Sir,

We would like to thank Dr Elgohary for the interest in and comments on our paper.1 The evidence for the relative importance of the immunomodulatory effects of oestrogen and progesterone remains inconclusive. In specifying the exact date of onset of acute uveitis, we were sensitive to the problem that there may be a short delay between immune ‘trigger’ and onset of symptoms. In view of this, it would be difficult to draw firm conclusions about the hormonal influence on particular days within the late phase of the menstrual cycle, at which time hormonal levels change precipitately. We can confirm that no patient was using concurrent oral steroid or immunosuppression at presentation.

We are also grateful for the comments on our paper1 by Leung and Lam. While we understand and appreciate the points made, we disagree with most. Almost all regularly menstruating women have excellent recall of the date of last menstrual period (LMP). Where there was doubt, the patient was not included in our study; there was therefore no recall bias on this parameter. We commented that it may be interesting to measure oestrogen and progesterone levels in such patients; however, the absence of these data does not affect the validity of our comments. Our proposition is not that uveitis commences at a particular hormone level, but that it may be precipitated by hormone withdrawal in the predisposed. It is not necessary to measure hormone levels to prove that a woman is in the late phase of the menstrual cycle, if the LMP date is known.

We strongly disagree that it is ‘not unusual’ for anterior uveitis to be relatively silent in the early phase of disease; all our patients had acute-onset anterior uveitis with clear memory of the date of onset of symptoms, and rapid attendance at our ophthalmic emergency department, usually within 24 h. No patient had acute-on-chronic disease and no patient had recently discontinued topical steroid. Where there was doubt, the patient was excluded. We disagree that the aetiology of uveitis could affect presentation; while uveitis such as that related to juvenile idiopathic arthritis (JIA) can indeed be asymptomatic, that is not relevant to acute-onset symptomatic uveitis presenting in adulthood. We can confirm that no patient had JIA-related uveitis. We disagree that measurement of objective signs of severity might be a more reliable indicator of onset than reported symptoms; there is a wide variation in the severity of inflammation in those presenting with acute anterior uveitis; there is no evidence that this has a bearing on the duration of inflammation before presentation. On the contrary, it is not unknown for patients with recurrent anterior uveitis to be aware of recurrence (and to present with symptoms) before cells are detected in the anterior chamber. Finally, although the incidence of uveitis in the premenstrual phase did not quite reach statistical significance, the incidence in the whole postovulatory phase did; having carefully stated potential sources of error, we believe that our conclusions remain valid.