Something tells me you are onbscuring friend's breakfast and dinner, and perhaps what goes in the salads.
A salad is better for diabetes than almost any alternative, and does reduce risk all else equal.
I don't have visibility of what he eats for breakfast and dinner. What I do know is that his overall diet has resulted in obesity.
My point is that an underweight person who consumes sugary beverages is at a lower risk than an obese person who eats salad for lunch, whereas the general public is likely to believe the opposite.
I once had an obese friend of mine tell me I was going to give myself diabetes -- despite the fact that I'm not even remotely overweight -- because I regularly drank soda. To him, his diet was good, because he ate salad for lunch and snacked from the office fruit bowl, whereas my diet was bad, because I drank soda.
There are likely many people out there who believe that because they eat salad for lunch they're "eating well" and reducing their risk of diabetes, when actually they're putting themselves at risk by being overweight.