Introduction: Consumption of confectioneries is a determinant of health risk. However, how health risk determines the intake of confectioneries remains unclear. The objective is to examine how waist circumference (WC) as a measurement of health risk influences the consumption of confectioneries among adults. The research question is that do high-risk people consume more confectioneries than low-risk people? Methods: A quantitative research design with a focus on establishing a correlation between the measurement of health risk and consumption of confectioneries was adopted. Secondary analysis of a nationally representative cross-sectional data was used. The population of interest was the Malaysian population, regardless of being obese or non-obese. Analyses stratified by body mass index (BMI) or WC were not conducted. An instrumental variable (IV) approach was used to estimate the regression of consumption of confectioneries. BMI was used as an IV for WC. In the first stage, we regressed WC on all exogenous variables, including BMI. Then, we replaced the original values of WC with the fitted values of WC. Results: BMI was an appropriate IV for WC. An additional cm of WC was associated with a 0.022 unit of reduction in the serving of confectioneries per week. The negative relationship between WC and consumption of confectioneries indicated that adults who had high health risk consumed lesser confectioneries than adults who had low health risk. Conclusion: Drawing from the IV regression results, the present study highlighted that people with high health risk, rather than people with low health risk, were less likely to consume confectioneries.

Introduction: This study examined how waist circumference (WC) varied across sociodemographic characteristics of Malaysian adults using a nationally representative data. Methods: Data from the Malaysian Adult Nutrition Survey 2014 (n = 2696) was used. In the sample, 46.55% were males and 53.45% were females. The outcome variable was WC. Waist circumference was categorised into four ordinal outcomes: very low, low, high and very high. The explanatory variables were age, monthly individual income, education, gender, marital status, ethnicity, employment status and household location. Ordered probit models were utilised. Results: The majority of males had a low WC (54.66%), and only a small proportion had a very high WC (0.96%). Similarly, a high proportion of females had a low WC (55.59%), and only 1.94% had a very high WC. Younger males were more likely to have a high WC than their older counterparts. Males who had monthly income of RM2000-2999 or RM3000-3999 were more likely to have a high WC compared with those who had monthly income of ≤RM999. Chinese and Indian males were more likely to have a high WC than Bumiputera males. For females, those who were married, widowed/divorced and Bumiputera were likely to have a high WC. Conclusion: WC was associated with various sociodemographic profiles of males and females. In particular, there were positive associations between the likelihoods of having a high WC and younger age group, higher income, Chinese, Indian, and being married. Our findings provided policy makers with better information on formulating intervention measures.