Background
Hypertension is one of the leading risk factors to health. Although the burden of hypertension is concentrated in older populations, it is now recognised that the origins of hypertension are rooted in the early stage of life. Elevated blood pressure in youth is a strong predictor of hypertension in adulthood. Previous studies found that life style factors, for example diet, can affect blood pressure status, but relevant studies in adolescents are lacking.
Carrying out epidemiology research online rather than using traditional face-to-face methods is becoming a novel method worth studying. These new methods, including recruiting participants through social network websites, completing questionnaire or dietary record online, etc., will allow researchers to obtain sample in a wider geographic area with lower costs in time and money, are more convenient and are more acceptable for participants to collect sensitive information.
Aims and objectives
The aims of the study it to develop a pilot online research method to explore the association between diet and blood pressure in adolescents. There are two main objectives: a) to find out key dietary factors that may affects blood pressure levels in adolescents using existing data, and b) to explore the feasibility of use new technology (for example online recruitment, online questionnaire, online dietary record, blood pressure measurements, etc.) in epidemiology studies.
Methods
Existing data was used to analyse the association between diet and blood pressure in 10 to 19 years old adolescents. National Diet and Nutrition Survey (NDNS) dataset was used as cross-sectional data, and Avon Longitudinal Study of Parents and Children (ALSPAC) dataset was used as longitudinal data. Data was cleaned and transformed. Linear regression, logistic regression and survival analysis were used to explain the association between diet and blood pressure levels.
The Online Research of Adolescents and Good Eating Study (ORANGE Study) was set up to explore the feasibility of online survey. It included two stages:
School-based stage: 13 to 19 years old participants were recruited from local school students, visiting students on university open-days and first year students in the University of Leeds. Questionnaires were used to find out the adolescents’ attitude to an online health research. How willingly they were to click on the advert and take part an online health research, which factor would affect their attitude were asked in the questionnaire. The accuracy of self-measured blood pressure values using a self-designed instruction was also tested in this stage.
Web-based stage: 579 participants were recruited through social network websites (Facebook and Instagram). Fifty participants were randomly selected to complete the whole study. They received blood pressure measuring tools by post. They were then asked to measure their own blood pressure, complete an online questionnaire and a 3-day online dietary record. The following factors were calculated and compared to those of the school-based stage to evaluate the feasibility of online survey: response rates, completion rates, time and money expenditure and sample characters.
Results and conclusion
7%~10% of the adolescents were hypertensive in NDNS and ALSPAC population. Dietary vitamin E intake was linked with lower blood pressure levels in crosssectional dataset, but not in longitudinal dataset. Dietary fat intake was linked with higher blood pressure levels in both cross-sectional dataset and longitudinal dataset.
Self-measured blood pressure of the adolescents showed acceptable accuracy when comparing to standard blood pressure values that measured by the researchers. In the ORANGE study, 89% of the participants used Facebook in the last week, 21% of the participants would like to click on the advert of online survey, and 32% of the participants would like to take part in the survey. In the pilot online study, response rate of the web-based sample was a bit higher than school-based stage. Compared with traditional school-based study, the time expenditure per valid participants on the webbased study is less, but the money expenditure is higher. Compared with national data (NDNS population), in both school-based stage and web-based stage, fewer participants under 16 years were recruited, fewer male participants were recruited, and more white participants were recruited. In summary, the web-based study expensed more money but less time, and the sample recruited is no worse than that in school-based sample.
