It’s a bit of both and other things.

You’re supposed to be able to identify a pill from its appearance. This can be done by a combination of colour, shape, size and symbols imprinted on it or something printed on the side of a capsule. At an individual person level it means that if you’re taking more than one type of pill at a time and pop them into your hand, you can see that you have the number and type you expect. At a bigger level, random pills could potentially be identified from a database

You’ve then got other things like the active drug. 1g of paracetamol is a standard dose for an adult, so you can’t go smaller than the volume of that amount. But you then have to add in any preservatives, binders etc needed

Generally you want tasteless, but it’s not always possible, even with pills. There might be a constituent of that has a horrible flavour so a nicer flavour is put in to mask it (eg in liquid drugs for kids). Other times there will be something that tastes horrendous and there’s no chance of covering it (eg 5mg soluble prednisolone tablets taste horrendous even if swallowed whole and seems it can’t be covered). Then you have a coating that might be needed. Some pills will need to release their contents after they’ve left the stomach and are in the intestines, so need a different coating than a simple water soluble one, and it might have a butter or sweet taste.

Colour is another thing. There’s a fair amount of research about the placebo effect of the colour of the pill. Such as yellow or red pills seeming to have a stimulant effect and blue pills are sedatives. This is with placebos. So if you are making a pill, it makes a bit of sense to match its colour to this research. But many manufacturers don’t do this, so it’s not taken into consideration often

No prob. There was something niggling at the back of my head that type 2 was higher than type 1 in identical twins both getting it, so I needed to check it myself.

Hi. This is a random add on about the identical twin thing. Did you mean in the study that this article was about? Or do you mean in general, because it’s only 30-50% of the other identical twin who goes on to develop DM1 after the other has it (90% if it’s type2)

Thanks. Had a re-read of the article and saw the bit about checking the autoantibodies in the participants and were in the early stages of diabetes.

Based on a trial of 76 people, with a delay of a few years and no mention of possible side effects, and you’re expected to pay nearly $200,000?

I couldn’t see in the report. Did all those at high risk go onto develop diabetes?