Age and Ageing 1994:23:200-203

Iron Status and Restless LegsSyndrome in the ElderlyS. T. O'KEEFFE, K. GAVIN, J. N. LAVAN

SummaryThe relationship between iron status and the restless legs syndrome (RLS) was examined in 18 elderly patientswith RLS and in 18 matched control subjects. A rating scale with a maximum score of 10 was used to assess theseverity of RLS symptoms. Serum ferritin levels were reduced in the RLS patients compared with controlsubjects (median 33 /ig/1 vs. 59 //g/1, p<0 .01 , Wilcoxon signed rank test); serum iron, vitamin Bn and folatelevels and haemoglobin levels did not differ between the two groups. Serum ferritin levels were inverselycorrelated with the severity of RLS symptoms (Spearman's rho —0.53, p <0.05). Fifteen patients with RLSwere treated with ferrous sulphate for 2 months. RLS severity score improved by a median value of 4 points insix patients with an initial ferritin =$ 18 /Jg/1, by 3 points in four patients with ferritin > 18 /xg/1, ^45 /ig/1 and by1 point in five patients with ferritin > 45 ^g/1, < 100 /Jg/1.

Iron deficiency, with or without anaemia, is an important contributor to the development of RLS in elderlypatients, and iron supplements can produce a significant reduction in symptoms.

IntroductionSeveral studies have reported low serum (or plasma)iron or ferritin levels in psychiatric patients who developakathisia during neuroleptic treatment compared withmatched controls who do not develop akathisia [1-3]. Acorrelation has also been reported between the degree ofiron depletion and the severity of neuroleptic-inducedakathisia [1, 2]. Other authors have failed to confirmthese findings [4—6], and the matter remains controver-sial.

An association between low serum iron levels or irondeficiency anaemia and development of the restless legssyndrome (RLS], a condition with many similarities toakathisia, was reported more than 30 years ago [7-10].Indeed, it was knowledge of this relationship thatprompted investigation of iron status in akathisia [1, 11].However, little systematic research into the relationshipbetween iron status and RLS has been reported in theintervening decades. Serum iron levels are of limitedvalue in assessing iron status because of the largenumber of clinical conditions which influence irontransport [12]. Serum ferritin, an indicator of ironstores, declines before a fall in serum iron and is nowregarded as the most useful indicator of iron deficiency[13]. In a recent prospective study of elderly hospitalpatients, we identified iron deficiency (defined as aserum ferritin ^ 1 8 /̂ g/1) in four (31%) of 13 patientswith insomnia due to RLS [14]. However, it is uncertainwhether depletion of iron stores, in the absence of overtiron deficiency, is associated with RLS, or whether theseverity of symptoms in RLS correlates with the degree

of iron deficiency. Also, vitamin B12 and folate deficien-cies are common in elderly patients and have beenreported in association with RLS [14, 15], but theimportance of these deficiencies in the pathogenesis ofRLS has not been examined in detail.

We studied serum iron, ferritin, folate and vitamin B12levels in 18 elderly patients with RLS and in 18 matchedcontrol patients. The influence of initial serum iron andferritin levels on the response to iron supplements in thepatients with RLS was also evaluated.

Subjects and MethodsThe study sample comprised 36 patients recruited from thewards and outpatient clinics of an acute-care geriatric unit.Eighteen patients with RLS were pair-matched with 18patients without RLS by age, sex, location (inpatient oroutpatient) and reason for referral. Subjects taking ironsupplements, non-steroidal anti-inflammatory agents, ome-prazole or H2 blockers and subjects with liver disease,malignancy (except skin cancer) or peripheral neuropathywere excluded. A diagnosis of RLS was made if a patient hadbilateral nocturnal leg discomfort which satisfied the followingcriteria [10]: the site included the calf or shin; the sensation wasaccompanied by an urge to move the legs and was relieved bymovement; symptoms were not of tingling, pins-and-needles,numbness, cramps or burning sensations alone; the patientwas not taking neuroleptic or anti-parkinsonian medications.The severity of the symptoms was assessed using the scaleshown in Table I. All RLS patients had a score of more than 2,and all controls had a score of 0. The RLS severity score wasreassessed after 4—6 weeks, before iron therapy was com-menced, in ten patients with RLS; the score was the same for

at Aston U

niversity on April 27, 2014

http://ageing.oxfordjournals.org/D

ownloaded from

IRON STATUS AND RESTLESS LEGS SYNDROME

Table I. Restless legs rating scale

Do you get an unpleasant restless feeling in your legs atnight which is relieved by moving the legs?

0 = never1 = rarely (less than once a month)2 = occasionally (less than once a week)3 = often (at least once a week)4 = almost every night

How distressing is this sensation?0 = no distressl=mi ld2 = moderate3 = severe

How long do these sensations usually last?0 = no time or a few seconds1 = < 30 minutes2 = > 30 minutes, < 1 hour3 = > 1 hour

Maximum score =10

eight patients and there was a change of 1 point for twopatients. Inter-rater reliability using the scale was excellent(Cohen's kappa = 0.98). Morning venous blood samples wereanalysed, as previously described, for estimation of full bloodcount, serum iron, ferritin, Bi2, folate and red-cell folate levels[14].

Patients with a serum ferritin ^ 4 5 /ig/1 were prescribed ironsupplements as ferrous sulphate 200 mg three times daily.Patients with serum ferritin levels between 46 and 100 /ig/1were informed that their blood tests showed the possibility of amild iron deficiency and a course of iron supplements might bebeneficial. If they agreed, they were also started on ironsupplements. The possibility of an improvement in symptomsof RLS or insomnia was not mentioned. Patients who werealready receiving treatment for restless legs or insomnia weremaintained on this treatment for the duration of the study.RLS severity score and blood tests were repeated after aminimum of 2 months' treatment with iron supplements.

Results were analysed by Wilcoxon matched-pairs signedrank test and Spearman's rank correlation.

ResultsThere were 13 women and five men and the median agewas 81 years (range 70-87) in each group. Of the 18patients in each group, five (28%) patients with RLSand two (11 %) control patients were taking benzodiaze-pines at the onset of the study.

Serum ferritin levels were significantly lower in theRLS patients than in the controls (median 33 /ig/1 vs. 59/ig/1, p<0.01) (Table II). Serum iron, vitamin B12 andfolate and haemoglobin and mean corpuscular volumevalues did not differ between the two groups. There wasa significant inverse correlation between serum ferritinand the restless legs rating score (Spearman's rho- 0 . 5 3 , p < 0.05) (Figure).

All of the patients with RLS who had serum ferritin< 100 /ig/1 (n = 1 7) were prescribed ferrous sulphate 200mg three times daily. One patient was unable to tolerateiron, and another died before follow-up. Median follow-

Imcg

/le

rrii

E

<DCO

120 -

100 -

80 -

60 -

40 -

20 -

0 2 4 6 8 10

RLS ssverity scoreFigure. Serum ferritin and RLS severity score in 18 patientswith RLS. Spearman's rho= —0.53, p<0.05.

up for the remaining 15 patients was 12 weeks (range 8-20 weeks). All patients showed an increase in serumferritin levels; the median change was + 34 /ig/1 (range+ 69 to +10 /ig/1). Median RLS rating score improvedfrom 5 (range 3-9) to 3 (0-5) (Table III). It was possibleto withdraw benzodiazepine therapy from three of thefive patients with RLS taking this medication.

Four patients in the RLS group and two controlsubjects had anaemia (Hb < 11 g/dl); six RLS patientsand two controls had serum ferritin < 18 /ig/1. Nosubject had folate deficiency. Two RLS patients had

Table II. Serum iron, B,2, folate and haematologic values inrestless legs (RLS) patients and controls. Values are median(range)

Variables

Iron (/imol/1)Ferritin (/ig/1)Vitamin B12 (ng/l)fFolate (/ig/l)tHaemoglobin (g/dl)Mean cell vol. (/im3)

R L S(n = 18)

11 (4-28)32.5 (6-124)

290(184-1134)6.3 (2.3-15.6)

11.7(8.3-14.1)82 (62-93)

Controls(n = 18)

14(6-25)59(12-189)*

305 (168-1890)5.7(2.5-13.7)

11.9(8.7-15.6)84 (64-92)

* p < 0.01; other differences are not significant,f Results were available for 16 of the controls.

Table III. Relationship between initial serum ferritin level andchange in RLS severity score in 15 patients treated with ironsupplements

Initial serum ferritin (/ig/1)

J£18 > 18, s?45 >45, < 100

No. of patients 6 4 5

Median change in severityscore (range) 4 (0-8) 3 (0-5) 1 (0-2)

Change in severity score5=2 points, no. (°o) 5(83) 3(75) 2(40)

at Aston U

niversity on April 27, 2014

http://ageing.oxfordjournals.org/D

ownloaded from

S. T. O'KEEFFE ET AL.

vitamin B12 deficiency, and both were treated withintramuscular vitamin B12 as well as with iron supple-ments: RLS score improved from 6 to 3 in one patientwith an initial vitamin B12 level of 184ng/land ferritinof7 A<g/1; there was no change in score for the other patientwho had a vitamin B12 level of 197 ng/1 and a ferritin of24 //g/1. Gastro-intestinal investigations in nine of theten RLS patients with ferritin levels <45 ^g/1 revealedpeptic ulcer disease (1), oesophagitis (3), carcinoma ofthe colon (1), benign rectal polyp (1) and no abnormality(3).

DiscussionThe significantly low serum ferritin levels in the RLSgroup, the correlation between ferritin levels and sever-ity of RLS and the improvement in symptoms with ironrepletion, especially in patients with lower ferritinlevels, suggest that depletion of iron stores, with orwithout anaemia, is related to the development of RLS.Vitamin B12 deficiency was an additional factor in onepatient, but folate deficiency was not important in thisstudy. This study was restricted to elderly patientsamong whom RLS is thought to be most common [16].One possible reason is the higher prevalence of medicalconditions, including iron deficiency, which are asso-ciated with the syndrome. However, there is no evidencethat susceptibility to akathisia increases with advancingage [17].

The explanation for the association between irondeficiency and RLS or akathisia may lie in altereddopaminergic neurotransmission [1]. Studies with iron-chelating agents suggest that the dopamine D2 receptoris an iron-containing protein [18]. Iron deficiency isassociated with hypofunction of this receptor, which canbe reversed with iron repletion [19]. Reduced levels ofdopamine and homovanillic acid in the cerebrospinalfluid of RLS patients [20], and the clinical response todopaminergic agents suggest that dopaminergic neuro-transmission is important in the pathogenesis of thesyndrome [21]. Postsynaptic dopamine D2 receptorblockade in mesocortica! pathways has been reported asthe likely cause of neuroleptic-induced akathisia [22]. Areduction of striatal D2 dopamine receptor binding siteswith age may contribute to the pathogenesis of thesedisorders in older patients [23].

There are various limitations to our study. We used asubjective scale to assess the severity of RLS. All-nightpolysomnography and videotaping have been used bysome investigators to quantify motor restlessness [24].However, these only provide information on a singlenight, while it is well recognized that the severity of RLSmay vary considerably from night to night. Also, thesetests do not take the distress caused by creepingsensations in the legs into account. Secondly, thetreatment phase of this study was not blinded. Telstedand colleagues have demonstrated a high placebo effectin the treatment of RLS [25]. However, our patientswere not informed that iron supplements might relievetheir symptoms. Also, many had been treated withbenzodiazepines without relief.

RLS is notoriously difficult to treat, and side effectsare common with agents such as carbamezapine, levo-dopa and clonazepam. This study shows the importanceof measuring serum ferritin levels in patients with thiscondition. The optimal cut-off point for serum ferritinin the diagnosis of iron deficiency depends on the pre-test probability of iron deficiency in the populationbeing studied. Recent work suggests that a cut-off forserum ferritin of 45 /ig/1 is appropriate in elderlypatients with [13] or without anaemia [26]. Elderlypatients with RLS have a particularly high prevalence ofiron deficiency. The response to iron supplements inRLS patients with a ferritin level ^45 /ig/1 is usuallyexcellent. A trial of iron therapy should also be consid-ered in patients with serum ferritin levels between 46and 100/ig/1, since some of these patients will also have agood response.

References1. Brown KW, Glen SE, White T. Low serum iron status

and akathisia. Lancet 1987;i: 1234—6.2. Barton A, Bowie J, Ebmeier K. Low plasma iron status

and akathisia. J Neurol Neurosurg Psychiatry 1990;53:671-4.

3. Horiguchi J. Low serum iron in patients with neuroleptic-induced akathisia and dystonia under antipsychotic drugtreatment. Ada Psychiatr Scand 1991 ;84:301-3.

4. Sachdev P, Loneragan C. Acute drug-induced akathisia isnot associated with low serum iron levels. Psychopharma-cology 1991; 103:138-9.

5. Zemes ZC, Rotrosen J, Angrist B, Peselow E, SchoentagR. Serum iron levels and akathisia. Biol Psychiatry1991;29:411-13.

6. Barnes TR, Halstead SM, Little PW. Relationshipbetween iron status and chronic akathisia in an in-patientpopulation with chronic schizophrenia. Br J Psychiatry1992; 161:791-6.

7 Nordlander NB. Therapy in restless legs. Ada Med Scand1953;145:453-7.

8. Ekbom KA. Restless legs syndrome. Neurology 1960;10:868-73.

9. Parrow A, Werner I. The treatment of restless legs. AdaMed Scand 1966;180:401-6.

10. Ekbow KA. Restless legs. In: Vinken PJ, Bruyn GW, eds.Handbook of clinical neurology vol. 8. Amsterdam: NorthHolland Publishing Company, 1970;311-20.

11. Blake DR, Williams AC, Pall WS, Fonseca A, Beswick T.Iron and akathisia. Br Med J 1986;292:1393.

12. Cook JD, Skikne BS. Iron deficiency: definition anddiagnosis. J Int Med 1989;226:349-55.

13. Guyatt GH, Patterson C, All M, et at. Diagnosis of iron-deficiency anaemia in the elderly. Am J Med 1990;88:205-9.

14. O'Keeffe ST, Noel J, Lavan JN. Restless legs syndromein the elderly. Postgrad Med J 1993;69:701-3.

15. Botez MI, Lambert B. Folate deficiency and restless-legssyndrome in pregnancy. N Engl J Med 1977;297:670.

16. Kreuger BR. Restless legs syndrome and periodic move-ments of sleep. Mayo Clin Proc 199O;65:999-1006.

17. Sachdev F, Loneragan C. The present status of akathisia.J Nerv Ment Dis 1991; 179:381-91.

18. Ben-Schacher D, Finberg JPM, Youdim MDH. Effectsof iron chelators on dopamine D2 receptors. J Xeurochem1985;145:999-lOO5.

at Aston U

niversity on April 27, 2014

http://ageing.oxfordjournals.org/D

ownloaded from

IRON STATUS AND RESTLESS LEGS SYNDROME

19. Ashkenazi R, Ben-Schacher D, Youdim MDH. Nutritio-nal iron and dopamine binding sites in the rat brain.Pharmacol Biochem Behav 1982;17(suppl l):43-7.

20. Montplaiser J, Godbout R. Restless legs syndrome andperiodic movements during sleep. In: Kryger MH, RothT, Dement WC, eds. Principles and practice of sleepmedicine. Philadelphia: WB Saunders, 1989;402-9.

21. Walters AS, Hening W. Clinical presentation and neuro-pharmacology of restless legs syndrome. Clin Neurophar-macolX 987;10:225-37.

22. Marsden CD, Jenner P. The pathophysiology of extra-pyramidal side-effects of neuroleptic drugs. Psvchol Med1980;10:55-72.

23. Antonini A, Leenders KL, Reist H, Thomann R, BeerH-F, Locher J. Effect of age on D2 dopamine receptors innormal human brain measured by positron emissiontomography and "C-raclopride. Arch Neurol1993;50:474-80.

24. Walters AS, Hening WA, Chokroverty S. Review and

videotape recognition of idiopathic restless legs syn-drome. Movement Dis 1991 ;6:105-10.

25. Telstad W, Sorensen 0 , Larsen S, Lillevold PE, StensrudP, Nyberg-Hansen R. Treatment of the restless legssyndrome with carbamezapine: a double blind study. BrM ed J 1984,288:444-6.

26. Joosten E, Dereymaeker L, Pelemans W, Hiele M.Significance of a low serum ferritin level in elderly in-patients. Postgrad Med J 1993;69:397-4OO.

Authors' addressesDepartment of Geriatric Medicine, Beaumont Hospital,Dublin 9

Address correspondence to: Dr S. O'Keeffe, UniversityDepartment of Geriatric Medicine, Royal Liverpool Univer-sity Hospital, Prescot Street, PO Box 147, Liverpool L69 3BX

Received 30 September 1993

at Aston U

niversity on April 27, 2014

http://ageing.oxfordjournals.org/D

ownloaded from