Analysis of 31 Families With an ApparentlyAutosomal-Dominant Transmission of MigraineWith Aura in the Nuclear Family

Vibeke Ulrich,* Michael Bjørn Russell, Steen Østergaard, and Jes OlesenDepartment of Neurology, Glostrup Hospital, University of Copenhagen, Glostrup, Denmark

We analyzed 31 families selected for an ap-parently autosomal-dominant mode of in-heritance of migraine with aura (MA) in thenuclear family. The nuclear families wereexpanded with first- and second-degreerelatives. All interviews were made by phy-sicians experienced in headache diagnoses.The criteria of the International HeadacheSociety were used. The population relativerisk among children in nuclear families wassimilar to the estimated population relativerisk of MA assuming an autosomal-domi-nant mode of inheritance. The populationrelative risk tended to decrease among first-degree relatives outside nuclear familiesand further among second-degree relatives.Both first- and second-degree relatives out-side the nuclear families had a statisticallysignificant lower risk of MA than expected.Thus, autosomal-dominant inheritance withor without reduced penetrance was un-likely. Autosomal-recessive inheritance wasunlikely because of the unequal sex distri-bution. Other modes of inheritance wereconsidered as well. Mitochondrial and X-linked inheritance were excluded becauseof paternal transmission. The female pre-ponderance was too low to explain sex-influenced inheritance. We conclude thatMA most likely has a multifactorial inheri-tance even in high-risk families with MA.Am. J. Med. Genet. 74:395–397, 1997.© 1997 Wiley-Liss, Inc.

KEY WORDS: migraine with aura; inheri-tance; genetic transmission

INTRODUCTION

Migraine with aura (MA, previously classic mi-graine) is one of the most common neurological disor-ders, affecting about 8% of the general population [Ras-mussen and Olesen, 1992a; Russell et al., 1995a]. At-tacks are initiated by aura, i.e., reversible visual,sensory, motor, and/or aphasic symptoms mostly last-ing less than 60 min [Headache Classification Commit-tee of the International Headache Society, 1988]. Theensuing headache is usually severe, unilateral, pulsat-ing, aggravated by physical activity, and accompaniedby nausea, and photo- and phonophobia [Rasmussenand Olesen, 1992a; Russell et al., 1996a]. Comparedwith the general population, first-degree relatives ofprobands with MA have a 3.8-fold increased risk of MA,while spouses to probands have no increased risk ofMA, emphasizing the importance of genetic factors[Russell and Olesen, 1995]. A complex segregationanalysis indicates that MA has multifactorial inheri-tance without generational differences, but genetic het-erogeneity could not be excluded [Russell et al., 1995b].Thus, subgroups of families with MA may have othermodes of inheritance.

The aim of the present study was to investigate thepossibility of some families having an autosomal-dominant mode of inheritance of MA. We selectednuclear families (parents and their offspring) with af-fected in both generations and expanded these nuclearfamilies in order to analyze the relative risk of MA infirst- and second-degree relatives.

MATERIALS AND METHODSData Collection

A total of 317 probands with MA was recruited fromthe Copenhagen Headache Clinic. They were inter-viewed by telephone using a semistructured headacheinterview. Of the 317 probands, 31 were part of anuclear family with one affected and one unaffectedparent and at least one affected and one unaffectedoffspring. These nuclear families were expanded withfirst-degree relatives. Second-degree relatives wereonly included if first-degree relatives were affected.Relatives aged 10 or above were included. All the in-terviews were made by two neurological residents

*Correspondence to: Vibeke Ulrich, Department of Neurology,University of Copenhagen, Glostrup Hospital, DK-2600 Glostrup,Denmark.

Received 16 October 1996; Revised 19 March 1997

American Journal of Medical Genetics (Neuropsychiatric Genetics) 74:395–397 (1997)

© 1997 Wiley-Liss, Inc.

(V.U. and S.Ø.) experienced in headache diagnoses.The criteria of the International Headache Societywere used [Headache Classification Committee of theInternational Headache Society, 1988].

Fifteen first-degree relatives were not interviewed; 8were deceased, 4 had an unknown address, and 3 de-clined to participate. Thirteen second-degree relativeswere not interviewed; 11 were deceased and 2 had anunknown address.

The project was approved by the Danish EthicalCommittees.

Statistical Methods

The risk of familial occurrence was assessed by esti-mating the population relative risk of MA in specifiedgroups of relatives [Weiss et al., 1982]. The risk wascalculated according to the following equation:

P(relative is affected | proband is affected)P(random member of the population is affected)

A family aggregation is implied when the risk ratiosignificantly exceeds 1.

Since the prevalence of MA depends on sex and age,the value of the denominator was adjusted according tothe distribution of sex and age in the group of relativesstudied [Russell and Olesen, 1995]. Hence, the stan-dardized population relative risk was estimated by di-viding the observed number of affected first- and sec-ond-degree relatives, respectively, by the expectednumber according to prevalence rates in the popula-tion. The expected number was calculated by addingthe products of the current sex- and age-specific ratesand the number of relatives within each correspondingsex-age category. The adjusted population relative riskwas estimated separately for children in nuclear fami-lies, and for first- and second-degree relatives outsidethe nuclear families. We also estimated the populationrelative risk, assuming that MA has an autosomal-dominant mode of inheritance.

Statistical analyses were performed using the SPSSBase System for Windows 7.0. Frequencies were com-pared by Fisher’s exact test and means by Student’st-test. A 5% level of significance was used.

RESULTS

The 31 probands (22 females and 9 males) had 132first-degree relatives (74 females and 58 males) ofwhom 46 (35 females and 11 males) had MA. Of the 30

second-degree relatives (17 females and 13 males), 5had MA (2 females and 3 males). Females were signifi-cantly more often affected than males (P 4 0.002), witha male:female ratio of 1:1.8.

Pedigrees

Transmission of MA from mother to daughter wasseen in 21 cases, mother to son in 11 cases, father todaughter in 9 cases, and father to son in 4 cases. Thus,maternal transmission was 2.5 times more frequentthan paternal transmission (P 4 0.03).

Population Relative Risk

Table I shows the sex- and age-standardized risk ofMA in relatives. The population relative risk amongchildren in nuclear families was similar to the esti-mated population relative risk of MA in first-degreerelatives assuming an autosomal-dominant mode of in-heritance. The population relative risk tended to de-crease among first-degree relatives outside nuclearfamilies and further among second-degree relatives.Both first- and second-degree relatives outside thenuclear families had a statistically significant lowerrisk of MA than expected, assuming an autosomal-dominant mode of inheritance.

Mean Age and Age at Onset

Table II shows the mean age and mean age at onsetof MA. The mean age at onset was significantly lowerin the third than in the second generation (P < 0.05).

Aura Characteristics

The distribution of various aura symptoms is shownin Table III. The three generations had similar aurasymptoms. Visual aura was experienced by nearly ev-eryone. Sensory, aphasic, and motor aura were succes-sively less frequent.

TABLE I. Sex- and Age-Standardized Risk of MA in Relatives

No. affectedPopulation

relative risk

Observed(O)

Expected(E)

Estimated(O/E) 95% CI

Assuming autosomal-dominant inheritance 6.33

Children in nuclearfamilies 36 6.23 5.78 4.46–7.50

First-degree relativesoutside nuclear families 10 2.98 3.36 1.72–5.46

Second-degree relatives 5 2.21 2.26 0.77–4.71

TABLE II. Mean Age and Age at Onset for MA inThree Generations

Mean age (range) Age at onset (range)

Generation 1 61.3 (39–83) 15.2 (3–35)Generation 2 35.0 (10–58) 17.8 (4–50)*Generation 3 20.6 (10–37) 9.4 (5–16)*

*P < 0.05.

396 Ulrich et al.

DISCUSSIONMethodological Considerations

Probands and first- and second-degree relatives wereinterviewed by physicians, since migraine assessed byproband report is not valid [Russell et al., 1996b]. Thediagnoses were made according to the criteria of theInternational Headache Society [Headache Classifica-tion Committee of the International Headache Society,1988], with the exception that first- and second-degreerelatives had no physical and neurological examina-tion. However, migraine diagnoses in the general popu-lation are rarely affected by this [Rasmussen andOlesen, 1992b]. We refrained from including relativesbelow age 10 years, because of the diagnostic uncer-tainty for children. The probands were selected from aclinic population and not from the general populationas in our previous studies [Russell and Olesen, 1995,1996; Russell et al., 1995a,b; Russell et al., 1996a,b].The nuclear families were selected with illness in twogenerations, and at least one affected and unaffectedoffsprings since some families could have an autoso-mal-dominant mode of inheritance. This excludesnuclear families with exclusively affected or unaffectedoffsprings. However, this exclusion is equally positiveand negative, since only one of the parents was af-fected.

RESULTS AND DISCUSSION

Our main result indicates that MA is unlikely tohave an autosomal-dominant mode of inheritance.First, the population relative risk in first-degree rela-tives outside nuclear families was significantly lowerthan estimated (6.33) given an autosomal-dominantmode of inheritance. Second, the population relativerisk tended to decrease further in second-degree rela-tives. The risk among second-degree relatives was ex-pected to be similar to that of first-degree relatives,since only second-degree relatives of affected first-degree relatives were included in the study. Third, thenumber of affected males and females was unequal.Autosomal-dominant inheritance with reduced pen-etrance is not likely, because of the unequal proportionof affected first- and second-degree relatives outsidenuclear families. Sex-influenced inheritance may beconsidered because of the female preponderance of MA.If so, females should have MA 10 times more frequentlythan males based on allele frequencies of 0.15 and 0.85,

respectively (prevalences: males 6.3%, females 9.6%)[Russell et al., 1995a]. However, we found that femalesonly had MA twice as often as males. Thus, sex-influenced inheritance is unlikely. Mitochondrial andX-linked inheritance is excluded in most of the familiesbecause of paternal transmission. Autosomal-recessiveinheritance is unlikely due to the unequal sex distri-bution. Our results are suggestive of multifactorial in-heritance and thus support our previous segregationanalysis [Russell et al., 1995b].

The earlier onset in the third than in the first andsecond generations can be explained by a lower meanage in the third generation. Onset of MA most oftenoccurs in the first four decades of life [Selby and Lance,1960; Rasmussen and Olesen, 1992a]. Thus, a higherproportion of relatives in the third than in the first andsecond generations may later in life develop MA. Thedistribution of aura symptoms was very similar acrossgenerations and did not differ from that of the generalpopulation [Russell and Olesen, 1996].

We conclude that MA most likely has a multifactorialinheritance even in these high-risk families with MA.

REFERENCES

Headache Classification Committee of the International Headache Society(1988): Classification and diagnostic criteria for the headache disor-ders, cranial neuralgias and facial pain. Cephalalgia [Suppl] 8:1–96.

Rasmussen BK, Olesen J (1992a): Migraine with aura and migraine with-out aura: An epidemiological study. Cephalalgia 12:221–228.

Rasmussen BK, Olesen J (1992b): Symptomatic and nonsymptomaticheadaches in a general population. Neurology 42:1225–1231.

Russell MB, Olesen J (1995): Increased familial risk and evidence of ge-netic factor in migraine. Br Med J [Clin Res] 311:541–544.

Russell MB, Olesen J (1996): A nosographic analysis of the migraine aurain the general population. Brain 119:355–361.

Russell MB, Rasmussen BK, Thorvaldsen P, Olesen J (1995a): Prevalenceand sex-ratio of the subtypes of migraine. A population based epide-miological survey of four thousand 40 year old males and females. IntJ Epidemiol 24:612–618.

Russell MB, Iselius L, Olesen J (1995b): Investigation of inheritance ofmigraine by complex segregation analysis. Hum Genet 96:726–730.

Russell MB, Rasmussen BK, Fenger K, Olesen J (1996a): Migraine withoutaura and migraine with aura are distinct clinical entities: A study offour hundred and eighty-four male and female migraineurs from thegeneral population. Cephalalgia 16:239–245.

Russell MB, Fenger K, Olesen J (1996b): The family history of migraineand allied vascular headache. J Neurol Psychiatry 23:23–32.

Weiss KM, Chakraborty R, Majumder PP, Smouse PE (1982): Problems inassessment of relative risk of affected individuals. J Chronic Dis 35:539–551.

TABLE III. Percentage Occurrence of Specific Aura Symptoms in Three Generations and in the General Population

Generation 1 Generation 2 Generation 3 General populationa

N 4 34 (%) CI 95% N 4 41 (%) CI 95% N 4 7 (%) CI 95% N 4 163 (%) CI 95%

Visual aura 88 73–97 100 91–100 100 59–100 99 96–100Sensory aura 46 27–62 39 24–55 29 4–71 31 24–39Motor aura 12 3–27 10 3–24 29 4–71 6 3–10Aphasic aura 21 9–38 21 9–36 43 10–82 18 12–25

aFrom Russell and Olesen [1996].

Autosomal-Dominant MA in 31 Families 397