You may also want to see if there is a correlation between drugs use and traumatic childhood events. Many people turn to drugs because it is their only way to cope with traumatic events that they have experienced.
Interestingly, drug use has decreased in America over the last thirty years. I would have presupposed that drug use is increasing because of wide availability and improved quality, but it turns out that most dabble in marijuana for a while and then just stop. Most habitual drug users are employed (easier to afford the drugs) and many have a social drug habit, but are not addicted in the usual sense of the word.
This is not to say that drugs are not a huge problem; drug use impacts a person's health directly and then indirectly impacts their family and friends. However, the famous War On Drugs has demonstrably failed; better solutions need to be researched and implemented.
You might look into medical information. Medical charts might trace family history and a user's drug history. I am not sure how complete or accurate they would be or how you would get access to them, but that would be one way to do it.
You might also want to consider what is missing in the statistics. Is stereotyping leading to more arrests and investigations of one group than another? You might also look for commonalities in these antecedents. As far as demonstrating this knowledge goes, you will need to present concrete facts and logical conclusions drawn from these facts. You will need to be able to prove how the antecedents and correlates effect drug use without showing bias. If you're writing a presentation, a chart or some type of graph might be helpful. If you're writing a paper, it might be helpful to create some type of graph for your own research purposes. Of course, you aren't likely to include this graph in a research paper, but it would help you see where your evidence is weak. It will also help you translate facts into definable trends.
This can be a dangerous thing to do as it can lead to stereotyping. However, as #2 makes clear, it is important to look at how drug use and offense seems to be more likely to occur in certain areas of the population. You might like to consider ethnicity along with socio-economic levels of welfare and whether these two causes can actually be linked thanks to various inequalities.
I thought it might be helpful to do some internet searching for you to see what might be out there on this topic. Some of the sites I found are listed below; I hope they prove helpful. The first link looks especially useful.
You need to look at the correlates to see if there are certain demographic characteristics that can predict drug use. For example, if poverty is correlated with drug use, then you might predict that the trend will be for more drug use right now as the US economy struggles.
You need to look at antecedents to see if there are factors that make a person more likely to use drugs over time. Again, if these factors (use of gateway substances, for example) seem to be on the rise, then we can also predict that drug use will rise.