"Let's see what "multifactorial" means in a practical sense. Consider someone with frequent depression who is visiting a friend today, pouring her heart out about her problems. How much could you have predicted the global depression and today's behavior by knowing about her biology? Suppose "knowing about her biology" consisted only of knowing what version of the serotonin transporter gene she has. How much predictive power does that give you? As we saw in chapter 8, not much--say, 10 percent. What if "knowing about her biology" consists of knowing the status of that gene plus knowing if one of her parents died when she was a child? More, maybe 25 percent. How knowing her serotonin transporter gene status + childhood adversity status + whether she is living alone in poverty? Maybe up to 40 percent. Add knowledge of the average level of glucocorticoids in her bloodstream today. Maybe a bit more. Toss in knowing if she's living in an individualist or a collectivist culture. Some more predictability.fn11 Know if she is menstruating (which typically exacerbates symptoms in seriously depressed women, making it more likely that they'll be socially withdrawn rather than reaching out to someone). Some more predictability. Maybe even above the 50 percent mark by now. Add enough factors, many of which, possibly most of which, have not yet been discovered, and eventually your multifactorial biological knowledge will give you the same predictive power as in the fractured-bone scenario. Not different amounts of biological causation; different types of causation."