earch 65 (2008) 131–133

Journal of Psychosomatic Res

A link between lunar phase and medically unexplained stroke symptoms:An unearthly influence?

Faheem Ahmada, Terence J. Quinnb,⁎, Jesse Dawsonb, Matthew Waltersb

aUniversity of Glasgow Medical School, Glasgow, UKbDivision of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow, UK

Received 10 December 2007; received in revised form 23 April 2008; accepted 6 May 2008

Abstract

Objective: Lunar effects on human behaviour and health havebeen postulated for centuries; associations between phase of themoon and both vascular and psychiatric disease have been reported.We hypothesised that admission with medically unexplained strokesymptoms would be influenced by lunar cycle. Methods: Alladmissions to the Western Infirmary Acute Stroke Unit are recordedin a comprehensive database. We analysed admissions betweenJanuary 1, 1993, and September 30, 2006 (inclusive). Associationbetween admission rate, phase of the moon, and a series of otherculturally significant dates was calculated using chi-square testing.Results: There were 7219 admissions during the study period,

⁎ Corresponding author. Gardiner Institute, Cardiovascular and MedicalSciences, Western Infirmary, G116NT Glasgow, UK. Tel.: +44 1412116395;fax: +44 1412112895.

E-mail address: tjq1t@clinmed.gla.ac.uk (T.J. Quinn).

0022-3999/08/$ – see front matter © 2008 Elsevier Inc. All rights reserved.doi:10.1016/j.jpsychores.2008.05.011

comprising 167 complete lunar cycles. Stroke admissions wereevenly spread throughout lunar phases (P=.72). Admission withmedically unexplained stroke symptoms was significantly increasedduring full moon phases (P=.023). There was no variation inadmission rates during other significant dates. Conclusion: Therewas a statistically significant association between admission to theacute stroke unit with a diagnosis of medically unexplained strokesymptoms and lunar phase. The reasons for this observed variationremain elusive; our data do not support a convincing biological orpsychosocial aetiology.© 2008 Elsevier Inc. All rights reserved.

Keywords: Cerebrovascular accident; Medically unexplained symptoms; Moon

Introduction

Belief in lunar influences on human health and behaviorhas been reported for centuries and persists amongst lay-public and medical staff [1]. Correlations between lunarcycle and psychiatric illness, neurological disease, andvascular disease have all been previously reported [2–4]but remain controversial [5]. A recent, nonsystematicreview of the literature describing lunar effects on healthand behaviours found almost equal numbers of paperssuggesting and refuting this association [6]. The divergentliterature precludes any formal meta-analysis, and as such,the question remains unanswered. To date, there are no

published data exploring lunar influence on cerebrovas-cular disease.

Admissions to an acute stroke unit include a cohort ofpatients with no demonstrable cerebrovascular disease.Within this population, there exists a consistent subgroupof patients for whom no medical diagnosis is reached,despite extensive investigation. The underlying pathologyin these patients with medically unexplained strokesymptoms (MUSS) is traditionally regarded as “psychia-tric” rather than “medical” [7]. Dependent on localexpertise, these patients can present to a number ofservices including general medical outpatients, neurology,and stroke-specific services, and as such, a betterunderstanding of MUSS should be of interest to mostpracticing physicians.

Recognising links between lunar cycle and psychiatricillness, we hypothesised that admissions with MUSS wouldbe increased on full-moon days.

132 F. Ahmad et al. / Journal of Psychosomatic Research 65 (2008) 131–133

Methods

All patients presenting with suspected acute stroke areadmitted to our dedicated stroke unit, approximating 800patients each year. Full clinical details including radiologicalresults and discharge diagnosis are prospectively recordedfor the West Glasgow Stroke Registry. Clinical, biochemical,and radiological data are reviewed and verified by a teamcomprising neurologist, radiologist, and other stroke physi-cians before a final diagnosis is recorded.

The diagnosis of MUSS is a diagnosis of exclusion,recorded once the complete team have concomitantlyreviewed the case and jointly agreed on the classification.In brief, MUSS is recorded if signs and symptoms areinconsistent with organic disease and comprehensiveinvestigations including imaging are negative. A completeaccount of the accepted definition of MUSS can be foundelsewhere [7,8].

We included all patients admitted from January 1, 1993,to September 30, 2006 (inclusive). To determine lunar phaseclassification, the lunar month was divided into therecognised quarters: new moon, first quarter, full moon,and last quarter. Each phase extended from the day of thenew quarter (inclusive) to the day preceding the beginning ofthe next. Patients were classified according to the lunar phaseor other date existent on day of admission. As most patientspresent to the unit early, this was considered a reasonablesurrogate for day of symptom onset.

Statistical analyses were performed using Minitab soft-ware (version 13.1, Minitab Inc, PA, USA). Chi squaredstatistics were used to evaluate association betweenadmission with stroke or MUSS, phase of the moon andseason of the year. We performed further analysis ofadmission rates for a series of dates with culturalsignificance in the United Kingdom: Friday 13th andHalloween (October 31st). Each analysis was across 4groups. The precise date of each quarter and othersignificant dates were obtained from Her Majesty's NauticalAlmanac Office (http://www.nao.rl.ac.uk/). In recognition ofthe disproportionate MUSS burden across differing sex andage, we performed an equivalent post hoc analysis limited tosingle sex and dichotomised (65 years) age groups.

Table 1Admissions to the acute stroke unit by lunar phase

Date/eventTotaldays

Incidence

Stroke MUS

All days (reference) 4985 6845 129Full moon 167 1735 (25.3%) 47 ⁎ (36.4%)New moon 167 1743 (25.5%) 30 (23.3%)First quarter 167 1658 (24.2%) 28 (21.7%)Last quarter 167 1709 (25.0%) 24 (18.6%)Halloween (Oct 31) 13 21 (0.3%) 0Friday 13th 21 33 (0.5%) 0

MUS, medically unexplained symptoms.⁎ P=.023.

Results

During the study period of 167 complete lunar cycles,7219 patients were admitted to the Acute Stroke Unit; 6845(94.8%) had a diagnosis of stroke (3504 female). Admissionswith stroke were evenly distributed throughout the lunarcycle (P=.72). There were 129 first presentations of MUSSduring the study period (1.78% of total admissions, 81female). A statistically significant increase in admission ratesfor MUSS was observed during the full-moon phase(P=.023). There was seasonal variation in MUSS, withadmissions increased during autumn and winter (P=.033)

Table 1. There was no association for stroke/MUSS andother culturally significant dates. MUSS were over repre-sented in female patients and in patients under 65 years old(Table 2). Lunar relationships were less apparent whenanalysis was restricted to single sex and age groups;however, these results must be interpreted with caution dueto small numbers in the male and older groups (P=.008 forb65 years old; P=.697 for N65 years old; P=.016 for femalegender; P=.448 for male gender).

Discussion

We report a significant increase in presentations withMUSS during the full moon. These results add credence tothe concept of lunar stimuli influencing human healthbehaviours. In synchrony with our findings, the majority ofprevious studies have reported a maximal lunar effect on fullmoon days [2–4].

We believe our data are robust, and the observedrelationship between MUSS and lunar phase is not artefact.The long duration of the study should reduce the effect ofextraneous temporal variations, although this can never befully excluded. Results are not explained by variation inoverall caseload of the unit, which did not fluctuatesignificantly during the study period. Composition of themultidisciplinary adjudicating committee was consistent,and criteria for referral to the cerebrovascular service did notchange throughout the duration of the study period. As withany scientific trial, play of chance may give an unexpectedpositive result, a well-cited example can be found in the ISIS2 [9] trial, where on multiple statistical testing, pooroutcomes were associated with astrological sign. However,for the small numbers of MUSS recorded, our data arecompelling with a statistically significant result on primaryhypothesis testing rather than on subgroup analyses.

Although not defined as a distinct entity in diseasemanuals such as the International Classification of Diseases,the syndrome of a stroke like presentation with extensivenegative investigations is well recognised in neurology,medicine, and psychiatry [7,8]. It is possible that somepatients labelled MUSS had undetected organic disease.

Table 2Demographics of acute stroke unit admissions comparing “true” stroke andmedically unexplained symptoms

All admissions (n=7219)

Incidence

Stroke MUS

Female sex 3504 81Male sex 3341 48P=.008AgeN65 4883 37Ageb65 1962 92Pb.001

133F. Ahmad et al. / Journal of Psychosomatic Research 65 (2008) 131–133

However, long-term prospective studies suggest that it isuncommon for patients with MUSS to be misclassified whena rigorous multidisciplinary approach to diagnosis is applied[8]. In defining MUSS cases, we used identical criteria tothose employed in earlier reports [7].

A convincing scientific mechanism for the association withlunar phase remains elusive. Disruption of the earth's electro-magnetic field or gravitational effect on water masses of thehuman body has been proposed [10]. Were these applicable toMUSS, a bimonthly change in presentation corresponding to thegreater gravitational influence of new moon and full moonwould be expected, but this was not observed.

We recognise that the underlying association may not bebiological and that the psychosocial constructs built aroundthe moon may better explain the effect of lunar phase. Tofurther investigate, we analysed frequency of MUSSadmissions during other periods—Friday the 13th andHalloween. These dates were chosen as they representperiods of spiritual/cultural significance in the UK, with noassociated biological pattern. Previous authors have studiedmedical admissions on religious holidays—for cerebralhaemorrhage, rates in a Japanese population are increasedon traditionally unlucky days [11]. We chose our significantdates in recognition of the multicultural population we serveand also acknowledging the general secular trends in theUnited Kingdom. Numbers were small as these dates occuronly annually; accepting this limitation, there was noobserved increase in MUSS. Perhaps the most importantfactor is the psychosocial meaning that these dates have forthe individual patient. At present, we do not collect detailedpsychological assessment data on MUSS patients. Futureprospective studies of MUSS may wish to incorporate suchan assessment.

The significant seasonal variation in MUSS is a novelfinding. The pattern of increased incidence in winter is inkeeping with other medically unexplained disorders [12] andwith affective disorders such as depression [13]. The role ofcircadian light rhythms and neurohumeral responses to thesestimuli is slowly being elucidated through animal and humanwork. Lunar effects on biological rhythms such as themenstrual cycle are found throughout nature [6], and it is

interesting to note that in keeping with other medicallyunexplained syndromes, younger female sex predominatesin MUSS. We await further work in this area, as the complexrelationship between novel environmental stimuli, health,and behaviours are unravelled.

Our findings have potential clinical application. Recogni-tion of increased MUSS incidence during specific timeperiods could aid in staff training and planning of servicedelivery. Diagnostic clues to MUSS will be of practicalbenefit as we make increasing use of hyperacute stroketherapies, often delivered within time windows that do notallow for definitive diagnosis. Cohorts of “strokes” receivingemergency thrombolytic therapy frequently include patientssubsequently proven to have MUSS [14].

We have identified a clear relationship between presenta-tion to a stroke unit with medically unexplained symptomsand lunar phase. The underlying mechanism remains elusivebut further work to corroborate our findings and to evaluateany proposed theoretical models seems justified. In parti-cular, prospective studies to define the physical andpsychological determinants ofMUSS admission are required.

References

[1] Danzl DF. Lunacy. J Emerg Med 1997;5:91–5.[2] Kung S, Mrazek DA. Psychiatric emergency department visits on

full-moon nights. Psych Services 2005;56:221–2.[3] Polychronopoulos P, Argyriou AA, Sirrou V, Huliara V, Aplada M,

Gourzis P, Economou A, Terzis E, Chroni E. Lunar phases and seizureoccurrence: just an ancient legend? Neurology 2006;66:1442–3.

[4] Takagi H, Umemooto T. Lunar cycles and abdominal aortic aneurysmrupture. J Vasc Surg 2004;40:1261.

[5] Gorvin JJ, Roberts MS. Lunar phases and psychiatric hospitaladmissions. Psychol Rep 1994;75:1435–40.

[6] Zimecki M. The lunar cycle: effects on human and animal behaviourand physiology. Postępy Hig Med Dośw (Advances in Hygiene andExperimental Medicine) 2006;60:1–7.

[7] Nazir FS, Lees KR, Bone I. Clinical features associated with medicallyunexplained stroke like symptoms presenting to an acute stroke unit.Eur J Neurol 2005;12:81–5.

[8] Crimlisk HL, Bhatia K, Cope H, David A, Marsden CD, Ron MA.Slater revisited: 6 year follow up study of patients admitted withmedically unexplained motor symptoms. BMJ 1998;316:582–6.

[9] ISIS-2 (Second International Study of Infarct Survival) CollaborativeGroup. Randomised trial of intravenous streptokinase, oral aspirin,both, or neither among 17,187 cases of suspected acute myocardialinfarction: ISIS-2. Lancet 1988;ii:349–436.

[10] Myers DE. Gravitational effects of the periods of high tides and thenew moon on lunacy. J Emerg Med 1995;13:529–32.

[11] Nakaguchi H, Teraoka A. Relationship between the occurrence ofintracerebral hemorrhage and holidays and traditionally unlucky daysin Fukuyama City, Japan. J Stroke Cerebrovasc Dis 2007;16:194–8.

[12] Pollmann L. Fibromyalgia: chronobiological aspects. ChronobiologyInt 1994;11:393–6.

[13] Wehr TA, Rosenthal NE. Seasonality and affective illness. Am J Psych1989;146:829–39.

[14] Tsivgoulis G, Hoover S, Frey JL, et al. Safety of thrombolysis in strokemimics. Stroke 2008;39:594–5.