I am using SPSS to analyze measurements of systolic and diastolic blood pressure in late adolescence. I want to make a binary hypertension variable, so I need to calculate percentiles for SBP and DBP. However, I need to adjust these for height, age, and gender. Is there a way I can easily do this in SPSS?
If you are proposing defining hypertension as an unusually high blood pressure, I can see two problems.
The first is that hypertension is not defined by how unusual the BP is but by its adverse impact on the person's risk of cardiovascular disease. BP levels that are statistically "normal" may carry very significant risk. Likewise, BPs that are statistically "abnormal" – such as the lowest percentiles of BP – carry no such implication.
Without a health outcome to define hypertension, you risk simply compromising your statistical power by dichotomising variables that are measured on a continuous scale. Categorising continuous variables begins to decrease study power appreciably when there are fewer than 10 categories, and by the time you dichotomise you can lose about 40% of your effective sample size. Plus you have a definition of hypertension that is not linked to a health outcome and is specific to a particular population.
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Physiological Processes
- Stress