Oxford Vegetarian Study


The Oxford Vegetarian Study (OVS), also known as the Study of Cancer in Vegetarians, began in 1980; 11,040 participants were recruited through the Vegetarian Society of the United Kingdom, publicity in local and national media and by word of mouth between September 1980 and January 1984. Participants joined the study by voluntarily completing and returning a diet and lifestyle questionnaire. The aim of the study is to investigate the long-term health of vegetarians and comparable non-vegetarians, with particular interest in cancer risk and mortality. The design is an observational cohort, based on recruiting healthy volunteer participants and following their long-term health through linkage to NHS information on incident cancers and causes of death. Participants in the study were flagged with the Office for National Statistics (ONS) and the cancer registries. Analysis of data arising from the study, occasionally in combination with data from the EPIC-Oxford study, has led to many publications in peer reviewed journals such as the British Medical Journal and British Journal of Cancer.

Data were initially collected by questionnaires on diet and lifestyle, returned by post in 1980-1984 by all 11,040 participants. Blood samples were collected by post in 1984-1986 from 3,773 participants and plasma prepared and stored; assays were conducted and papers reporting the findings were published between 1987 and 1992, and the plasma samples have subsequently been destroyed. Food diaries (four day records) were collected by post in 1985-1986 from 5,551 participants.

In 1994, we wrote to all surviving participants in the Oxford Vegetarian Study, informing them of the findings to date and enclosing a list of published papers, and inviting them to join the European Prospective Investigation into Cancer (EPIC)-Oxford study. Our Privacy Notice for the Oxford Vegetarian Study can be accessed here


As well as follow-up through questionnaires, each participant in The Oxford Vegetarian Study is followed up for cancer diagnosis, and for causes of death through NHS and the Office for National Statistics. In England and Wales this linkage is done through NHS Digital. To do this, details including name, NHS Number and date of birth plus a unique study identifier are supplied by the study investigators to NHS Digital. This linkage is carried out under Section 251 of the NHS Act 2006. NHS Digital is able to provide us with information about people who may have passed away (mortality data) and cancer diagnosis data. This information includes month and year and cause of death, and month and year of cancer diagnosis and type of cancer, supplied by NHS Digital on behalf of the Office for National Statistics and is sourced from civil registration data. NHS Digital then supply the mortality and cancer information together with the unique identifier with no names or other details back to the study. The information provided will not be linked back to the participant record, but stored in a database which will be used for analysis, where an individual participant cannot be identified. For participants in Scotland equivalent data linkage is performed Public Benefit and Privacy Panel for Health and Social Care and for Northern Ireland this linkage is done by the Central Services Agency.

When we receive the mortality and cancer information, we do not see any identifying information and do not link it to identify the person. All subsequent analyses use only subsets of de-identified data. The data will be held only at the Cancer Epidemiology Unit at the University of Oxford. No data is released or shared in any form that would enable individual participants to be identified.

Participants can request more information about how we use their data and also find out about their options to withdraw consent for us to use their data in our FAQ section.

Data gathered on the cohort from 1980 to 2016 has resulted in 18 publications in peer reviewed journals as listed in the OVS publications section.

Results arising from the most notable publications have shown that:

  • cross-sectional analyses of study data and assays using the plasma samples showed that vegans had lower total- and LDL-cholesterol concentrations than did meat eaters; vegetarians and fish eaters having intermediate and similar values (Thorogood et al, 1987)
  • plasma total cholesterol was strongly positively associated with the Keys score (Thorogood et al, 1990)
  • meat and cheese consumption were positively associated and dietary fibre intake was inversely associated with total cholesterol concentration in both men and women (Appleby et al, 1995)

  • Previous Article
  • Next Article
Copyright © EPIC-Oxford