Always tired with low ferritin — will iron help?
The tiredness perhaps: in non-anaemic iron-deficient adults iron eased self-reported fatigue (4 trials, all women, SMD −0.38) but not VO2max or timed performance (10 trials).
Main caveat: Women. Healthy women in 15 of 18 trials; the fatigue meta-analysis is women only, of uncharacterised athletic status, two of its four trials from clinics.
Houston and colleagues asked whether iron helps fatigue and physical capacity in adults who are iron-deficient but not anaemic.
They found 18 randomised placebo-controlled trials (1170 participants, 1989-2015), almost all in healthy women aged 17-55, using oral ferrous sulphate (13 trials; mean 86.9 mg of elemental iron a day), intravenous iron (3 trials) or intramuscular iron (1 trial) for a mean of 46 days.
Iron reduced self-reported fatigue (SMD −0.38, 95% CI −0.52 to −0.23; 4 trials, 714 women; I² 0%), with moderate certainty, and the effect was seen with both oral and intravenous iron.
It did not improve VO2max (SMD 0.11, −0.15 to 0.37; 9 trials, 235 participants), 15-km time trials (SMD −0.09; 2 trials) or time to exhaustion (SMD 0.25, −0.22 to 0.73; 4 trials), all low certainty. Haemoglobin rose 4.01 g/L and ferritin 9.23 (printed as µmol/L).
The authors conclude that iron eases subjective fatigue but not objective capacity, and that patients and practitioners could consider iron-rich foods or supplements for fatigue "in the absence of documented anaemia".
Our reading: the honest expectation for an athlete who is always tired with a low ferritin — tiredness may ease, a performance gain is not shown — but the fatigue evidence is from women attending clinics, not athletes, and blinding of oral iron is imperfect.
Protocol
Search: Medline, Embase, CENTRAL, CINAHL, SPORTDiscus and CAB Abstracts from inception to 31 October 2016, trial registries, forward citation searches; English only; 11 580 citations screened in duplicate.
Eligibility: RCTs of adults (18 and over) with iron deficiency without anaemia; oral, intramuscular or intravenous iron at any dose; placebo or active comparator; outcomes at least 28 days after starting.
Participants: n = 1170 (598 iron, 572 control) in 18 trials; healthy women in all but three trials; trial populations ranged from women with fatigue from clinics (Vaucher, Verdon), premenopausal women with fatigue (Favrat, Krayenbuehl, intravenous iron), physically active or untrained women, female athletes and endurance runners, to university endurance runners of both sexes.
Interventions: oral ferrous sulphate in 13 trials (713 participants; mean 86.9 ± 49.1 mg elemental iron a day, range 16-200), intravenous iron in 3 (395; mean total 767 mg), intramuscular in 1 (16); therapy 1 to 112 days (mean 46), follow-up 28 to 112 days (mean 57).
Outcomes: primary — self-reported fatigue and objective physical capacity (VO2max, time trial, time to exhaustion); secondary — anaemia, haemoglobin, ferritin, adverse events.
Analysis: RevMan, random effects, SMD or mean difference; subgroups by sex, athletic status, route, duration, follow-up and risk of bias; GRADE for primary outcomes.
Risk of bias: one trial low risk; the rest unclear (mostly unclear randomisation); blinding of participants adequate in 13 of 18.
Funding and interests: "This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors"; competing interests: none declared.
Key Results
| Outcome | Trials (participants) | Effect (95% CI) | I² | GRADE |
|---|---|---|---|---|
| Self-reported fatigue | 4 (714) | SMD −0.38 (−0.52 to −0.23) | 0% | moderate |
| 15-km time trial | 2 (79) | SMD −0.09 (−0.53 to 0.35) | 0% | low |
| Time to exhaustion | 4 (69) | SMD 0.25 (−0.22 to 0.73) | 0% | low |
| VO2max | 9 (235) | SMD 0.11 (−0.15 to 0.37) | 0% | low |
| Haemoglobin | 12 (298) | MD 4.01 g/L (1.22 to 6.81) | 48% | — |
| Ferritin | 14 (616) | MD 9.23 (6.48 to 11.97), unit printed µmol/L | 58% | — |
The fatigue trials: four trials (two oral at 80 mg a day, two intravenous) in women, scored with four different fatigue scales; in the intravenous iron sucrose trial (Krayenbuehl), fatigue did not differ overall at 12 weeks but improved in the subgroup with ferritin ≤15 ng/mL or transferrin saturation ≤20%.
Other capacity measures: one trial (20 participants) found less decline in knee-extension force after 6 min of exercise with iron; 3 of 16 other capacity measures improved significantly.
Subgroups: route of administration and follow-up did not change the fatigue effect; athletic status and route did not change the capacity results; sex and risk of bias could not be tested (all fatigue and capacity trials in women, all of unclear risk).
Adverse events: sparsely reported; nausea increased in the two intravenous trials, GI intolerance in the intramuscular trial, not in the oral trials that reported it; adherence did not differ (RR 1.0).
Units: ferritin is reported as µmol/L in the abstract and results, while Table 1 gives ferritin limits in µg/L; our reading is that µg/L was meant.
Mechanisms
Measured: fatigue scores fell and haemoglobin rose slightly with iron in women without anaemia; VO2max and timed performance did not change.
Authors' interpretation: whether the fatigue effect translates into quality of life, work or athletic performance is uncertain; unmasking by stool colour and GI effects could bias subjective fatigue, yet fatigue fell with oral and intravenous iron alike; follow-up (mean 57 days) may be too short for changes in muscle metabolism.
Authors' interpretation: earlier physiological experiments describing VO2max gains independent of haemoglobin (via muscle oxidative capacity) were not borne out in the pooled trials.
Inference: the one fatigue trial with a subgroup found benefit only with ferritin ≤15 ng/mL or low transferrin saturation, in line with the larger ferritin response from very low starts in REF-1409; the trials do not tell who will feel better.
Practical Application
Always tired with low ferritin — will iron help? The tiredness perhaps: in non-anaemic iron-deficient adults iron eased self-reported fatigue (4 trials, all women, SMD −0.38) but not VO2max or timed performance (10 trials).
Set the expectation honestly: fatigue may ease (moderate certainty, from women of uncharacterised athletic status); a capacity gain is not shown when there is no anaemia (low certainty, small trials). This is the source the base lacked behind the PHYS-36 card ("Always tired, ferritin 28").
Tiredness is not a diagnosis: not from this paper; the base's own practice: when signals persist after a lighter week, ask about illness, under-fuelling, iron, sleep and life stress before more rest or load (SYNTH-29), and screen for low energy availability (SYNTH-54); these trials enrolled people whose iron deficiency had been measured.
The authors' closing suggestion (iron-rich foods or supplements for fatigue without anaemia) applies to people with documented iron deficiency; not from this paper, the base's own practice: no iron without a blood test, with dose and schedule from a physician (SYNTH-06).
Against the other performance sources: REF-1413 (A-) pooled a VO2max gain in iron-deficient endurance athletes; REF-1409 (A-) found a non-significant trend; REF-1422 (B+, abstract) found VO2peak no different from placebo after IV iron in active women, while fatigue scores improved within the iron group (no between-group test reported) — consistent with this pattern.
Limitations
Women. Healthy women in 15 of 18 trials; the fatigue meta-analysis is women only, of uncharacterised athletic status, two of its four trials from clinics.
Small capacity trials. 235 participants across 9 VO2max trials; low certainty.
Risk of bias unclear in 17 of 18 trials; blinding of oral iron is hard (stool colour, GI effects).
Short follow-up (mean 57 days) and sparse reporting of adverse events.
Mixed routes and definitions. Oral, intravenous and intramuscular iron pooled; deficiency defined by each trial (maximum ferritin 15 to 50 µg/L).
Search to 2016, English only; newer athlete trials (REF-1422) are not included.
Units. The ferritin difference is printed in µmol/L.
Funding and interests: no specific grant; none declared.
How this entry was checked and changed
Why this level
a meta-analysis of placebo-controlled RCTs with a published protocol and GRADE (rubric A), read from the full text (Europe PMC, CC BY-NC 4.0). The fatigue result is moderate certainty with no heterogeneity; the physical-capacity nulls are low certainty (small trials, imprecision, unclear risk of bias in 17 of 18 trials)
Full text (checked 2026-10-28)
We included 18 unique trials and 2 companion papers enrolling 1170 patients (full text)
Iron supplementation was associated with reduced self-reported fatigue (SMD −0.38; 95% CI −0.52 to −0.23; I^2 0%; 4 trials; 714 participants) (full text)
Not associated with differences in maximal oxygen consumption (SMD 0.11; 95% CI −0.15 to 0.37; I^2 0%; 9 trials; 235 participants) and timed methods of exercise testing (full text)
15 km time trials: SMD −0.09; 95% CI −0.53 to 0.35 (2 trials, 79 participants); time to exhaustion: SMD 0.25; 95% CI −0.22 to 0.73 (4 trials, 69 participants) (full text)
Haemoglobin MD 4.01 g/L; 95% CI 1.22 to 6.81; I^2 48%; 12 trials; 298 participants; serum ferritin MD 9.23 µmol/L; 95% CI 6.48 to 11.97; I^2 58%; 14 trials; 616 participants (full text)
In oral iron trials the mean daily elemental iron dose was 86.9 mg (±49.1 mg; range: 16–200 mg) (full text)
In one trial fatigue improved in the subgroup with the lowest iron stores (ferritin ≤15 ng/mL or transferrin saturation ≤20%) (full text)
Healthy women comprised the study population in 15 of 18 included trials (full text)
Funding: no specific grant; Competing interests: None declared (full text)