A community based case-control study was conducted in all six villages of
Gudjig Subdistrict, Soong Noen District, Nakhon Ratchasima Province, Thailand.
Although the subdistrict is located in a rural area, there has been a high prevalence
of hypertension over the past few years (Gudjig Primary Care Unit, 2009).
Cases were persons > 35 years whose systolic blood pressure was >140 mmHg, and diastolic blood pressure >90 mmHg, following WHO-ISH hypertension guidelines (WHO and ISH, 2003) and who had been taking antihypertensive medication for at least 3 months according to the medical records of the Gudjig Primary Care Unit. Cases were sampled proportionately by a simple random sampling from all 6 villages. Controls were persons aged > 35 years whose systolic blood pressure was < 120 mmHg and whose diastolic blood pressure was < 80 mmHg, sampled from people who lived in the same house or the nearest house to the subject. In the case where there was more than one control in the house, a control was sampled using simple random sampling. Sample size was calculated using an unmatched case control formula with Epi Info (2002) software, using the proportion of cases and controls with a body mass index (BMI) among cases and controls (8.90% and 8.15%, respectively), an expected odds ratio for high body mass index (4.01), a 95% confidence level and 80% power for the study (Mitzumoto, 2004).