When people are asked to write about emotional upheavals for as little as fifteen minutes a day for three days, their physical and mental health often improves. Not a particularly shocking finding. But in 1986, when the first expressive writing study was published, the idea caused quite a stir.

The discovery of expressive writing was a mix of chance, dead ends, and the fitting together of a glorious puzzle. After graduate school, most of my research focused on the psychology of physical symptoms. When, how, and why did people notice symptoms and sensations? Although most studies were lab experiments, I was interested in the ways people reported and acted on their symptoms in the real world. Around this time, a student told me about her roommate, who was secretly gorging and purging large amounts of food almost every day. Later, other undergraduates admitted they were doing the same thing. There was virtually nothing in the research literature about bulimia at the time. Why not pass out a questionnaire to several hundred students to get a sense of its incidence and correlates?

Working with a small group of undergraduates, we devised a lengthy questionnaire that asked about eating behaviors, food preferences, and family eating practices. We threw in some random questions, asking about things such as how they got along with their parents, their health problems and behaviors, and “Prior to the age of 17, did you ever have a traumatic sexual experience?” Why did we include the sexual trauma question? No reason. It just sounded interesting.

The questionnaire went out to 800 undergraduates. Although we didn't find much about eating disorders, one powerful and unexpected finding emerged: People who claimed to have experienced a sexual trauma were more likely to have every health problem we asked about. The same pattern emerged on a large survey I conducted with the magazine Psychology Today a few months later. Both women and men who said they had experienced a sexual trauma were twice as likely to have been hospitalized in the previous year for any cause. They also reported that they were more likely to have been diagnosed with cancer, high blood pressure, ulcers, colds, flus, and almost every other disease we asked about.