Using genetic data to understand the causes of age-related macular degeneration

Reference # RPGF1806\67
Date June 2019 - Present
Funding £268,848
Project lead Dr Reecha Sofat & Dr Valerie Kuan
Organisation University College London

Published November 2020

UCL
Summary

When investigating diseases of ageing and their causes, it can be very difficult, time-consuming and expensive to conduct randomised controlled trials. Instead, Drs Reecha Sofat and Valerie Kuan are using a technique called Mendelian Randomisation to understand the underlying risk factors for the sight loss condition age-related macular degeneration, shedding light on new ideas for prevention and treatment.

As clinicians, we felt that what was really important was that we could bring tangible findings almost directly to the public. We wanted to be able to say “if you do this, you can potentially prevent the onset of AMD”. In a way, what we’re doing now is even more direct clinically than our original plan

Dr Valerie Kuan

When investigating age-related diseases and their causes, it can be very difficult, time-consuming and expensive to conduct randomised controlled trials – the ‘gold standard’ way of testing scientific ideas – where participants are randomly assigned into groups receiving different treatments or interventions. And in many cases this may simply be unethical or impractical.

Instead, to understand the relationships between different health factors and a disease, we often use observational studies, where we compare groups of similar people who are already doing different things (such as comparing the incidence of a particular disease between smokers and non-smokers). But these kinds of studies are open to potential bias and can only tell us about correlations between risk factors and health, not causation.

For example, if a study shows that smokers are more likely to develop a particular condition, does smoking actually cause it, or do people who smoke also tend to do other things that are the true cause of the disease?

With this in mind, we’re using a technique called Mendelian Randomisation to look at risk factors for age-related macular degeneration (AMD), the biggest cause of sight loss in the UK.

Mendelian Randomisation is a technique that comes as close to doing a randomised controlled trial as possible without having to do one

Dr Reecha Sofat

Mendelian Randomisation relies on the fact that, across large groups of people, genes are allocated randomly at conception. Our genes influence our predisposition to certain diseases or behaviours. People can then be grouped based on their genetic code as a proxy for their disposition to particular diseases or behaviours – such as whether they have high cholesterol or are a smoker.

This is similar to randomising clinical trial participants, with their genetic code being the one variable that you’re testing – equivalent, for example, to a drug that participants would otherwise be randomised to, only in this case the genetic code is used as a proxy for the drug. These groups can be statistically analysed to see whether having a particular genetic variation – for example, those who are less likely to smoke – are more or less likely to have a particular condition, such as AMD. If AMD is equally likely in all groups, then it suggests that smoking isn’t a root cause of the disease. But if the genetic variation that influences smoking is associated with AMD using Mendelian Randomisation, this means that we can be more confident that smoking is a cause, because this is the only variable that is different between the groups. Any other genetic and lifestyle differences between the groups are evened out across all the groups due to the large number of participants in the study.

A change of plan

We were originally planning to investigate the role of different types of fat in the blood on AMD, but another research group published similar findings before we could even start our analysis. In science, it’s not unusual that others will be working on similar questions and there’s always a chance you can be ‘scooped’, so we had to come up with a ‘Plan B’.

We decided instead to look at a range of modifiable risk factors for AMD. As clinicians, we felt that what was really important was that we could bring tangible findings almost directly to the public. We wanted to be able to say, “If you do this, you can potentially prevent the onset of AMD”.

When we spoke to the Dunhill Medical Trust they were really supportive about our change of plan and recognised that it was a good way to address the problem. In fact, what we’re doing now is of even more direct clinical relevance than our original proposal.

Close-up model of DNA double helix on a black background
By using people’s genetic code as a proxy for their disposition to certain behaviours or physical characteristics, we’re able to identify risk factors for the development of AMD. If these risk factors are modifiable (behaviours that can be changed, for example) we can potentially prevent people from developing AMD in the first place.
Findings that can change public health messages

So far, we’ve completed the work investigating modifiable risk factors associated with the development of AMD. Our analysis is reaching the final stage and we hope to publish these findings very soon. We think these will demonstrate an important public health message around AMD.

In terms of future research, the holy grail is larger data sets

Dr Valerie Kuan

In the remaining time of our project, we’re going to use Mendelian Randomisation to investigate the role that different biological pathways play in the development of AMD. We’ll then use all of this information to identify potential drugs that could be repurposed to prevent AMD from developing in the first place. In terms of our future research, the ‘Holy Grail’ is larger data sets. We would love to have access to larger data sets that allow us to investigate other causes of AMD and identify targets for prevention or treatment to help save people’s sight.

Share: