This kind of study from singapore would produce a very big data bias or wrong perception. For example, Singapore's Chinese population are from very small geography regions of their ancestral home mostly from south and south-east costal regions, see the picture. These regions are genetically very closed (less mixed with the rest of China). Study of Sigaporian Chinese would give an unreliable result regarding Chines in general.
This kind of data bias and faulty perception can be seen elsewhere. For example, northern Europeans can be close to 10cm taller than southern Europeans on average. A statistic taken in one European country will certainly give the wrong result when applying caucasian in general.
Indian is also a very diverse population. If Singaporian Indians are from a small region of India, the result of the study would be missleading as well. Only study of Malay is more reliable because they are local and same as Malaysians just over the border.
Similar study has been made in Taiwan and BMI standard officially adjusted in the same way. Guess what, Taiwan's 60% or more population are from Fujian province over the straight. Mainland China however does not set a different BMI standard officially, nor is a study conducted because it is meaningless for a continental sized country.