It is well known that there is an obesity epidemic in the United States. Ethnic variations in the epidemic indicate that African Americans and Latinos are at higher risk than the general population. It is also established that people with severe and persistent mental illness die 25 years earlier than the general population and that this early mortality seems tied to preventable conditions, including obesity. Considering the range of medical complications that result from obesity, we were compelled to provide a treatment program to our clients, who are urban Latinos with severe mental illness and low incomes.

Behavioral weight control techniques are effective for some overweight and obese people with severe mental illness. However, there are limited data on generalizability of these results to real-world settings and almost no data on the effectiveness of such approaches in higher-risk populations, such as low-income Latinos with severe mental illness.

The Washington Heights Community Service (WHCS) provides psychiatric services for people with severe mental illness living in Northern Manhattan. Approximately 75% of the WHCS population is Hispanic, which creates a unique opportunity to assess the efficacy of a behavioral treatment in this high-risk group.

The HEALTH Project (Healthy Eating and Activity Among Latinos Treated in the Heights) was started in March 2008. This program was based on a course developed by Rohan Ganguli, M.D., at the University of Pittsburgh, and we adapted it to meet the needs of our population. We reviewed the literature and conducted interviews with patients, clinicians, schizophrenia specialists, and cultural experts to develop our adaptations, which are described below. Adaptations for urban and low-income population characteristics were in three areas. First, the original program referred to shopping in malls, traveling by car, and using dishwashers—unlikely activities for our patients, who live in Manhattan—so instead, we refer to urban activities, such as taking the subway.

Second, many of our patients are on low incomes and obtain food via food banks, soup kitchens, and food stamps. Given this reality, we recommend low-cost options whenever recommending specific foods. We also highlight the cost per person of the healthy snacks we provide in class. Grocery shopping in neighborhoods with low socioeconomic status has been associated with higher body mass index, possibly because of lower quality of available foods in disadvantaged neighborhoods. Thus we visited bodegas and supermarkets in the neighborhood and found that a wide range of healthy foods was readily available, including whole wheat bread, skim milk, sugar-free frozen desserts, lowfat cheese, sugar-free beverages, and good-quality fresh produce. We discussed these options with participants and distributed information on local farmers' markets that accepted food stamps.

Home exercise videos were successful in Dr. Ganguli's original program, but some of our participants do not have VCR or DVD equipment. Our third adaptation, then, was to make exercise videos available for use in the clinic, either to patients working out together or on their own. Other types of low-cost exercise are encouraged, especially walking outdoors. We also provide pedometers and scales because they are too expensive for participants to purchase themselves.

The other adaptations for our Latino population were cultural. Because many of the patients speak Spanish, a bilingual staff member runs the lessons and discussions. Handouts and workbooks are available in both English and Spanish.

Because our participants are mostly Dominican-American, we incorporate cultural influences on food, diet, and exercise that might be relevant for them. For example, lessons on food choices include items such as plantains and chicken stews, which are common in Dominican cuisine. We encourage exercise that has cultural ties (Merengue dancing, for example). We also address the cultural belief that being somewhat overweight is viewed as desirable.

The centrality of family in Hispanic cultures is well documented. We invite family members to a special class to encourage their involvement in the program. This provides an opportunity for the group leader to explain the intervention, answer questions, and solicit support for participants.

Specific ethnocultural group composition helps participants support each other through the course.

We have enrolled 50 patients into this program so far. These patients enjoy participating, and some have lost weight. It is our ultimate goal to assess the feasibility and efficacy of this modified behavioral course.