June 23, 2021

expert reaction to observational study on hormonal contraception and glaucoma

An observational study published in The British Journal of Clinical Pharmacology looks at hormonal contraceptive use and risk of developing glaucoma in women of reproductive age.

Dr Sarah Hardman, Director of the Clinical Effectiveness Unit of the Faculty of Sexual and Reproductive Healthcare (FSRH), said:

“In a study like this, we don’t have information about all the factors that might contribute to glaucoma occurring – these might differ between individuals that used or did not use hormonal contraception and lead to confounding: it is difficult to be certain that the effect seen is caused by the contraception rather than by other unmeasured factors. The results of other studies on this subject have been inconsistent, with some suggesting increased risk of glaucoma and some indicating no increase. The data are too limited for us to draw definite conclusions as to the relative effects that different hormonal contraceptives might have on risk of glaucoma, thus we can’t make any recommendation as to which contraceptive method to use to minimise glaucoma risk.

“It is possible that use of hormonal contraception does genuinely increase glaucoma risk, but the incidence of glaucoma in women of reproductive age is extremely small – in the sample of almost 5 million women in this study, only 2366 cases of glaucoma were observed – that is <0.05% of the population. This means that even if the risk was doubled by use of hormonal contraception, the chances of an individual developing glaucoma would remain very small.”

“Any small increase in glaucoma risk would have to be weighed against the very significant benefit of achieving effective contraception to avoid unplanned pregnancy and non-contraceptive benefits such as management of heavy menstrual bleeding.”

Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:

“There are two features of this study that mean that it shouldn’t be taken as alarming by women who use hormone contraceptives. The first is that it’s an observational study. The women involved weren’t asked by the researchers to use, or not use, hormone contraceptives. Indeed the data were collected generally before these researchers were involved at all. The study is based on records of medication and of the health of the women. So there were many differences between the women who used hormonal contraception, and those who didn’t, and also differences between women who used different contraceptive types, other than their contraceptive use. Any differences in diagnoses of glaucoma between women who used or didn’t use hormonal contraceptives might have been caused, entirely or in part, by some of these other differences, and not by their contraceptive use. The researchers did make some statistical adjustments to allow for some of these differences, on which they had data, and they also used a statistical method that matched women on the basis of their age, their body mass index (BMI), and when they entered the study. The association that they found between hormonal contraceptive use and glaucoma still remained after these adjustments. However, that still does not show that the higher rate of glaucoma diagnoses in women who currently used hormonal contraceptives must be caused by the contraceptive use. For one thing, the researchers couldn’t adjust for differences on which they had no data, and as they point out, that includes whether the women smoked. So a study like this can’t establish that using hormonal contraceptives causes glaucoma, or even that they contribute to causing a higher glaucoma risk. That certainly is one possible explanation, but there are other possible explanations.

“Perhaps more importantly, though, the risk of glaucoma in women like those in this study is low, as the research paper and the press release are careful to point out. Over the average two years’ follow-up time, only about five women in every ten thousand in the study had a glaucoma diagnosis. To investigate what effect the use of hormonal contraception might have on that low risk, the researchers temporarily made an assumption that the increased risk of glaucoma in those who used hormonal contraceptives was entirely caused by the contraceptives, beyond the adjustments they had made to allow for possible effects of the other differences for which they had made adjustments. As I’ve said, that’s quite a strong assumption, because the study can’t establish cause and effect and there may be other differences that come into the pattern of cause and weren’t adjusted for. But, under this assumption, the researchers calculated that 2.6% of glaucoma cases in this population could have been prevented if hormonal contraceptive use were eliminated completely. That is, if that assumption about cause and effect were true, completely getting rid of hormonal contraceptive use in this population would get rid of about 1 in 40 of the glaucoma cases, so, very roughly, one or two glaucoma cases for every 100,000 women in the population. If other factors than hormonal contraceptive use also come into the causation of glaucoma, that reduction will be even smaller.

“The advice in the press release, that women taking hormonal contraceptives who experience visual changes should have the symptoms checked out by an ophthalmologist, clearly makes sense anyway, despite the low risk and the doubts about cause and effect – though surely it would make sense for anyone (male, female, regardless of contraceptive use) to have visual changes checked out properly.”

‘The association between hormonal contraceptive use and glaucoma in women of reproductive age’ by Kate Hogden et al. was published in The British Journal of Clinical Pharmacology at 5.01am UK TIME on Wednesday 23 June 2021.

DOI: 10.1111/bcp.14920

Declared interests

Prof Kevin McConway: “I am a Trustee of the SMC and a member of its Advisory Committee.  However, my quote above is in my capacity as an independent professional statistician.”