1350

times We have tested the hypothesis that brain oedema might bedetectable during migraine headache, using MRI.Four women aged 23-51 with common migraine for a mean of 13

years volunteered for study. MRI was done between 2 and 3 h afterthe onset of a typical attack: each patient had severe headache(predominantly left sided in 2 and right sided in 2), accompanied bynausea and photophobia. Analgesia was withheld until after MRI.MRI was done again within 1 month, avoiding the week ofmenstruation: no patient had had an attack in the previous 7 days.MRI was done at 0-26 T, and multiple axial images were obtained

with sequences from which T and T calculated images wereproduced with the manufacturer’s algorithms. From the two MRIexaminations, matching sets of images were selected at the level ofthe frontal horns. From each set, brain and hemisphere dimensionsand T, and T2 for frontal and occipital grey and white matter weremeasured. A screen cursor was moved around the brain region ofinterest and the mean of three readings was obtained. The resultswere analysed with the Mann-Whitney test for non-parametricdata.No focal or generalised MRI abnormalities were seen in any

patient. The percentage change between scans done during andbetween attacks for all indices ranged from -4-1% to +9-8%(mean + 0-9%). Comparison of during- and between-attack resultsshowed that the greatest difference for brain size measurements was+ 2-3% (maximum antero-posterior diameter) and for Tl and T2was +2-68% (T1 of frontal grey matter). No difference wassignificant, including those between values from the hemisphereipsilateral to the headache and from the contralateral side.Although other workers have suggested that MRI abnormalities

compatible with infarcts are often seen in the brains of migrainepatients,45 none of our subjects showed on either T or T weightedimages localised abnormalities that were compatible with an infarctor other lesion. Although clinical detail is sparse in previous reports,we feel that in a young patient (under 45) with confirmed commonmigraine, the finding of high T2 areas in the brain probablyindicates a cause other than migraine, and this should be

investigated. We also conclude that pain attacks in common

migraine are probably not caused by a change in whole brain orhemisphere size, since even large space-occupying lesions rarelycause a headache resembling migraine. Nor are they likely to beassociated with a substantial increase in white or grey matter water

content, which virtually excludes localised or generalised brainoedema (either cytotoxic or vasogenic) as a pathogenetic mechanismaccounting for symptoms, including pain, in common migraine.We thank the patients for their cooperation in delaying treatment and

tolerating the MRI during their headaches, and the British MigraineAssociation for their generosity in funding the MRI studies.

National Hospital,Queen Square,London WC1N 3BG, UK

D. BARNES

J. N. BLAUD. P. E. KINGSLEYB. E. KENDALL

1. Goltman AM. The mechanism of migraine. J Allergy 1936; 7: 351-55.2. Blau JN, Solomon F. Migraine and intracranial swelling an experiment of nature.

Lancet 1985; ii: 718.3. Barnes D, McDonald WI. A magnetic resonance imaging study of experimental

vasogenic edema and its response to dexamethasone. Mag Res Imag 1988; 7:125-31.

4. Kaplan RD, Solomon GD, Diamond S, Freitag FG. The role of MRI m theevaluation of a migraine population, preliminary data Headache 1987; 27: 315-18.

5. Rothrock JF, Walicke P, Swenson MR, Lyden PD, Logen WR. Migrainous stroke.Arch Neurol 1988; 45: 63-67.

Incentive spirometry versus routine chestphysiotherapy

SIR,—Dr Hall and colleagues (April 20, p 953) attempt toevaluate postoperative chest physiotherapy versus incentive

spirometry, but their results do not support their conclusion andraise important points.

Firstly, this study perpetuates the myth that physiotherapytechniques consist solely of tip, bash, breath, and cough. In modemclinical practice more emphasis is placed on positioning to achievemaximum ventilatory function, and on early mobilisation to

increase functional residual capacity.

Secondly, their patient selection criteria include low-risk surgicalpatients (eg, appendicectomy) who would rarely requirephysiotherapy in the UK. More importantly, the very patientswhom physiotherapists would judge most likely to benefit frompostoperative therapy were specifically excluded-ie, those withpre-existing pulmonary disease.

Therefore, Hall et al merely show that chest physiotherapy is ofno benefit when it is not indicated. We would urge surgeons not todiscount appropriate physiotherapy in favour of what they may seeas a cheaper plastic alternative.

Physiotherapy Department,King’s College and Dulwich Hospitals,London SE5 9RS, UK

ANNETTE PARKERSHARON VERNE

Delayed visual maturationSiR,—Your April 20 editorial brings to mind an ophthalmic

adage-"beware of telling a parent his/her child cannot see, becauseyou can often be wrong". While this is recognition of the existenceof delayed visual maturation (DVM), it also highlights our inabilityto measure the vision of infants, and the inevitable consequence ofthis is a poor understanding of the natural history of visual disordersof childhood. Fortunately, with the advent of clinically applicabletests such as the acuity card procedure, visual acuity can now bemeasured simply and repeatedly even in the very young. Thesenon-invasive investigative aids are proving to be of practical value inthe assessment of severity of visual deficits, in monitoring of theclinical course of infant vision disorders (with and withouttreatment), and in providing fascinating insight into the basis ofearly visual function.You recognise that DVM is a clinical spectrum. We now believe it

should be extended even further.1 Independently, in two centres inthe UK and USA, we have monitored a group of children whopresented as blind in early infancy with severe and permanentocular disorders such as coloboma, optic nerve malformation(hypoplasia), or retinal abnormality (Leber’s amaurosis). Althoughthere are anecdotal reports of success, visual improvement is notgenerally believed possible in the presence of such severe oculardisorganisation. Nevertheless, of our 11 infants 8 showed somevisual improvement and the pattern of development was broadlysimilar in all. Because of the underlying ocular disorder the degreeof visual improvement was necessarily limited and the childrenremained legally blind.The value of this change for the child and family should not be

underestimated. First, despite remaining legally blind, the childgains valuable vision-mediated skills, and this vision facilitates earlyindependent mobility. Second, this information is important forthose involved in care. Third, stimulation is often undertaken forvisually impaired children, and, clearly, knowledge of the naturalhistory of these disorders is a prerequisite to the understanding ofthe success, or failure, of any such treatment. Fourth, it raises theissue of the possible influence of ocular disorders and visualexperience, normal or abnormal, on visual pathway maturation.With the spectrum of DVM becoming increasingly broad, so as

to include even infants with obvious ocular defects, a word ofcaution is warranted. There is another well-known clinical saying-"never openly anticipate to parents visual improvement, for ifwrong this causes far more anxiety and turmoil than a more cautiousapproach". This statement is true, and it is important to emphasisethat, unfortunately, far from all children who are blind in infancyimprove (3 infants of our 11 did not). Both clinical sayings quotedhere greatly oversimplify the issues. Although prediction of futuredevelopment is not possible at present, the clinician can statewhether an infant can see, and how much, and parents can beimmediately and appropriately updated.Department of Ophthalmology,Birmingham and Midland Eye Hospital,Birmingham B3 2NS, UK ALISTAIR R. FIELDER

Department of Ophthalmology,Children’s Hospital,Boston, Massachusetts, USA

D. LUISA MAYERANNE B. FULTON

1 Fielder AR, Fulton AB, Mayer DL. The visual development of infants with severeocular disorders. Ophthalmology (in press).