Sarah Lewington, Professor of Epidemiology and Medical Statistics and Director of Graduate Studies (taught-courses)
19 November 2020
Sarah has been with the department for nearly three decades now. As Director of Graduate Studies, she enjoys meeting new students and believes our diversity at the workplace is 'the best thing.'
15 minutes with Sarah Lewington
The ’15 minutes with…’ blog series aims to provide a glimpse into the career paths and personal lives of the researchers and employees in our department and showcase the wide range of roles and science conducted. In this blog series, we ask researchers, non-research staff and students, nominated by the students themselves, to reflect on their career to date – their proudest accomplishments and their lowest points – and how they came to be where they are today.
We will publish on a regular basis in the coming months – we hope you will enjoy it.
Today, we continue this series with an interview with Sarah Lewington, Professor of Epidemiology and Medical Statistics and Director of Graduate Studies for taught courses.
Could you describe your current role in one sentence?
I am an epidemiologist and statistician leading a group that does global health research, and I’m also the Scientific Director for the MSc in Global Health Science and Epidemiology.
What is your favourite thing about it?
I think diversity is the best thing. I love the science and I really enjoy the people I work with, who are so enthused and committed to what they’re doing. I’m lucky to work with such a range of people – statisticians, programmers, medics, and policy people. As Director of Graduate Studies, I love meeting the students – they give a real buzz to the department. It’s always really interesting to hear where they come from and why they decided to do a Master’s in Global Health Science and Epidemiology.
...and what is your least favourite thing about it?
Jetlag! I love visiting collaborators in their home environment, but I am not good (ie grumpy) if I don’t get eight hours of sleep so I generally suffer terribly from jetlag after a long-haul economy flight. More seriously, I worry about raising money and it’s frustrating to be spending precious writing time on grant applications instead of scientific papers. But, it’s a reality because we cannot do any research without the funds to do it.
Moving on to the science, could you tell us about what science you’re most excited about or interested in in your own work?
The work I do is based around the classical (or some might call them boring!) risk factors: smoking, alcohol, blood pressure, and adiposity. I think what excites me is how important they still are, and how little we actually know about them outside high-income countries in terms of epidemiology. There’s quite a lot of data on the prevalence of smoking in low and middle-income countries (from great surveys like the Global Adult Tobacco Survey), so we know quite a lot about that, but one of the exciting things was discovering a link between smoking and TB in South India. We’d never have seen that in a high-income country: there’s not enough TB.
People think ‘Why can’t you just apply the findings from high-income countries? Does it matter where you do the study?’ The studies we do show that sometimes it does matter because these associations do differ and the relative importance of risk factors differs depending on circumstances. In Mexico, for example, many people smoke intermittently and casually – you can buy a single cigarette on a street corner in Mexico. Where else in the world can you properly study the effects of occasional smoking? You certainly couldn’t look at it in the UK, as people don’t tend to smoke like this in the UK.
And then turning it into policy – how can you make sure people live as healthy a life as possible, or at least give people and governments the information they need to decide where they’re going to put their money policy-wise?
And what about the science that you’re most interested in beyond your own work?
Probably public health. I’m an epidemiologist, so I don’t really do public health, but I feel passionately about public health – improving access to healthcare and therefore improving health outcomes. Not just globally, but within the UK- in one UK city there is an eight-year gap in life expectancy between a wealthy and poor area of the same city.
Outside the medical field, I’ve always been interested in education. It would be interesting to apply some of what we’ve learned in medicine, such as about the need for randomised evidence. I’m sure there is loads of really good research done in education, but it’s a bit like epidemiology in that if you change something, it takes 10 years to have an effect. In a school, changing something now is unlikely to affect the kids doing their GCSEs (UK exams at age 16) next year – it’s going to affect the kids coming in who will experience it for several years. Unless there is something desperately wrong in a school or education system, you’re tweaking it. You’re making improvements, but they’re not massive effects, so you need to study them carefully and systematically.
Another one is the criminal justice system. The aim of prison should be to reform, but does it reform? Or does it just punish? Does it deter people from doing it again? I’m interested in using what we’ve learned in epidemiology and medicine in these other fields, and presumably the same would be true going the other way too.
What’s the thing that you are most proud of in your professional life?
I was incredibly proud to get my Chair. I had a very unassuming beginning to my academic career. In school, I was definitely not a high achiever so I think the idea when I left school that I would eventually be a professor at Oxford University would have been laughed out of the park! I was at a comprehensive school in Wigan in northwest England until I was 16, so I would really like to go back and say to those kids that anything is possible if you work hard.
I wasn’t the brightest – I worked hard and then I found something I was passionate about. I wasn’t passionate about my O Levels (previous UK exams at 16). I did my homework and I did it well, but I didn’t have any great interest. I liked maths when I was doing A Levels (UK exams at 18), but I didn’t love it at that stage. And then I went to university and found I really liked statistics. I hadn’t done statistics at school, but I really enjoyed it and then did a module in medical statistics in my final year and thought “Wow, this combines what I’m good at with what I’m interested in!”
Academically, I’m very proud of the Prospective Studies Collaboration, which was the first project I worked on and now has more than 6000 citations. We didn’t know it was going to be that successful or that important. I came to Oxford in 1995 as a statistician to work on it, and I also asked if I could do a PhD working on it – in those days you couldn’t formally do them part-time, so as well as doing it for my job, I worked on the writing-up in my spare time.
We brought together data from 61 prospective studies on a million people. This was back in the 1990s before anyone had heard of UK Biobank (and was part of the inspiration for UK Biobank, by showing that you needed huge numbers of people). We brought together cohorts of thousands and tens of thousands, and it was only with data on a million people did we really see the true relationships. For example, we saw that the effect with blood pressure is a log-linear relationship throughout the range.
It’s easy to forget that in the mid-1990s people were worried about lowering blood pressure in old people, in case they did more harm than good (ie there was an inverse association). And at all ages, there was concern about lowering blood pressure too far, because in the 1990s the received wisdom was that it was a J-shaped relationship. It was about getting enough numbers, getting rid of the people who already had disease and doing the analysis correctly and then suddenly it (the J-shape) just disappeared. It’s beautiful – the image comes into sharp focus. It’s an iconic picture.
I think it changed the way people thought about the size of study needed in epidemiology, and also about how you need to think really carefully about your biases. Some of those biases can now be overcome using genetics, but that can also bring its own limitations: you don’t have the wide range, as genetically determined blood pressure is narrower, so you can see that it’s causal, but can only look in that narrow range. The Prospective Studies Collaboration was aimed at blood pressure and cholesterol, and cause-specific mortality, and we also looked at BMI, obesity and diabetes. We only had four publications, but all in The Lancet and all very highly cited. So I’m proud that we did just those four, and not 30 that nobody remembers.
But it was hard. It took almost seven years to get that publication. I was lucky this was in the 1990s when CTSU (our Clinical Trial Service Unit and Epidemiological Studies Unit) could afford to keep me on the general ledger. But even then, it was hard because you’re not competitive for fellowships without the lines on your CV early on. I think there’s a compromise that we need to think about – as PhD students and early career researchers, you need lines on your CV, even if in 10 to 15 years you look back and hope no one reads them!
These blockbuster New England Journal, JAMA, and Lancet papers take years – years you don’t have. It’s one of the reasons the department created the Early Career Research Fellow posts – population health research isn’t lab research and it does take a long time. Getting that balance right between the quality and quantity of publications can be hard to do for an individual.
And your biggest failure or disappointment?
Being rejected for University Research Lecturer (an academic title) three times! That was quite humbling, and I can laugh about it now, but the first one was genuinely “What have you done for seven years?” I did have some publications, but I’d really focused on the Prospective Studies Collaboration – which had only just been published. Later, of course, it had several thousand citations, but that was beside the point.
After that, I had three children (2001-2007) so I was working part-time. Then I applied again and they said “You’ve not done much for the last 10 years”. I think that’s the huge difference Athena Swan has made – pre-Athena Swan they were just looking at CVs in terms of “you’ve been here 15 years, what have you done?”, not at the fact that you started 15 years ago, but you’ve actually only worked 7.5 of those years.
These things are tough – they’re tough to take when you’re working hard and you are compromising your family life. You don’t do research for these awards, but it’s still hard to be told that you and your work are not worthy of them.
Could you outline your route to where you are today?
I realised in sixth form (the last two years of school) that I quite liked maths and physics. I had a female physics teacher who must have wished she’d been an engineer because she managed to persuade five of the 12 girls in my physics class to do engineering. So I went off to university to do engineering, barely able to wire a plug. I quickly realised that this was a mistake! That was probably one of the hardest years of my life. Away from home, excited to go to university and it wasn’t living up to it. Not only was I not living up to it, but I was worried that I wouldn’t be there the next year. I couldn’t carry on this engineering course, which was making me very unhappy.
I started to look around at other universities and degrees. Leicester, the university I was at, had a modular degree (which was quite unusual for the 1980s). So I changed to the modular degree, using some of the maths I’d done, and studied pure maths, statistics, computing and psychology. That was really when my academic career took off – up until then I’d got there by working hard, but then I suddenly really enjoyed the maths and really enjoyed the statistics, to the point that I became a bit of a geek. But even at the end of it, I didn’t even think of staying on in academia – it wasn’t anything that had been done in my family, so I never even contemplated it.
I went to work in the city for a year – I hated that, so that was another crunch moment of “where do I want to go with my life?” Then I saw a job advertised for a medical statistician – I thought “that was what I really enjoyed”, so I applied and worked in the pharmaceutical industry, Boehringer Ingelheim, for five years. I went on a few short courses and really enjoyed them, but I always felt I was missing the basics. I felt like I was building without foundations – I was running analyses, but I didn’t have the theory.
So I took a sabbatical to do a master’s at Southampton in medical statistics, and it was there I found that I really loved epidemiology. The design of clinical trials is interesting, but as a statistician, the analysis isn’t so interesting – if they’re well designed, the analysis sort of falls out. I was writing my dissertation on meta-analysis, and as an MSc student, I rang up Richard Peto (who had written one) out of the blue to ask something. I can’t remember the details, but we had a long conversation. Afterwards, I didn’t really want to go back to the pharmaceutical industry – I wanted to go into academia. I wrote him a letter, and he said “Why don’t you come up for a chat?”. That was 26 years ago and that was how I came to be in CTSU working on epidemiology. I’ve worked with Richard since then, and it’s inspiring, it’s exciting.
Could you tell us one thing about you outside work?
I spend a lot of time at the weekends standing beside football pitches, watching girls on a Saturday and boys on a Sunday. The rest of the time I seem to spend ferrying teenagers around – and occasionally I get to go out for a drink myself!
If you could give one piece of advice to yourself what would that be?
[Laughs] Don’t do engineering!
No, I think follow your interests and believe in yourself. And work hard but have fun. I think that was one of the things I really learned at university – you can have fun and work hard. I was quite disciplined and worked hard in the day, but didn’t miss out on anything I wanted to do. I didn’t spend a lot of time in coffee shops, but I played lacrosse, I played squash, I went out, I went to gigs, and I had a full social life. But definitely: follow your interests and believe in them.
Thank you so much!
Sarah Lewington is a Professor of Epidemiology and Medical Statistics and the Director of Graduate Studies for taught courses in the department. She is also the leader of the global population studies group and a research fellow at Green Templeton College.
She is on X at @SarahLewington2.