So speaking wrt primary prevention only, this is my approach as adapted from the various Endocrine guidelines (AACE, ADA are the big ones):
Triglycerides under 500 should be treated with lifestyle modification. In and of themselves, a mildly elevated elevated triglyceride level (150-499) doesn't seem to give much of a risk for heart disease/pancreatitis. If they have other lipid abnormalities and are at high risk for heart disease, some people treat, but certainly not in the relatively young population. It's not in the guidelines, but in this population I'd probably consider adding some fish oil. I'd also doubly make sure that I screened them for DM, hypothyroidism, and alcohol abuse, which are the most common causes of a non-genetic hypertriglyceridemia.
Triglycerides >500, would consider addition of fibrates. Here the primary risk you're worried about about is the pancreatitis. I'd also doubly make sure I counsel my patients to avoid anything close to problem drinking, because that risk is more than additive with severe hypertriglyceridemia.
For patients <40 with an LDL >190, I might consider treatment. The atypical antipsychotics might muddy the waters, but even still the most concerning part would be a genetic dyslipidemia. I'd personally likely send these patients to your neighborhood Endocrinologist.
Other issues? It really truly depends on risk factors.
For patients <40 with diabetes, if they have no other risk factors (smoking, htn, family hx of early cad, etc), I would typically do nothing. If they do have other risk factors, the ADA recommendations say put them on a statin. AACE starts getting into treating LDLs to target. Really, whoever is managing their diabetes should take care of this, so I think as the psychiatrist you're mostly off the hook.
For patients <40 without diabetes and abnormal lipids but an LDL <190 you're going to run into no good recommendations. The best I can say would be to stick them into a risk calculator that they aren't filtered out of, which out of all of them (Framingham, MESA, UKPDS, Reynolds, ACC/AHA), you're probably stuck with just Framingham (the rest require you to be older, have DM, have a hsCRP or a coronary calcium score, etc).
Framingham Coronary Heart Disease Risk Score - MDCalc.
If Framingham calculator puts their 10-year risk >20%, I would consider treatment. If it's 11-19% and they have risk factors (smoking, HTN, early CAD in the family, etc), I *might* consider treatment after carefully talking about it with them. Framingham is less accurate for women and statins are teratogenic anyway, so I'd be very hesitant to put a young woman on one just based on this alone. If they're <40 and their 10-year risk is <10%, I'd probably do nothing but counsel them about lifestyle.