Search

IntroductionNoncommunicable diseases (NCDs) are among the leading causes of death and contributes to approximately 70% of all deaths worldwide. Elevated body mass index (BMI) is recognized as a major risk factor for NCDs. Several studies have determined various BMI cut-offs for classifying dyslipidemia, hypertension, diabetes or at least one cardiovascular risk factor (CRF) 2,3,4 . • For Malaysian adults, the proposed optimal BMI cut-off points for the risk of diabetes mellitus, hypertension and hypercholesterolemia varied from 23.34kg/m2 to 24.14kg/m2 for men and from 24.04kg/m2 to 25.4kg/m2 for women3 . However, they were determined based on data collected in 2006, more than a decade ago. ObjectiveThe aim of this study is to verify the sensitivity and specificity of these cut-off points for the Malaysian adult population using data from a recent national survey Methods[Refer to Results]. ResultsThe total number of respondents aged ≥ 18 years was 10,472 (4785 males and 5687 females). Majority aged between 31 to 59 years. 51.0% Malay, 21.5% Chinese, 11.1% Other bumiputras, 10.6% other ethnic groups and 5.8% Indian. The prevalence of BMI above 23.0 kg/m2 was 66.0% (BMI range from 14kg/m2 to 55kg/m2). Discussion / ConclusionOur findings indicate that the proposed BMI cut-offs (23.3 to 24.1 kg/m2 for men and 24.0 to 25.4 kg/m2 for women) correctly identified more than 67% of those with NCDs and correctly identified more than 39% of those without.The BMI cut-off points proposed by the previous study showed acceptable sensitivity but relatively low specificity.Therefore, we suggest that the use of these proposed BMI cut-offs for classification of overweight among Malaysian adults should be revised periodically to increase its sensitivity and specificity for the NCDs screening and weight management programs.

Background Reliability studies are commonly used in questionnaire development studies and questionnaire validation studies. This study reviews the sample size guideline for Cronbach's alpha test. Methods Manual sample size calculation using Microsoft Excel software and sample size tables were tabulated based on a single coefficient alpha and the comparison of two coefficients alpha. Results For a single coefficient alpha test, the approach by assuming the Cronbach's alpha coefficient equals to zero in the null hypothesis will yield a smaller sample size of less than 30 to achieve a minimum desired effect size of 0.7. However, setting the coefficient of Cronbach's alpha larger than zero in the null hypothesis could be necessary and this will yield larger sample size. For comparison of two coefficients of Cronbach's alpha, a larger sample size is needed when testing for smaller effect sizes. Conclusions In the assessment of the internal consistency of an instrument, the present study proposed the Cronbach's alpha's coefficient to be set at 0.5 in the null hypothesis and hence larger sample size is needed. For comparison of two coefficients' of Cronbach's alpha, justification is needed whether testing for extremely low and extremely large effect sizes are scientifically necessary.