and post-supplement (week 6). As serum iron status indicators (Hb, sFer and sTfR) are not acutely influenced by food intake, the subjects were not required to fast before blood draws. Assessment of iron status was not standardized by menstrual cycle phase as Hb, Hct, red cell volume and sFer are not altered by phases of the cycle (Belza et al., 2005). All blood samples (7 ml) were collected via a butterfly needle inserted into an antecubital vein with the subject seated. A 0.5 ml aliquot of whole blood was immediately removed from the initial 7 ml blood sample and used for Hb and Hct analysis. The remaining amount of blood was collected into a serum separator tube and centrifuged (2000 g) at 4°C for 15 min in a Marathon 22100R centrifuge (Fisher Scientific, Pittsburgh, PA, USA). The separated serum was aliquoted into 1 ml cryogenic vials and stored at −70°C. To control for potential variation in assay conditions, each subject's pre- and post-supplement samples were analyzed simultaneously.

Iron status measurements

Hb was assayed in whole blood immediately following sample collection using cyanomethemoglobin (procedure no. 320, 321, 325, Stanbio Laboratory, Boerne, TX, USA) according to the method of Drabkin and Austin (1935). Hct was determined using the microhematocrit method. sTfR and sFer were determined by using commercial enzyme immunoassays (Ramco Laboratories, Stafford, TX, USA) (Flowers et al., 1986; Punnonen et al., 1998). Serum total iron (sFe), unsaturated iron binding capacity (UIBC), total iron-binding capacity and transferrin saturation were determined using quantitative colorimetric assays (procedure no. 370, Stanbio Laboratory, Boerne, TX, USA) according to a modified method of Persijn et al. (1971).

Statistical analyses

Data were examined to verify normality of distribution by examining skewness and kurtosis. sFer showed a skewed distribution, and statistical analyses were performed on natural-log-transformed data. Independent Student's t-test was used to test group differences at baseline. Statistical analyses included 2 × 2 analysis of variance (ANOVA) with repeated measures on time (pre- and post-supplement) to examine group (iron and placebo) and time main effects and group-by-time interactions for measures of iron status and performance variables. In the case of a significant main effect or interaction, one-way ANOVA was used to test for significant changes within group over time. Multiple linear regression analysis (GLM) was used to test whether initial iron status modified the effects of iron supplementation on performance variables. Models were constructed with post-treatment performance variables as the dependent variables and with baseline performance, baseline iron status and group-by-baseline iron status as the independent variables. Statistical significance was indicated at P<0.05. Pearson correlations were used to examine the relationships between change in iron status and change in performance variables from pre- to post-supplementation. Pearson correlations were also used to determine if training load (energy expended during exercise) was related to exercise performance. All statistical analyses were performed using SAS statistical software version 8.0 (SAS, 1999). Results are presented as means±standard deviations (s.d.).


Subject characteristics

Twenty recreationally trained men (n=3) and women (n=17) participated in this study. The treatment groups were balanced with respect to gender. The iron group had nine female subjects and one male subject, whereas the placebo group had eight female subjects and two male subjects. The iron supplement and placebo groups did not differ in age (28.1±5.1 and 27.7±4.4 years, respectively) or height (1.63±0.03 and 1.68±0.03 m, respectively). Body weight and composition did not differ between the two groups before or after the supplementation, nor were there significant changes during the supplementation period (Table 1).

Table 1: Subject characteristics before and after 6 weeks of iron or placebo treatment

Nutrient intake

Dietary iron intake was not different between groups or from pre- to post-supplement. The average daily iron consumption of both the iron and placebo groups was near the RDA for women (18–50 years) of 18 mg/day. The iron group consumed 17.5±5.3 and 19.8±4.6 mg and the placebo group 19.2±11.7 and 17.6±8.0 mg, pre- and post-supplement. The iron group consumed significantly more vitamin C than the placebo group before supplementation (149.3±96.1 vs 63.0±24.2 mg). Post-supplementation, the vitamin C intake of the iron group had significantly decreased to 112.1±68.9 mg vitamin C and that of the placebo group significantly increased to 103.0±50.0 mg vitamin C. There were no significant differences between groups or over time for energy, macronutrients, fiber or micronutrients other than vitamin C (data not shown).

Exercise training

Subjects were asked to maintain their pre-supplement exercise training schedule throughout the 6 weeks of supplementation. Subjects recorded exercise time, mode, pace and intensity in a daily exercise log. The Compendium of Physical Activities (Ainsworth et al., 1993) was used to estimate EE in kilojoules, based on the assumption that resting metabolic rate is equivalent to 4.186 kJ/kg body weight for all subjects. There was no difference between groups over the 6 weeks in average weekly exercise EE or total EE for the 6 weeks (Table 2). Energy expended during training was not correlated with any exercise performance outcome (data not shown).

Table 2: Energy expenditure (kJ) from training logs during the 6 weeks of iron or placebo treatment

Compliance with supplementation

Compliance with supplementation was verified using pill counts of the number of pills left in the pill bottle when returned at the end of 6 weeks and the training logs where subjects recorded the date and time capsules were taken. On average, subjects in the iron group ingested 98±8.2% and the placebo group 99±5.4% of their supplements. There was no significant difference in compliance between the two groups; frequency of side effects recorded in the daily training log did not differ between groups.

Iron status responses to iron and placebo treatments

There were no significant group differences in any measure of iron status at baseline (Table 3), including sFer (P>0.2). For sFer, there was a significant group-by-time interaction (P=0.01); the iron-supplemented group exhibited significant increases in sFer (P=0.003), but the placebo group did not show any significant changes. There were no significant main effects for other iron status indicators, which were in the range for normal values. Baseline ferritin, sTfR and the sTfR to ferritin index did not modify the response to iron supplementation.

Table 3: Indicators of iron status before and after 6 weeks of iron or placebo treatment

Maximal oxygen consumption test

Results from the VO2peak test verified that subjects were recreationally active (Table 4). Maximal oxygen consumptions (l/min or ml/kg/min) for iron and placebo groups were not different from each other at either time point and did not change over time (Table 4). There was a significant group-by-time interaction for VT expressed as oxygen consumption relative to body weight or as a percentage of VO2peak. VT (ml/kg/min) was not significantly different between iron and placebo groups at baseline, and, although the iron group did not change after 6 weeks of iron supplementation, the placebo group showed a significant decrease in VT after 6 weeks of placebo treatment. Moreover, subjects with lower sFer concentrations at baseline exhibited a greater change in VT with iron treatment (Figures 1, 2). In addition, there was a significant positive correlation between the change in log ferritin and change in VT from pre- to post-supplement (Figure 3).

Table 4: VO2peak data before and after 6 weeks of iron or placebo treatment
Figure 1
Figure 1

Relationship between baseline log sFer and change in VT. Iron group (long dash, r=−0.68, P=0.03), placebo group (dotted, r=0.25, P=0.49).

Figure 2
Figure 2

Individual pre- and post-supplementation VT data for iron (▪) and placebo () groups.

Figure 3
Figure 3

Relationship between change in log sFer and change in VT from pre- to post-supplementation (r=0.488, P=0.029).

Submaximal test

During the submaximal test, subjects were tested at 60% VO2peak (Table 5). Average RER for the 60-min submaximal test was not different between groups before or after supplementation (Table 5). There was a significant negative correlation between the change in log sFer and change in RER from pre- to post-supplement (Figure 4) using the entire study population, indicating that an increase in sFer was associated with a reduction in RER. Average HR during the submaximal test did not differ between groups at either time point. The group-by-time interaction was not significant; however, there was a significant main effect of time (P=0.016), which was due to an increase in HR in the placebo group (P=0.023) but not in the iron group (Table 5). EE, measured in kilojoules, was not different between iron and placebo groups during the submaximal test and did not change for either group from pre- to post-supplement. Both groups averaged 1.9 kJ/min during both tests (Table 5). There was a significant main effect of time (P=0.042) for average workload during the submaximal test; however, the group-by-time interaction was not significant (Table 5). It was hypothesized that the iron group would be able to exercise at a higher workload while remaining at the same relative intensity (i.e. 60% VO2peak) after supplementation; so the simple effects of time for different levels of group were analyzed by one-way repeated measures ANOVA. The average workload maintained during the submaximal test increased in the iron group from pre- to post-supplementation (P=0.048; Table 5). Gross energetic efficiency was not significantly different between groups at either time point. There was a significant main effect of time (P=0.05), but the group-by-time interaction was not significant (Table 5). It was hypothesized that the iron group would become more efficient after supplementation while exercising at the same relative intensity (60% VO2peak); so the different levels of group were analyzed by one-way repeated measures ANOVA. Iron supplementation resulted in a statistically significant increase in gross energetic efficiency, that is, greater work output per energy expended (Table 5).

Table 5: Submaximal exercise data before and after 6 weeks of iron or placebo treatment
Figure 4
Figure 4

Relationship between change in log sFer and change in submaximal RER from pre- to post-supplementation (r=−0.435, P=0.05).


The purpose of this study was to examine the effects of iron supplementation on endurance performance in iron-deficient, nonanemic, recreationally trained men and women. As in most of the previous studies, sFer and Hb concentrations were used to identify potential subjects. In addition, sTfR and the sTfR/log sFer index, which have received attention only recently, were also measured and used to classify subjects as iron-deficient but not anemic (Suominen et al., 1998). In the present study, iron supplementation of iron-deficient, nonanemic trained individuals preserved VT and significantly increased gross energetic efficiency during submaximal endurance exercise compared with placebo treatment. The effect of the iron supplement could be attributed to changes in iron status and, in the case of VT, the response to iron treatment was greatest in subjects with the greatest potential for improvement (i.e. low pre-supplementation ferritin).

Iron status responses to iron and placebo treatments

After 6 weeks of iron treatment, iron-deficient subjects exhibited a significant increase in sFer from pre- to post-supplementation (11.7.±7.3–20.8±11.6 μg/l; Table 2). It should be noted that the average post-supplementation sFer was barely within the normal range; several studies classify subjects as iron-deficient with sFer concentrations between 20 and 25 μg/l (Newhouse et al., 1989; Rowland et al., 1989; Klingshirn et al., 1992; LaManca and Haymes, 1993). Although it was not statistically significant, iron supplementation resulted in a decline in sTfR (6.58±1.45–5.96±1.17 mg/l, P=0.176) and sTfR/log sFer index (9.98±10.60–4.96±1.58, P=0.072). Low statistical power may have compromised our ability to detect differences between groups. A larger dose of supplemental iron may have been needed to improve sTfR in these highly active subjects. Endurance exercise increases whole body iron turnover and iron losses via the gastrointestinal tract, hematuria and hemoglobinuria, elevating the estimated average requirement for athletes by 30–70% (Food and Nutrition Board, Institute of Medicine, 2001). In theory, functional iron indicators, that is, sTfR, should improve with supplementation before repletion of stores. Thus, one would expect sTfR to significantly decrease after 6 weeks of iron supplementation. However, sTfR may be increased by skeletal muscle growth (Murray-Kolb et al., 2001) and by eythropoiesis (Fillet and Beguin, 2001), which would diminish the response to iron supplementation.

Maximal oxygen consumption test

As expected, iron supplementation did not affect maximal oxygen consumption (Table 3) (Davies et al., 1984; Celsing and Ekblom, 1986; Newhouse et al., 1989; Klingshirn et al., 1992; Zhu and Haas, 1998b; Hinton et al., 2000). As Hb concentration was above the threshold for normal iron status for both groups at pre- and post-supplementation testing, it was anticipated that VO2peak values would remain relatively constant. Iron supplementation did not significantly increase VT, but prevented the decline observed in the placebo group. As a result of the decrease in VT, the placebo group had an earlier onset of excess production of carbon dioxide, thus decreasing the workload required to achieve VT (Table 3). Individuals with lower baseline iron stores demonstrated greater improvements in VT after iron supplementation than individuals with higher baseline ferritin. Although the placebo group, as a whole, did not exhibit a significant decrease in sFer, some individuals in the placebo group experienced a decrement in sFer from pre- to post-supplementation. Correlations between changes in iron status and in performance measures take individual responses into account. A significant positive correlation was found between change in log sFer and change in VT (Figure 3).

Few studies have examined the effects of iron deficiency without anemia on VT (Newhouse et al., 1989; Zhu and Haas, 1997). Newhouse et al. (1989) failed to show a relationship between VT and sFer after iron supplementation. Zhu and Haas (1997) also found no differences in VT between iron-deficient and iron-replete untrained women. However, it should be noted that some subjects placed in the normal iron group by Zhu and Haas would have been classified as iron-deficient in the present study. It is plausible that VT is affected by iron deficiency because it is closely linked to the lactate threshold (the onset of blood lactate accumulation). In the iron-deficient state, oxidative capacity of the muscle is reduced (Finch et al., 1976; Davies et al., 1984; Willis et al., 1987). Thus, the workload required to initiate the onset of lactate accumulation (i.e. VT) is decreased.

Submaximal test

There was a significant inverse correlation between the change in log sFer and change in RER from pre- to post-supplement (Figure 3). This finding is consistent with increased oxidative capacity of skeletal muscle as functional iron is repleted and is supported by animal studies (Finch et al., 1976; Davies et al., 1982).

Iron-deficient anemic rats have an increased dependence on glucose metabolism at rest and during submaximal exercise (Brooks et al., 1984; Gregg et al., 1989). Brooks et al. (1984) observed that resting iron-deficient rats had significantly increased concentrations of blood glucose and lactic acid. Also, the iron-deficient animals had a significantly higher rate of glucose turnover than controls with normal iron status, showing more metabolite recycling via the Cori cycle (i.e. lactate conversion to glucose). Submaximal endurance exercise produced significantly greater blood lactate and glucose turnover in the iron-deficient rats (Gregg et al., 1989).

The results of human studies examining the effects of iron deficiency without anemia on glucose and lactate metabolism during exercise are equivocal (Schoene et al., 1983; Rowland et al., 1989; Lukaski et al., 1991). Schoene et al. (1983) studied the effect of 2 weeks of iron treatment on exercise performance in trained, iron-deficient female athletes. Exercise performance remained unchanged, but blood lactate levels at maximum exercise decreased significantly. Lukaski et al. (1991) found iron deficiency to be associated with an increase in peak carbon dioxide production, RER and blood lactate, as well as a decrease in total oxygen uptake and aerobic EE. Treadmill endurance times significantly improved after iron supplementation, but differences in submaximal oxygen consumption or ventilation were not found (Schoene et al., 1983; Klingshirn et al., 1992). The inconsistent results from human studies may be attributed to the range of iron supplement doses and treatment times, different tests of endurance performance, large variation in exercise training status of subjects and wide range of study sample sizes. Discrepancies between animal and human studies may relate to experimental design. Animal studies allow for greater control of iron status via feeding of iron-deficient diets and phlebotomy. Moreover, animal studies are conducive to direct measurement of functional and storage iron. Thus, the negative findings in some human studies may be a result of indirect measures of iron status.

We previously reported that 4 weeks of iron supplementation, concurrent with aerobic training, improved performance in a 15-km time trial by 10% (Hinton et al., 2000). In addition, iron-supplemented subjects worked at a significantly lower percentage of their VO2peak and at a significantly higher work rate. These results indicate that iron supplementation of deficient subjects increases work capacity for a given exercise intensity relative to VO2peak.

In the present study, iron supplementation also resulted in an increase in gross energetic efficiency to do work (Table 5). As EE did not change from pre- to post-supplementation, the increased work output at the same relative input resulted in an increase in gross energetic efficiency for the iron group. Improved energetic efficiency after iron supplementation has also been reported by Zhu and Haas (1998b). In their study of untrained women, increased efficiency was attributed to a decrease in EE at the same work rate.

During the 6 weeks, supplement compliance among subjects in iron and placebo groups was high, 98 and 99% respectively. Therefore, improvements in iron status were attributed to the iron supplement and not to dietary modification by subjects as verified by pre- and post-supplementation diet records. Daily training logs kept by the subjects showed that the average amount of energy expended per week in exercise did not change during the course of the study and did not differ between groups at any time point (Table 2). Energy expended during training was not correlated with any exercise performance outcome (data not shown). Therefore, changes in any exercise performance variable from pre- to post-testing cannot be explained by an exercise training effect.

In conclusion, 6 weeks of iron supplementation significantly increased sFer and endurance capacity in trained men and women. Iron supplementation preserved aerobic function by maintaining VT during maximal exercise and by increasing gross energetic efficiency during submaximal work compared to placebo treatment. Increases in sFer were also correlated with reductions in RER during submaximal exercise. These results suggest that even marginal repletion of iron stores in iron-deficient individuals either improves or maintains aerobic function. Future studies examining changes in iron-dependent enzymes and proteins in skeletal muscle in response to iron supplementation of iron-deficient nonanemic individuals are warranted.


  1. , , , , (1993). Compendium of physical activities: classification of energy costs of human physical activities. Med Sci Sports Exerc 25, 71–80.

  2. , , (1995). The frequency of anemia and iron deficiency in the runner. Med Sci Sports Exerc 45, 27–35.

  3. (1996). Assessment of iron status. Clin Biochem 29, 209–215.

  4. , , , , (2005). Day-to-day variation in iron-status measures in young iron-deplete women. Br J Nutr 94, 551–556.

  5. , , (1986). The effects of fitness-type exercise on iron status in adult women. Am J Clin Nutr 43, 456–463.

  6. , , (1984). Increased glucose dependence in resting, iron-deficient rats. Am J Physiol 253, E641–E666.

  7. , , , , (2001). Marginal iron deficiency without anemia impairs aerobic adaptation among previously untrained women. Am J Clin Nutr 75, 734–742.

  8. , , , (2004). Tissue iron deficiency without anemia impairs adaptation in endurance capacity after aerobic training in previously untrained women. Am J Clin Nutr 79, 437–443.

  9. , , , , , (2003). Iron supplementation improves progressive fatigue resistance during dynamic knee extensor exercise in iron-depleted, nonanemic women. Am J Clin Nutr 277, 441–448.

  10. , (1986). Anemia causes a relative decrease in blood lactate concentration during exercise. Eur J Appl Physiol 55, 74–78.

  11. , , , , (1988). Effects of iron deficiency on endurance and muscle enzyme activity in man. Med Sci Sports Exerc 18, 156–161.

  12. , , , (1999). Anaemia and iron deficiency in athletes. Sports Med 27, 229–240.

  13. , , , , (2000). Iron status of highly active adolescents: evidence of depleted iron stores in gymnasts. Int J Sport Nutr Exerc Metab 10, 62–70.

  14. , , , , , (1984). Distinguishing effects of anemia and muscle iron deficiency on exercise bioenergetics in rats. Am J Physiol 246, E535–E543.

  15. , , , , (1982). Muscle mitochondrial bioenergetics, oxygen supply, and work capacity during dietary iron deficiency and repletion. Am J Physiol 242, E418–E427.

  16. , (1935). Spectrophotometric studies. II. Preparations from washed blood cells; nitric oxide hemoglobin and sulfhemoglobin. J Biol Chem 112, 51–65.

  17. , (2004). Prevalence of iron depletion and anemia in top-level basketball players. Int J Sport Nutr Exerc Metab 14, 30–37.

  18. , (2001). Monitoring of erythropoiesis by the serum transferrin receptor and erythropoietin. Acta Clin Belg 56, 146–154.

  19. , , , , , (1976). Iron deficiency in the rat: physiological and biochemical studies of muscle dysfunction. J Clin Invest 58, 447–453.

  20. , , , , , (1986). A serum ferritin assay for prevalence studies of iron deficiency. Am J Hematol 23, 141–151.

  21. Food and Nutrition Board, Institute of Medicine (2001). Dietary Reference Intakes for Vitamin A, Vitamin K, Arsenic, Boron, Chromium, Copper, Iodine, Iron, Manganese, Nickel, Silicon, Vanadium, and Zinc. National Academy of Sciences: Washington, DC.

  22. , , , , (1977). Physical work capacity and metabolic stress in subjects with iron-deficient anemia. Am J Clin Nutr 30, 910–917.

  23. , , (1989). Acute anemia results in an increased glucose dependence during sustained exercise. J Appl Physiol 66, 1874–1880.

  24. , , , (2000). Iron supplementation improves endurance after training in iron-depleted, non-anemic women. J Appl Physiol 88, 1103–1111.

  25. , , , (2002). No effect of menstrual cycle phase on glucose kinetics and fuel oxidation during moderate-intensity exercise. Am J Physiol Endocrinol Metab 282, E752–E762.

  26. , (1978). Generalized equations for predicting body density of men. Br J Nutr 40, 497–504.

  27. , (1980). Generalized equations for predicting body density of women. Med Sci Sports Exerc 12, 175–182.

  28. , , , , (1992). Effect of iron supplementation on endurance capacity in iron-depleted female runners. Med Sci Sports Exerc 24, 819–824.

  29. , (1993). Effects of iron repletion on VO2peak, endurance, and blood lactate in women. Med Sci Sports Exerc 25, 1386–1392.

  30. , , (1988). Anthropometric Standardization Reference Manual (abridged edn). Human Kinetics Books: Champaign, IL.

  31. , , (2002). Iron deficiency – United States, 1999–2000. Morb Mortal Wkly Rep 51, 897–899.

  32. , , (1991). Altered metabolic response in iron-deficient women during graded, maximal exercise. Eur J Appl Physiol 63, 140–145.

  33. , , (1984). Iron metabolism and ‘sports anemia’, a hematological comparison of elite runners and control subjects. Acta Med Scand 216, 157–164.

  34. , , (2000). Iron status in female endurance athletes and in non-athletes. Int J Sport Nutr Exerc Metab 1, 260–276.

  35. , , , (2002). The assessment of frequency of iron deficiency in athletes from the transferrin receptor-ferritin index. Int J Sport Nutr Exerc Metab 11, 42–52.

  36. , (1970). Physical work capacity and maximum oxygen uptake in treadmill and bicycle exercise. Med Sci Sports 2, 118–123.

  37. , , , , , (2001). Resistance training affects iron status in older men and women. Int J Sport Nutr Exerc Metab 11, 287–298.

  38. , , , (1989). The effects of prelatent/latent iron deficiency on physical work capacity. Med Sci Sports Exerc 21, 263–268.

  39. , , (1971). Determination of serum iron and latent iron-binding capacity (LIBC). Clin Chem Acta 35, 91–93.

  40. , , (1984). Precision of ventilatory and gas exchange alterations as a predictor of the anaerobic threshold. Eur J Appl Physiol Occup Physiol 52, 173–177.

  41. , , (1997). Serum transferrin receptor and its ration to serum ferritin in the diagnosis of iron deficiency. Blood 89, 1052–1057.

  42. , , (1998). Serum transferrin receptor, ferritin and Tfr-F index in identification of latent iron deficiency. Eur J Haematol 60, 135–137.

  43. , , (1979). Determination of anaerobic threshold by the ventilation equivalent in normal individuals. Respiration 38, 36–42.

  44. , , , (1989). The effect of iron therapy on the exercise capacity of nonanemic iron-deficient adolescent runners. Am J Dis Child 142, 165–169.

  45. SAS Institute (1999). Statistical Analyses System, version 82. SAS Institute Inc.: Cary, NC.

  46. , , , , , (1983). Iron repletion decreases maximal exercise lactate concentrations in female athletes with minimal iron-deficiency anemia. J Lab Clin Med 102, 306–312.

  47. , (2005). Prevalence of iron deficiency with and without anemia in recreationally-active men and women. J Am Diet Assoc 105, 975–978.

  48. (1956). The gross composition of the body. Adv Biol Med Phys 4, 239–280.

  49. , , , (1998). Serum transferrin receptor and transferrin receptor-ferritin index identify healthy subjects with subclinical iron deficits. Blood 92, 2934–2939.

  50. , , , (1987). Effects of iron deficiency and training on mitochondrial enzymes in skeletal muscle. J Appl Physiol 62, 2442–2446.

  51. , (1997). Iron depletion without anemia and physical performance in young women. Am J Clin Nutr 66, 334–341.

  52. , (1998a). Response of serum transferrin receptor to iron supplementation in iron-depleted, nonanemic women. Am J Clin Nutr 67, 271–275.

  53. , (1998b). Altered metabolic response of iron-depleted nonanemic women during a 15-km time trial. J Appl Physiol 84, 1768–1775.

Download references


We gratefully acknowledge the expertise of Dr Tom Thomas and Dr Rich Cox. We thank Scott Rector for his assistance with the phlebotomy and exercise testing and Denise Vultee for her help in preparing this manuscript.

Author information


  1. Department of Nutritional Sciences, University of Missouri, Columbia, MO, USA

    • P S Hinton
    •  & L M Sinclair


  1. Search for P S Hinton in:

  2. Search for L M Sinclair in:

Corresponding author

Correspondence to P S Hinton.

About this article

Publication history







The authors have no conflicts of interest.