| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
Research Centre for Gender and Health (A.J.P., W.J.B.), The University of Queensland, QLD, AUSTRALIA
Discipline of Nutrition and Dietetics, Faculty of Medicine and Health Sciences (D.C.K.R.), The University of Queensland, QLD, AUSTRALIA
School of Human Movement Studies (W.J.B.), The University of Queensland, QLD, AUSTRALIA
Address reprint requests to: Amanda Patterson, Ph.D., Research Centre for Gender and Health, The University of Newcastle, University Drive, Callaghan, New South Wales 2308, AUSTRALIA. E-mail: whajp{at}alinga.newcastle.edu.au
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Design: Randomised controlled trial to compare supplement and dietary treatment of iron deficiency.
Subjects: 44 iron deficient (serum ferritin <15 µg/L or serum ferritin 15–20 µg/L, plus two of the following: serum iron <10 µmol/L, total iron binding capacity >68 µmol/L or transferrin saturation <15%) and 22 iron replete (hemoglobin 
120 g/L and serum ferritin >20 µg/L) women 18 to 50 years of age were matched for age and parity.
Interventions: Iron deficient women were randomly allocated to either iron supplementation or a high iron diet for 12 weeks.
Measures of Outcome: Iron deficient and iron replete participants had iron studies performed and completed the Piper Fatigue Scale (PFS) and the SF-36 general health and well-being questionnaire at baseline (T0), following the 12 week intervention (T1) and again after a six-month non-intervention phase (T2). The SF-36 includes measures of physical (PCS) and mental (MCS) health and vitality (VT).
Results: MCS and VT scores were lower and PFS scores were higher for iron deficient women (diet and supplement groups) than iron replete women at baseline. Both intervention groups showed similar improvements in MCS, VT and PFS scores during the intervention phase, but mean increases in serum ferritin were greater in the supplement than the diet group. PCS scores were not related to iron status.
Conclusions: Treatment of iron deficiency with either supplementation or a high iron diet results in improved mental health and decreased fatigue among women of childbearing age.
Key words: iron deficiency, women, fatigue, mental health, diet therapy, supplement
Abbreviations: ALSWH=Australian Longitudinal Study of Women’s Health • GHQ=General Health Questionnaire • MCS=Mental Component Summary score of the SF-36 • PCS=Physical Component Summary score of the SF-36 • PFS=Piper Fatigue Scale • SF-36=SF-36 general health and well-being questionnaire • T0=baseline or pre-intervention • T1=post-intervention • T2=follow-up • VT=Vitality subscale of the SF-36
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
There is a widely held belief that iron deficiency affects general health by causing chronic tiredness. Several studies have examined associations between nonspecific symptoms such as tiredness, weakness, headache, breathlessness, dizziness, irritability and the like and iron deficiency. For example Ballin et al. [7] and Beutler et al. [8] reported improvements in symptoms, including tiredness, after iron supplementation in adolescent girls and iron deficient women. However, other studies such as those by Elwood and Hughes [9] and Morrow et al. [10] do not support these findings.
The physiological effects of iron deficiency are likely to have a broader impact on general health and well-being than just fatigue. To our knowledge, there have only been two studies which have directly examined the effects of iron deficiency on health and well-being using a validated scale. Fordy and Benton [11] and Rangan et al. [12] used the General Health Questionnaire (GHQ), which is a mental health assessment tool. Fordy and Benton [11] found no association between serum ferritin levels and GHQ scores for 297 male and female students. Rangan et al. [12] also found no association between serum ferritin and GHQ scores for female university students, but anemic subjects in this study did report more psychological distress.
The Australian Longitudinal Study of Women’s Health (ALSWH) recently provided the opportunity to examine the relationship between self-reported iron deficiency (based on information from their physicians) and general health and well-being in a large sample of women. Between baseline in 1996 and follow-up in 1998, 795 women out of 8,869 women 45 to 50 years of age in 1996 reported having had a diagnosis of iron deficiency. Decreases in physical health, mental health and vitality scores, as measured by the SF-36 general health and well-being questionnaire, were more marked among this group of women than among those who were not diagnosed with iron deficiency during this period [13].
In order to understand the public health impact of iron deficiency, it is important that its full effects on general health, well-being and daily functioning are systematically investigated. In this study, the effects of iron deficiency and its treatment on general health, well-being and tiredness were examined in women 18 to 50 years of age, who were participating in a randomised controlled trial to compare the efficacy of dietary and supplement treatment on iron deficiency.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Sample-size calculations for the overall study design were based on observation of a 40% increase in serum ferritin (4 µg/L based on an estimated average serum ferritin of 10 µg/L and standard deviation of the change in serum ferritin of 6 µg/L) for the dietary intervention group. Eighteen subjects per group were determined adequate to provide power of 80% to detect a change of 4 µg/L at a 95% confidence level, but a target of 25 per group allowed for withdrawals.
Eligibility criteria for inclusion of women in the study were no major illness, childbearing age, menstruation, non-pregnant, no hysterectomy, 18 years of age or over (or 16 to 18 years of age with parental consent), hemoglobin 90 g/L. Iron deficiency was defined as either serum ferritin <15 µg/L or serum ferritin 15–20 µg/L with two other hematological parameters indicative of iron deficiency (e.g., serum iron <10 µmol/L, total iron binding capacity >68 µmol/L, transferrin saturation <15%). Hematological criteria for inclusion in the iron replete group were hemoglobin 120 g/L and serum ferritin >20 µg/L.
After confirmation of iron status, each woman completed the SF-36 general health and well-being questionnaire and the Piper Fatigue Scale.
Treatment
Women with iron deficiency were randomly allocated within their age (<20, 20–29, 30–39, 40–50 years) and parity (0, 1–2, 3–4 children) category to one of the two treatment groups (diet or supplements) and counseled individually on the treatment protocol.
Participants allocated to the supplement group were asked to take a 350 mg ferrous sulphate supplement (equivalent to 105 mg of inorganic iron) on an empty stomach, daily for 12 weeks. Compliance was measured using a diary system.
Participants allocated to the diet group were asked to follow a high iron diet for the 12 week intervention period. The dietary intervention was designed to provide approximately the recommended daily intake of absorbed iron (2.25 mg per day). Details of the dietary intervention have been reported elsewhere [14]. Participants were given "meat vouchers" with which to purchase lean beef or lamb during the intervention and received counseling from a dietitian on the diet regime. Compliance was measured by recording consumption of iron containing foods and enhancers of iron absorption for three of the twelve weeks of the intervention (weeks 1, 5 and 9).
Blood samples were collected from all participants at baseline (T0) after 12 weeks of the intervention (T1) and again after a further six-month follow-up period (T2). The SF-36 and Piper Fatigue Scale were also repeated at T1 and T2.
The study protocol was approved by The University of Newcastle’s Human Ethics Committee.
Assessment of General Health and Well-Being
The SF-36 general health and well-being questionnaire is a brief questionnaire that can be self-administered in five to ten minutes. It contains 36 items scored as eight multi-item scales plus a one item measure of self-evaluated change in health status. The eight scales are Physical Functioning, Role Physical, Bodily Pain, General Health, Vitality, Social Functioning, Role-emotional and Mental Health. These scales combine further to produce two summary scores: the Physical Component Summary score (PCS) and the Mental Component Summary score (MCS) [15]. Published reliability statistics for the SF-36 have exceeded recommended standards for measures used in group comparisons. In most studies, reliability coefficients equal or exceed 0.80 [16].
As most of the subscales have significant ceiling effects, the analysis here was restricted to consideration of changes in the two summary scores, which are normally distributed, and in the Vitality subscale score (VT), which is also normally distributed and is directly related to the a priori hypothesis relating to iron deficiency and fatigue. The summary scores were compared with norms for the reference population (women 18 to 54 years of age from the 1995 Australian National Health Survey [17]) such that the population average is set at 50. A score below 50 indicates worse physical or mental health than the reference population, while a score above 50 indicates better health than the reference population [15]. Vitality scores are not standardised, and the mean score for women 18 to 54 years of age from the 1995 Australian National Health Survey was 63.3 [17].
Assessment of Fatigue
The Piper Fatigue Scale (PFS) was designed as a self-administered research instrument to measure subjective fatigue patterns in a variety of populations [18].
In its current form (1995), the PFS comprises 22 items which combine to produce four subscales, each measuring a different dimension of subjective fatigue (Behavioral/Severity, Affective Meaning, Sensory and Cognitive/Mood). The 22 items are scored on a numerical scale from "0" to "10," and component items are averaged to calculate the four subscales and a Total Fatigue Score [19].
The current version of the PFS has undergone extensive testing and has been shown to have good reliability and validity [19]. Concurrent validity estimates have been determined by correlating the subscales with the mood disturbance scores of the Profile of Mood States and with the Fatigue Symptom Checklist subscales and Total Fatigue Scores. Correlations between these tests range from 0.31 to 0.57 [19].
Data Analysis
Linear mixed models were used to investigate differences in variables between the three groups (control, diet and supplement) at baseline (T0), following the 12-week intervention (T1) and after the six-month follow-up (T2). This type of model allowed the correlation between observations for each individual to be taken into account, so that the differences between each group at each time period could be tested. Eight hypotheses were used to test whether there were differences between the control, diet and supplement groups at T0 and T2, whether there were changes over time for any group and whether there were differences between the diet and supplement groups at any of the time points. As the procedure involved eight simultaneous comparisons, a significance level of p<0.00625 was used (ie. p=0.05 divided by 8) [20].
Because serum ferritin values were not normally distributed, the natural log of serum ferritin was used for statistical analyses.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Mean serum ferritin at T0 was significantly higher for the control group than for the diet and supplement groups (Table 1) and did not change significantly over the duration of the trial. While baseline values for serum ferritin changed over time in both intervention groups, the patterns of change were different, with a more marked increase in the supplement group during the first three months than in the diet group.
|
In summary, iron status improved for both the diet and supplement groups over the duration of the intervention trial, while values for the control group did not change.
General Health and Well-Being (SF-36)
Mean Vitality scores (VT) for the diet and supplement groups were very much lower at baseline than the mean for the control group (Table 2). Scores for both the diet and supplement groups increased significantly over time such that, at T2, there were no longer any differences between the three groups.
|
Mean MCS scores for the diet and supplement groups were similar to each other at baseline (T0), but significantly lower than the mean for the control group. No difference could be detected between mean scores for the diet and supplement groups at T1, nor for the three groups at T2; however, a significant increase was observed for the diet group, but not the supplement group, over time (Table 2).
Piper Fatigue Scale (PFS)
Mean values for the four subscale scores and the Total Fatigue Score were significantly higher at baseline for the two intervention groups than for the control group (Table 3). During the intervention these scores improved in both the diet and the supplement groups, but did not change over time for the control group. These improvements were statistically significant for the Behavioral/Severity, Affective Meaning and Total Fatigue Scores in the supplement group and for the Sensory score in the diet group.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
It was expected that changes in general health measures would mirror changes in iron status, but this was not the case. SF-36 and PFS scores for the diet intervention group improved in a similar manner to those seen in the supplement intervention group, despite greater improvement in serum ferritin in the latter. This might represent a psychological advantage of dietary treatment over supplement treatment in terms of locus of control or a perception that diet is the healthier and more natural treatment option. Alternatively, the advantage for the diet over the supplement may lie with the diet itself because, although the diet was designed to improve dietary iron intake and iron absorption, it is impossible to alter one dietary component in isolation.
In contrast to the results of Rangan et al. [12], who reported an association between GHQ scores and iron status only for anemic women, the relationships between serum ferritin and indicators of mental health, vitality and fatigue at baseline in this study were not affected by the exclusion of anemic women from the analyses. However, the improvements in VT in the supplement group and in PFS scores in both intervention groups were considerably smaller and not statistically significant when women with anemia were excluded from the analyses. The improvements in MCS in both groups and in VT in the diet group appeared to be independent of the baseline severity of iron deficiency.
General Health and Well-Being
Baseline PCS scores for the control, diet and supplement groups were not significantly different from each other and were close to 50. However, while the mean MCS score for the iron replete control group was close to the comparison population mean of 50 at baseline, MCS scores for the iron deficient women in both intervention groups were significantly lower (40.2 and 43.0). VT scores for the iron deficient women were also significantly lower than the mean for women 18 to 54 years of age from the 1995 Australian National Health Survey (63.3 cf. 38.8 and 45.7) [17].
The relative burden of iron deficiency can be demonstrated by comparing these findings with MCS and VT scores for women from the ALSWH who report various chronic conditions and symptoms. For example, mean MCS and VT scores for ALSWH women with diabetes, heart disease or osteoporosis were higher (MCS about 42–46 and VT about 46–51) than those for the iron deficient women in this study. However, ALSWH participants who reported chronic migraines and painful joints had similar MCS and VT scores (MCS about 40–42 and VT about 44–48) to those found in this sample of iron deficient women [21].
The mechanism by which iron deficiency may affect mental health is unknown [12]; however, iron has important roles in the functioning of several neurotransmitters, including dopamine, serotonin and catecholamines [22]. The concept of vitality encompasses both physical and mental dimensions, and the effects of iron deficiency are likely to result from the combination of mental health and physical performance effects. Physical performance has been shown to be reduced in iron deficiency as a result of inadequate cellular energy production (iron is a critical component of many mitochondrial enzymes) and the effects on oxygen transport when hemoglobin is reduced in iron deficiency anemia.
Fatigue
To our knowledge, this is the only study to date which has investigated the role of fatigue in iron deficiency using a multi-item validated scale. While the results support the concept of increased fatigue in iron deficiency, it is difficult to determine the relative burden of this fatigue. The Piper Fatigue Scale (PFS) was designed to measure subjective fatigue patterns in a variety of populations but, due to the research interests of the developers, has predominantly been used with cancer patients. Mean Total Fatigue Scores for the control, diet and supplement groups at baseline were 2.6, 4.2 and 4.7, respectively. Corresponding means for patients with breast cancer, multiple types of cancer and pregnant women were 4.5, 4.4 and 3.6, respectively [18,19]. Thus, the degree of fatigue experienced by the iron deficient women was similar to that for cancer patients and greater than that for pregnant women up to the 40th week of gestation.
Previous randomised placebo controlled studies have used a subjective one-item measure to explore the role of chronic tiredness (or fatigue) in non-anemic iron deficiency. Beutler et al. [8] found that women who were iron deficient showed more symptomatic improvement with iron therapy than with placebo, while women with normal iron stores showed similar improvements with both iron and placebo treatments. However, Morrow et al. [10] found no significant difference between iron and placebo in producing improvement in tiredness.
One limitation of the baseline fatigue data from this study is that the women were aware of their iron status at the time that they completed the PFS. It is probable that most of the women were aware of the public perception that iron deficiency causes fatigue, and this may have influenced their responses. For this reason, it was envisaged that the Vitality subscale of the SF-36 would act as a check for the PFS scores, as fatigue and vitality are really observations of the opposite poles of the same dimension. The SF-36 Vitality subscale scores are unlikely to be affected in the same way as the PFS scores, as it would be more difficult for anyone completing the SF-36 to identify which items were directly assessing vitality, if indeed they were aware that vitality was being assessed. As the baseline PFS data correlated strongly with the baseline results for the SF-36 Vitality subscale (R=-0.7), it is concluded that the baseline PFS scores represent a reasonable measure of subjective fatigue.
In contrast with the baseline assessments, completion of the PFS at T1 and T2 was done prior to the results of individual iron tests being divulged. Therefore, these results are unlikely to be biased to the same extent, although it might be assumed by participants that their iron status had improved due to the treatment and that their fatigue levels should drop. However, once again there was strong agreement between PFS and SF-36 Vitality subscale results, suggesting that the changes were not solely due to this belief. In fact, the increase in Vitality and simultaneous decrease in PFS scores after the intervention could in part be considered as validation of the measurements. However, unlike the Vitality scores, the PFS scores did not return to the level of the controls by T2. This may be due to a time-delay effect, the fact that a number of participants remained iron deficient at the conclusion of the study (11 and 3 for the diet and supplement groups, respectively) or may again be a psychological consequence of the public perception of a link between iron deficiency and fatigue.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Similar improvements in SF-36 and PFS scores for the diet and supplement intervention groups, despite greater improvements in serum ferritin in the latter, may suggest advantages in terms of general health and well-being for the use of dietary intervention rather than supplementation in the treatment of iron deficiency. Notwithstanding, confirmation that improvements in general health were in fact the result of increases in serum ferritin, and not merely due to the psychological impact of being studied or changes in dietary patterns due to a heightened interest in nutrition, is required. These results should therefore be reproduced in a randomised controlled trial of iron supplements and placebo, using a cohort who are naïve to their iron status.
|
ACKNOWLEDGMENTS |
|---|
Received September 17, 2000. Accepted May 24, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. E Murray-Kolb and J. L Beard Iron deficiency and child and maternal health Am. J. Clinical Nutrition, March 1, 2009; 89(3): 946S - 950S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M.-A. Grondin, M. Ruivard, A. Perreve, H. Derumeaux-Burel, I. Perthus, J. Roblin, F. Thiollieres, and L. Gerbaud Prevalence of Iron Deficiency and Health-related Quality of Life among Female Students J. Am. Coll. Nutr., April 1, 2008; 27(2): 337 - 341. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. J Zhou, R. A Gibson, C. A Crowther, P. Baghurst, and M. Makrides Effect of iron supplementation during pregnancy on the intelligence quotient and behavior of children at 4 y of age: long-term follow-up of a randomized controlled trial Am. J. Clinical Nutrition, May 1, 2006; 83(5): 1112 - 1117. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F Verdon, B Burnand, C-L F. Stubi, C Bonard, M Graff, A Michaud, T Bischoff, M de Vevey, J-P Studer, L Herzig, et al. Iron supplementation for unexplained fatigue in non-anaemic women: double blind randomised placebo controlled trial BMJ, May 22, 2003; 326(7399): 1124. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
