Question: What kind of fallacy would this be?

I think the mother is seeing correlation and claiming causation. People drink energy drinks; some people who drink energy drinks die. The mother doesn't state any facts supporting a link between drinking energy drinks and dying. There are many other possible explanations.

I don't know what fallacy the son is making, other than just using the definition of "spouse." By definition, if you don't have a spouse, the chances of your being killed by a spouse are 0.

At this point, the mother might clarify: "OK, smarty-pants, the risks of dying are greater if you drink an energy drink than if you don't." (This is still the same correlation/causation fallacy, unless there is data to support a causation.) "How about we expand your example to include people in relationships as well as married people. That eliminates any definition issues. Is your assertion still true; are people more likely to be killed by someone in a relationship with them than not?"

Hopefully both parties are learned enough in logic to make these extensions of their arguments.