Important health benefits of meditation may be conferred as much by what people actually do and say in their daily lives as by how they self-perceive the effect of training on their behavior and emotions. To test this hypothesis, and to explore whether specific elements of meditation (e.g. compassion vs. mindfulness) have different effects on real-world social behavior, the current project proposes to use the Electronically Activated Recorder (EAR) in a randomized, longitudinal design to objectively measure whether meditation does indeed make people behave in happier, kinder and more prosocial ways in their daily lives. This design will permit an exploratory analysis of how such behavioral transformations might relate to subjective reports of meditative effects. To accomplish these objectives, 108 medically healthy adults will be randomized to 8 weeks training in compassion meditation (n=36), Mindfulness Attention Training (MAT) (n = 36) or to an active control condition consisting of a health education discussion group (n=36). Prior to randomization and again upon completion of these interventions, all subjects will participate in an EAR protocol to evaluate the effect of meditation training on their real-world prosocial and affiliative behavior. Based on prior research and direct pilot data from 25 participants, prosocial and affiliative behavior will be assessed as the amount of (1) shown empathy and affection, (2) laughing vs. sighing, (3) arguing and complaining, (4) time spent interacting with others vs. alone, (5) meaningful conversations, (6) socially inclusive (i.e. use of "we"/us") vs. self-focused (i.e. use of "I"/me") first person language. Following the first EAR assessment, subjects will commence 8 weeks of compassion meditation training, MAT, or attendance in the health discussion group. Following these interventions subjects will participate in EAR monitoring identical to the initial assessment. To assess whether changes in daily behavior are sustained over a longer period, all subjects will undergo a final EAR assessment 6 months following completion of the study interventions. Self-report measures of social integration and support and behavioral/lifestyle variables that might be associated with EAR outcomes will be evaluated prior to each EAR assessment and treated as covariates.

Inclusion Criteria:

• Males and females between the ages of 25 - 55 at study entry
• Ambulatory and in good medical health (see exclusion criteria below for specifics)
• Ability to read/understand English

Exclusion Criteria: Potential subjects will be excluded for any serious ongoing medical or psychiatric condition that might influence results or increase risk of stated participation, including but not limited to:

• Malignancy
• Auto-immune disorder
• Neurologic disorder
• Endocrinopathy; chronic infection (i.e. human immunodeficiency virus, hepatitis B or C)
• Any renal, hepatic or hematological abnormality (other than history of mild anemia)
• Current major depression or major depression requiring hospitalization or resulting in suicide attempt in past year
• Current active suicidal ideation as assessed by the Structured Clinical Interview for DSM-IV (SCID): current DSM-IV substance abuse
• Schizophrenia or bipolar disorder type 1
• BMI ≥ 30
• Any other current/past condition that might increase the risk of participation.