Associations Between Lifetime Histories of Iron Deficiency, Anxiety, Depression and Multiple Pain Conditions: An Observational Study Using a Large-Scale National Database

Abstract

Purpose

Our aims were to investigate in adults, as we have shown in three paediatric samples, associations between a history of iron deficiency (whenever acquired and treated) and history (> 3 months) of primary, overlapping and multisite chronic pain conditions, and to investigate associations between history of iron deficiency, anxiety, and depression.

Participants and Methods

We recruited adult twins and their families from the Australian Twin Registry database. Responders addressed questionnaires about lifetime history of iron deficiency >3 months, diagnosed and treated by a physician, and a history >3 months of listed pain conditions, anxiety and depression. The emphasis was on life history, not current status, of each of these conditions and such questions were the only practical method. Logistic regression, using generalized estimating equation to account for correlations within twin families, was employed to assess relationships between iron deficiency and pain conditions and the psychological conditions.

Results

Of the 1519 adult responders, 1326 addressed the criteria for iron deficiency, and the prevalence was 38.7% in females and 2.4% in males. Iron deficiency as statistical predictor was significantly associated with migraine (Odds ratio (OR)=1.70, p=0.001), headache (OR=1.37, p=0.03), recurrent abdominal pain (OR=1.97, p<0.001), dysmenorrhea (OR=2.18, p<0.001), anxiety (OR=1.88, p<0.001) and depression (OR=2.06, p<0.001). Iron deficiency, anxiety and depression, as statistical predictors, were univariately and multivariately associated with the number of pain conditions (all p<0.05).

Conclusion

A history of iron deficiency in a community sample of adults was associated with most individual, primary, overlapping and multisite pain conditions tested, and multivariately associated with the number of pain conditions, and with anxiety and depression. We hypothesize, consistent with these results, that previous or current iron deficiency, even when treated, influences pain sensitivity and multiplicity directly and through known effects on anxiety and depression.

Introduction

In research with children and adolescents,^1–3 we have shown associations between a history of physician confirmed iron deficiency and multiple pain conditions. The pain conditions were pediatric forms of primary,⁴ overlapping,⁵ and multisite,⁶ chronic or recurrent pain. Iron deficiency has been reported to intensify acute pain,⁷ and to be associated with conditions or syndromes characterized mainly by single site or widespread chronic pain, including migraine,^8–11 headaches,^11,12 widespread pain,¹³ fibromyalgia syndrome,^14–17 chronic neck pain,¹⁸ and glossal pain.¹⁹ However, association between history of or current iron deficiency and multiple pain conditions has not been shown in adults. It is relevant to acknowledge the influence of iron deficiency on sleep impairment, anxiety and depressed mood,²⁰ which are risk factors for augmented and chronic pain. In turn, there are bidirectional associations between chronic pain and sleep impairment,²¹ anxiety,²² and depression,²³ (The latter three are also bidirectionally associated with each other). Thus, we have considered whether influence of a history of iron deficiency on pain is direct or may be mediated by a history of anxiety and/or depression.

Strong support for the relationships between iron deficiency and nociception and pain conditions has come from two publications based on animal experiments during neurodevelopment. Dowling et al showed that induced iron-deficiency sensitizes mice to acute pain stimuli and formalin-induced nociception.²⁴ Based on hot-plate reaction time, iron deficient mice had a lower acute pain threshold than control mice after 4 and 15 weeks. In addition, iron deficient mice had an increased chronic pain response compared with the control mice in the late phase of the formalin test after 1, 4, and 15 weeks of iron deficiency. This increased pain response was accompanied by an elevated expression of c-Fos immunoreactive cells at the ipsilateral dorsal horn, suggesting that iron deficiency indirectly increases cell activity at the spinal cord level. Yoo et al provided evidence that early-life iron deficiency evokes sex-independent effects on nociception in developing mice,²⁵ including an exacerbation of postsurgical pain during adulthood. The complex set of differences in adult pain behaviours shown in mice exposed to early-life iron deficiency was suggested as reflecting the multiple key roles that iron plays in the maturation of the peripheral and central nervous systems.²⁶

Yoo et al emphasized the importance of early life iron deficiency in altering nociception,²⁵ and our three published studies were in children and adolescents,^1–3 most of whom would have acquired ID in infancy or for females from menstruation. Iron deficiency particularly affects children younger than 5 years,²⁶ premenopausal (especially pregnant) women,²⁷ and people in low- and middle-income countries.²⁷ In the USA, 11% of children aged 6 months to 5 years have been estimated to have iron deficiency.²⁷ In less developed countries, the prevalence estimates in preschool children ranged from 11% to 26%.²⁸ In premenopausal non-pregnant women, the prevalence estimate of iron deficiency was 15% in the USA and 14–30% in less developed countries.²⁷ In the USA, 18% of pregnant women were estimated to have iron deficiency.²⁹ We anticipated that an adult sample would have lower associations between iron deficiency and pain conditions. The more profound effect of iron deficiency on pain in children and adolescents would be the result of early life influences on the developing nervous system while a substantial proportion of adults acquire iron deficiency post-development.

Aiming to confirm the associations between history of iron deficiency and multiple pain conditions in adults, we conducted a study of such associations using data acquired in a twin family study of adults with painless and painful phenotypes of restless legs syndrome (manuscript in preparation).³⁰ A history of iron deficiency (positive responses to 3 questions) was the most practical method to obtain life history for a large-scale cross-sectional survey. Serum ferritin was impractical and most subjects who had iron deficiency had been treated. A further aim of the current study was to investigate the inter-relationships between history of iron deficiency, anxiety, depression, and multiple overlapping (including primary) chronic pain conditions. We also tested for multiple associations between the overlapping (including primary) and multisite chronic pain conditions.

Methods

Data acquired for this study were obtained from a twin family project, administered by Twins Research Australia,³¹ entitled Contrasting Painful and Painless Restless Legs Phenotypes: An Adult Twin Family Study.³² Twin families are convenient and valuable epidemiological resources,^33,34 but we required to control for the analyses of twins and acknowledge the increased prevalence of iron deficiency in twins,³⁵ at least as newborns (it is not known whether children of twins retain iron deficiency risk into adult life). We recruited consenting adult twins and their families, including spouses and adult offspring, from the Australian Twin Registry database. The initial approach was by email, with the following contact attempts when there was no response: one Email reminder, one SMS reminder, one phone call reminder, and one final repeat email. Inclusion criteria were: twin adults between the ages of 30–55 years; children of included twins, where the offspring is aged 18–25; co-parents of the children of twins (30–55). Children 3–11 years and adolescents 12–17 years were included in the assessment of history of iron deficiency to provide a perspective on the age distribution of diagnosis of iron deficiency. They were not otherwise included in the analyses. Zygosity was not relevant to this study. The maximum accessible number of twin pairs aged 30–55 (and their families) in the Twin Registry were approached. Assuming a 15% response rate from previous experience of surveys of similar complexity, 1800 twin individuals were expected to participate in this study, and that number met power/sample size requirements for essential questions in the study.

Participants were requested to respond to demographic questions. Consistent with previous publications,^2,3 all responders were asked if they had a lifetime history of iron deficiency of >3 months duration. Supplementary questions were included to substantiate the categorization of history of iron deficiency: If yes, was it diagnosed by a doctor (physician)? Did your mother have iron deficiency treated by a doctor? Did your mother take this treatment when she was pregnant? Were you given iron tablets over 3 months, or given iron by injection? A history of iron deficiency (3+ months) was accepted if it was confirmed by a physician and if the treatment criterion was met.

All responders were asked if they had the following pain conditions for at least 3 months during their life, and whether a physician diagnosed it and whether it was treated, age span of relevant symptoms and whether it was still active. The qualifying questions were important for some conditions such as migraine for which self-diagnoses are common. We have generally used the term condition for simplicity and acknowledge that some conditions such as migraine are variously stated in the literature to be a condition, disorder, syndrome and disease. The painful conditions were in two categories: overlapping (including primary pain) pain conditions migraine, non-migraine headaches, recurrent abdominal pain including irritable bowel syndrome, dysmenorrhea, and growing pains (as a child) and chronic multisite pain conditions widespread pain and fibromyalgia syndrome, low back pain, other spinal pain, sciatica, and other chronic pain. Anxiety and depression responses were also considered positive by the same criteria (at least 3 months history at some time in the life course and diagnosed by a doctor) and were referred to as disorders. There was no suitable validated measure or questionnaire which covered 3+ months at any time across the lifespan.

The protocol was approved by the Sydney Children`s Hospitals Network Human Research Ethics Committee, HREC Reference 2019/ETH00023, and approved by The Director, Twins Research Australia.

Statistical Analyses

Summary statistics for all adults who met the criteria for a history of iron deficiency, pain conditions, and psychological diagnoses are presented by case numbers and percentages in Tables 1 and 2. Logistic regression, using generalized estimating equation (GEE) to account for correlations within families,³⁶ was employed as the estimation method to compare the difference between males and females for each of psychological and primary and chronic conditions. The same regression method was also used to assess relationships between the history of iron deficiency as predictor and each primary and chronic pain condition, as well as psychological conditions, as outcomes. Ordinal logistic regression was used to assess relationships between the number of overlapping (including primary) and multisite chronic pains in individuals (reflecting total burden of pain) as outcome and the history of deficiency, anxiety and depression as predictors. All analyses were conducted using commercial software STATA 16 software (http://www.stata.com).

Table 1.

Distribution of History of Iron Deficiency by Age and Gender

Age (Years)	All			Female			Male			Pa
	N	n	%	N	n	%	N	n	%
3–11	259	11	4.25	133	7	5.26	126	4	3.17	0.4043
12–17	232	22	9.48	100	17	17.0	132	5	3.79	0.0007
>17	1326	348	26.2	871	337	38.7	455	11	2.42	<0.0001

Notes: N = sample size; n = cases with history of iron deficiency; % = percent with history of iron deficiency; ^ap-value (p) compares equality in proportions between males and females. Note that all subsequent analyses involved subjects > 17 years.

Table 2.

Summary Statistics for Life Prevalence (>3 Months) of Psychological Disorders and Overlapping (Including Primary) and Chronic Pain Conditions, for All Adults and for Females and Males

	All (N = 1326)		Female (N = 871)		Male (N = 455)			p
	n	%	n	%	n		%
Psychological disorders
Anxiety	407	30.7	309	35.5	98		21.5	<0.001
Depression	320	24.1	240	27.6	80		17.6	<0.001
Primary and chronic pain conditions
Migraine	190	14.3	161	18.5	29		6.37	<0.001
Headache	301	22.7	236	27.1	65		14.3	<0.001
Recurrent abdominal pain	189	14.3	164	18.8	25		5.49	<0.001
Growing pains	93	7.01	66	7.58	27		5.93	0.236
Chronic pain (diverse)	49	3.70	39	4.48	10		2.20	0.020
Chronic Spinal Pain	80	6.03	59	6.77	21		4.62	0.130
Dysmenorrhea			117	13.4

Notes: n = number of cases; % = percentage; p = p-value for female male differences. *Recurrent.

Results

Of the 4019 twin families invited and who met all eligibility criteria, 1519 adults responded (955 twins, 342 co-parents and 222 adult children) of whom 1,326, aged 18.3–65.4 years, answered questions on their history of iron deficiency. Of the 1326 individuals available for the history of iron deficiency analyses, 172 were children of twins (≥18 years), 278 were co-parents and 876 were twin individuals.

Table 1 shows the distribution of history of iron deficiency by age and gender using all data, including children 3–11 years and adolescents 12–17 years. We emphasise that the study is about adults but that iron deficiency is not infrequently acquired in early life, hence the inclusion of young family members in the table to make the point. This is not an accurate determination of age when the iron deficiency was diagnosed, rather the percentage in the 3–11 and 12–17 and >17 (adult) age groups for whom there was a history of iron deficiency meeting all 3 criteria. The prevalence of history of iron deficiency in the total adult sample (>17 years) was 26.2%, with higher rates observed among females (38.6%) than males (2.4%; p <0.001).

Table 2 shows summary statistics for life prevalence (>3 months) of psychological conditions and overlapping (including primary) and chronic pain conditions, for all adults and for adult females and males. Female dominance was found for psychological conditions, primary pain conditions, and miscellaneous chronic pain conditions, except for growing pains (p = 0.236) and chronic spinal pain (p = 0.130).

In Table 3 we present results for associations between history of iron deficiency and overlapping (including primary) pain conditions, chronic pain conditions, and psychological conditions, adjusted for age whenever significant. A history of iron deficiency was significantly associated with the female dominated primary (overlapping) pain conditions of migraine, headache, recurrent abdominal pain, dysmenorrhea, chronic pain and with female dominated anxiety and depression. The initial categories of chronic pain, diverse and spinal, were combined because the original sample sizes were relatively small.

Table 3.

Association Between History of Iron Deficiency (Predictor) and Lifetime History of Each Overlapping (Including Primary) Condition, Chronic Pain Conditions and Psychological Disorders (N = 1326)

	OR	p	95% CI
Migraine	1.70	0.001	1.23–2.33
Headache	1.37	0.027	1.04–1.80
Recurrent abdominal pain	1.97	<0.001	1.44–2.71
Growing painsa	1.45	0.117	0.91–2.29
Dysmenorrhea (female)	2.18	<0.001	1.48–3.22
Chronic pain conditionsb	1.62	0.041	1.02–2.58
Anxietya	1.88	<0.001	1.47–2.41
Depression	2.06	<0.001	1.58–2.69

Notes: ^aAdjusted for age. Sample size for dysmenorrhea was 871. ^bCombining diverse and spinal chronic pain categories. All analyses were conducted using binary logistic regression, except the chronic pain conditions where we used ordinal logistic regression.

As shown in Table 4, history of iron deficiency, anxiety and depression as statistical predictors were all significantly associated with the number of pain conditions per individual as outcome. In both univariate and multivariate analyses, anxiety and depression had higher odds ratios (OR) than the history of ID. The highest OR was found for individuals who had both anxiety and depression (OR 4.01, 95% CI 2.94–5.45, p<0.001).

Table 4.

Association Between Number of Pain Conditions (0–6^a) in Individuals (Outcome) and History of Iron Deficiency, Anxiety, and Depression

	Univariate			Multivariate
	OR	p	95% CI	OR	p	95% CI
Iron deficiency	1.69	<0.001	(1.30, 2.20)	1.41	0.012	(1.08, 1.83)
Anxiety	3.12	<0.001	(2.41, 4.04)	2.06	<0.001	(1.54, 2.77)
Depression	3.45	<0.001	(2.62, 4.53)	2.28	<0.001	(1.67, 3.11)

Notes: Ordinal logistic regression was fitted using generalised estimating equation. ^aHistory (>3 months) of migraine, non-migraine headaches, recurrent abdominal pain, growing pains, diverse chronic pain conditions, low back pain.

The pain conditions were extensively associated with each other as determined by logistic regression, adjusted for age and gender, and most associations remained in the multivariate analysis (Supplementary Table S1). The most significant associations were for dysmenorrhea, chronic spinal pain and migraine, and least for growing pains which had a relatively small sample size.

Discussion

In this study we have found, in an adult community sample of twins and their adult family members, associations between a history of physician-confirmed and treated iron deficiency and primary,⁴ overlapping,⁵ and multisite pain conditions (as defined in adults),³⁶ also the number of pain conditions per individual. These results are consistent with results in our pediatric publications,^1–3 and with other publications showing iron deficiency-pain associations,^7–19 and with animal experiments.^24,25 We also showed significant associations between a history of iron deficiency as predictor and anxiety and depression which, in turn, were associated as predictor variables with the number of pain conditions per individual. These multiple associations, consistent with previous publications, strengthen the validity of our choosing history of iron deficiency, doctor diagnosed and treated, as the only practical method to survey life history.

As anticipated, the strength of associations in adults was less than in the pediatric samples,² probably because of the higher percentage of sampled young people who acquire iron deficiency in the first 5 years of life when iron deficiency impacts on neurodevelopment.^26,27 The age of acquisition and diagnosis of iron deficiency in responders reporting a history of iron deficiency could not be accurately determined in this survey, however, as shown in Table 1, 4.3% of responders in the 3–11 years age group (5.3% of females, 3.2% of males), 9.5% in the 12–17 years age group (17% of females, influenced by menstruation, and 3.8% of males) and 26.2% of adults >17 years (38.7% of females, 2.4% of males) reported a history of iron deficiency which was confirmed and treated. Namaste et al estimated 11% detection of low iron stores (ferritin) in preschool children in international samples.²⁸

We emphasize that all subjects with a history of iron deficiency had been treated with iron supplementation, although it could not be determined whether iron stores were satisfactorily repleted. The extent to which iron supplementation can overcome the neurocognitive influences of iron deficiency varies across published studies,^37–39 and may be minimal or partial. Our results are consistent with partial benefit from treatment at best. It is possible that our survey of history of iron deficiency across the life course may have underestimated the numbers who truly had iron deficiency. However, the numbers who responded positively to the 3-part question were consistent with published epidemiological studies and, if the numbers who had iron deficiency were in fact higher, the associations may have been stronger.

Iron deficiency, impaired sleep, anxiety, depression, and chronic pain are commonly co-morbid with each other,^20–23 just as we have found in this study (except for sleep impairment which was not included in this study). The extent to which iron deficiency directly (through the nervous system) causes increased pain sensitivity and vulnerability and the extent to which iron deficiency-influenced impairment of anxiety, and depression (also sleep) impact on pain sensitivity and vulnerability is complex and could not be determined using the cross-sectional design. The stronger associations between anxiety and depression and the number of pain conditions per individual than the association between history of iron deficiency and number of pain conditions suggest that the indirect influences are substantial.

Does Iron Deficiency Causally Influence Pain Sensitivity and Vulnerability?

There is no direct pathophysiological evidence in humans that iron deficiency causes alteration of nociception, pain sensitivity and vulnerability, but the animal experiments are persuasive.^24,25 We require, at this stage, to rely on causal reasoning or inference and draw on the Bradford Hill viewpoints for causal inference in associations in observational studies.⁴⁰ Evidence that iron deficiency causally influences pain sensitivity and vulnerability in this cross-sectional study in adults cannot be claimed, but the cumulative evidence of the three pediatric and 13 other publications in adult humans, applying the critical perspective of Rothman and Greenland to the Bradford Hill viewpoints,⁴¹ is persuasive. Potential confounders include sleep impairment, anxiety, depression,^2,3 all of which are associated with multiple chronic pain conditions and, in our and other studies,^20–23 with history of or confirmed iron deficiency (low iron stores) as statistical predictor.

Regarding the viewpoint of consistency, there are repeated observations of the association in different populations and under different circumstances in the total of 16 human publications.^1–3,7–19 For iron deficiency to be causative it must precede the nociception effects (temporality). This applies in the animal experiments,^24,25 and frequently does in humans, especially in the pediatric samples as discussed above.^1–3

Biological plausibility is an important viewpoint and is especially evidenced by the cited animal experiments.^24,25 These mouse experiments provide strong supportive evidence in favor of a causal relationship between iron deficiency, especially acquired in early life, and pain sensitivity and vulnerability. It is acknowledged that experiments with induced iron deficiency in rats using substantially different methods did not reach the same conclusions,^42,43 but they did not refute the findings of Dowling et al and of Yoo et al,^24,25 Experimental evidence in humans for iron deficiency causing increased pain sensitivity and vulnerability is lacking. The best model, quasi-experimental, would be a study in frequent blood donors but no such study has been published.

We conclude that a causal influence of iron deficiency, directly and indirectly, on pain sensitivity and on vulnerability to pain regions and conditions is an appropriate hypothesis to be further tested. Further testing should ideally include seeking clinical evidence for temporality and biological gradient, focus on potentially mediating variables especially impaired sleep, anxiety, depression and on multisensory sensitivity, and on neurobiological mechanisms. All subjects with a history of iron deficiency had been treated. Our study could not determine when the iron deficiency occurred, and a challenging long-term prospective study would be required to do so.

Potential Mechanisms Whereby Iron Deficiency May Influence Nociception and Pain

There are extensive cognitive, motor and neurobiological consequences of iron deficiency in early human life,^20,44–46 assessed in childhood, adolescence and in young adults,⁴⁷ but nothing enabling direct insight into nociception and pain.

Studies in young adults, especially women of childbearing age, with iron deficiency (with or without anemia), without knowledge of when it was acquired, have shown association with a range of behavioral and brain activity (electroencephalography) indicators of neurocognitive dysfunction, possibly reflecting low brain iron availability.^{44,45,48–52} A significant but unknown proportion of such young adults would have been iron deficient or at risk of iron deficiency from early life and thereby potentially more at risk of adverse consequences than those who acquired ID specifically in adulthood.

Searching for evidence that iron deficiency specifically acquired in adult life has neurocognitive consequences is challenging. Potentially, the best resource is studies on adult blood donors. Studies which include initial serum ferritin estimations are generally unavailable. Zalpuri et al and Di Angelantonio et al reported clinical consequences of iron deficiency from repeated blood donations,^46,47 but there were no pain-related assessments.

Generally, neuroimaging and pathological studies for low iron in adult human brains have not enabled direct insight into neurobiology relevant to nociception and pain. Rather, studies inspired by an interest in restless legs syndrome have been insightful for mechanisms of iron deficiency in the central nervous system, including pathways related to adenosine, glutamine and dopamine,^53–58 which may be relevant to pain sensitivity and vulnerability. Formulation of a hypothesis about the nervous system mechanisms underlying the influence of iron deficiency on nociception and pain requires review of animal studies, including Dowling et al and Yoo et al, and that is in process.^24,25

A community sample of women showed higher pressure pain sensitivity than men,⁵⁹ and healthy women showed more experimentally induced central sensitization than men.⁶⁰ Given the much higher prevalence of iron deficiency in women, it is interesting to hypothesize that a history of iron deficiency might contribute to their propensity to higher pain sensitivity and central sensitization than men. There is no objective study of induced or clinically occurring central sensitization from childhood through adolescence to adulthood.

The Importance of Multiplicity: Iron Deficiency and Multiple Pain Conditions, Primary, Overlapping and Multisite

It is notable that this study and our preceding pediatric publications showed associations between a history of iron deficiency, anxiety and depression as statistical predictors with multiple pain conditions,^1,2 the overlapping pain informal classification (including primary pain conditions),^5,61 and multisite pain.⁶ The multiplicity of pain associations with each other (Table S1) indicates that, if there is association with an individual pain condition, the likelihood of additional pain associations is increased. The more pain conditions per individual, the stronger was the association with iron deficiency, anxiety and depression (Table 4).

Overlapping and multisite chronic pain conditions collectively are complex traits which, compared with single chronic pain conditions,^5,6,61 have points of difference in causal influences, neurobiology and outcomes.^62–68 We have found no recent reviews of this important concept, hence the detail which follows. Coggon et al showed that extensive pain,⁶² affecting 6–10 anatomical sites, was reported much more frequently than would be expected if the occurrence of pain at each site were independent (674 participants vs 42 expected) and showed that extensive pain differs importantly in its associations with risk factors from pain that is limited to only a small number of anatomical sites.

Persons with multisite chronic pain are more likely to have higher chronic pain impact than persons with single-site chronic pain.⁶³ Living with multiple pain conditions and multiple bodily areas affected by pain is associated with greater pain severity, pain interference, emotional distress, lower quality of life, and poor psychosocial functioning.^64,65 Recently, multisite pain has been found to be associated, potentially causally, with increased cardiovascular disease risk, including myocardial infarction and stroke^66,67 and with dementia.⁶⁸

Neurobiological features reported as distinguishing multiple pain conditions include impaired axonogenesis shown by functional genomics.⁶⁹ Children who subsequently developed multisite pain had increased neural activity in superior parietal /primary somatosensory and motor cortices and decreased activity in the medial prefrontal cortex.⁷⁰ They also exhibited stronger functional connectivity between the salience network, somatosensory, and default mode network regions.

The frequent co-occurrence of pain conditions suggests common risks. Parental transmission of multisite chronic pain was reported by Zadro et al.⁷¹ Genome-wide studies, some with brain structure analysis, have shown substantial genetic influence on multiple pain conditions. Khoury et al compared genetic determinants of chronic single-site versus multisite pain in the UK Biobank.⁷¹ They found that different genetic signals underlie chronic single-site and multisite pain with much stronger genetic contributions for the latter. The findings of Johnston et al supported the proposition that chronic pain involves a strong nervous system component.^6,72 Multisite chronic pain had a significant polygenic component. Additional gene-level association analyses identified neurogenesis, synaptic plasticity, nervous system development, cell-cycle progression and apoptosis genes as enriched for genetic association with multisite pain. Genetic correlations were observed between multisite pain and a range of psychiatric, autoimmune and anthropometric traits, including major depressive disorder, asthma and Body Mass Index. Johnston and co-authors subsequently identified sex differences and a potential role for the dorsal root ganglion additional to the central nervous system in multisite chronic pain.⁷²

Zorina-Lichtenwalter et al identified genetic risk for multiple distinct pain disorders across individuals using 24 chronic pain conditions,⁷³ and genomic structural equation modeling revealed a general factor explaining most of the shared genetic variance across all pain conditions and a second more specific factor explaining genetic covariance across musculoskeletal pain conditions. Cross-reference with previous genome-wide association studies showed genetic overlap with cognition, mood, and brain structure.⁷³ These results identified common genetic risks and suggested neurobiological and psychosocial mechanisms. Schirle et al used a polygenic risk score approach to investigate genetic influences on chronic overlapping pain conditions.⁷⁴ Significant findings in this study provide evidence supporting previous hypotheses that a shared polygenic influence involving central sensitization may underly chronic overlapping pain conditions and can guide future biomarker and risk assessment research.

Numerous other risk factors for multiple pain conditions have been identified, including female gender, older age, multiple physically stressing occupational activities,⁶² adverse lifestyle factors,⁷⁵ sleep impairment,⁷⁶ multisensory sensitivity,⁷⁷ “eveningness”⁷⁸ restless legs syndrome,⁷⁹ mental health,⁸⁰ and others especially in combination.⁸¹

Strengths and Limitations

The twin family responders enabled access to a relatively large sample size of adults with adequate numbers of cases with a history of iron deficiency. The inclusion of twins might suggest confounding in the analyses, but the methods to assess associations, specifically logistic regression using generalized estimating equation (GEE) to account for correlations within families,³⁶ were appropriate to overcome such confounding. The totality of the evidence from our pediatric,^1–3 and adult studies adds to the confidence in the validity of the evidence for association between history of iron deficiency and the multiple pain conditions. The associations between the primary, overlapping and multisite pain conditions supports the concept of common causal mechanisms in such multiple pains, including iron deficiency, anxiety and depression as tested in this study. It was not valid to perform mediation analyses for anxiety and depression between history of iron deficiency and pain conditions because of the known bidirectional association between chronic pain and those psychological conditions (shown also in this study).

The approach letter to twin families mentioned restless legs syndrome and this potentially resulted in a bias towards responders who had that condition. Such bias would have probably increased the numbers with a history of iron deficiency, anxiety and depression (also impaired sleep). Consequently, the probability of association of iron deficiency with multiple pain conditions may have increased. Nevertheless, a history of iron deficiency remains a risk factor or risk marker for multiple pain conditions.

A significant limitation is that we are unable to identify the temporal relationships between acquisition of iron deficiency and the pain conditions and psychological disorders in individuals. The probability is that in most responders the iron deficiency preceded the clinical condition and if this were not the case it is unlikely that the significant associations would have been achieved. It is this limitation that has been a significant influence on the presentation of this paper as hypothesis generating.

The published associations between a history of iron deficiency and anxiety, depression and restless legs syndrome are dominantly unidirectional. Adolescent doctor-confirmed anxiety and iron deficiency acquired in adulthood for example could occur by chance and have no valid potential causal relationship. Such uncommon temporal associations which may misleadingly increase the association would probably be more than offset by those whose risk of anxiety, depression or restless legs syndrome were prevented by the administration of iron supplements. The associations between history of iron deficiency, anxiety, depression and restless legs syndrome in our pediatric publications^1–3 did not have such potential wide temporal disparity and were stronger.

Female preponderance was evident in the sample. In those who responded to the history of iron deficiency questions there were 871 adult females and 455 adult males. In the adults (18–65 years), 38.7% of females met the iron deficiency criteria and 2.42% of males. Although our sample included twins which may have introduced some bias (increased participation rate in those with a history of iron deficiency), this contrast is unsurprising given that, in an Australian study of new blood donors,⁸² 12.0% of women and 1.3% of men were iron deficient. In a longitudinal study of Australian women’s health,⁸³ 31% of middle-aged women reported “ever having low iron” diagnosed by a doctor. The preponderance of women with the multiple pain conditions,^4–6,84 and a history of anxiety and depression was consistent with published prevalence rates.⁸⁵ We would have included impairment of sleep in the study of associations with iron deficiency, but regrettably an appropriate questionnaire had not been included in the primary database accessed for this study.

Future Research, Clinical Implications and Considerations

We suggest that iron deficiency mechanisms which enhance the risk of nociception and pain have features in common with the identified causal influences on multiplicity of pain conditions. However, further clinical research is required to determine the extent to which there is a direct neurobiological effect of iron deficiency on nociception, especially on the developing nervous system, or the extent which the iron deficiency influence on multiple pain conditions relates to its co-morbidity with sleep, anxiety and depression.

Management and therapeutic implications of the results of this study are necessarily provisional in individuals who have passed the development stage. The strongest guidance is that it is critical to test all pregnant women and their newborn infants for iron deficiency and to initiate optimal supplementation, by intravenous infusion if necessary. Generally, menstruating and older females should also be closely monitored for iron deficiency, especially during adolescence when there is a greater chance of therapeutic response to supplementation. All individuals who have a history of iron deficiency should be assessed for sleep impairment, anxiety and depression.

Conclusion

We have shown, in a community sample of adults from twin families, that a history of iron deficiency was associated with multiple overlapping (including primary) and chronic pain conditions. The analyses suggest that iron deficiency might have a direct effect on pain sensitivity and vulnerability, and potentially a substantial indirect effect through it`s known associations with anxiety and depression. Studies in mice have provided support that iron deficiency is probably causal in altering nociception and vulnerability to pain, however we accept that our results are best considered as hypothesis generating and in need of further testing.

Disclosure

The abstract of this paper was presented with interim findings as a free paper at the Annual Scientific Meeting of the Australian Pain Society, Darwin, April 2024. The poster’s abstract was not published. Dr Elias Karroum reports personal fees from Noctrix Health Inc., outside the submitted work. The author(s) report no conflicts of interest in this work.