The height data focuses on people aged 20 to 49 years old. Below age 20, height growth may still occur and above age 50 people start physically shrinking. Measured height is preferred over self-reported height as evidence suggests that respondents tend to overestimate their own stature (Gorber et al., 2007). This self-reporting bias varies according to age, sex, education, mode of interview, and purpose of the survey. Data from a recent systematic review suggests that unweighted average over-estimation of height from self reports by general adult male and female populations may be roughly about 1 cm in both cases (Gorber et al., 2007). When height of age cohorts aged 20-24 years old is compared to that of those aged 45-49 years old to examine how recent adult height has been changing, some or all of this measurement error may be removed by the differencing. Where possible, data was obtained directly from specialised official health surveys.