Stress may take two paths in depression.

Psychiatrists have long distinguished between endogenous depressions, thought to reflect mainly biological influences, and reactive, or nonendogenous, depressions, pegged as the aftermath of particularly distressing life experiences. Yet researchers sometimes find that equally stressful events precede both types of depression, creating questions about the usefulness of these diagnoses.

Two new studies, both published in the July Archives of General Psychiatry, suggest a solution to this puzzle. Recent stressful incidents provoke a large majority of all initial bouts with depression, regardless of how those cases are diagnosed. However, depression typically waxes and wanes. For people experiencing its endogenous form, recurrences often appear with minimal environmental nudges, while people with reactive depression slide back into melancholy in response to highly stressful events.

Much previous research has concentrated on individuals who have experienced several periods of depression. This work may thus underestimate the role of stressful events in initiating endogenous depression, conclude George William Brown, a psychologist at the University of London, and his coworkers. Evidence corroborating that of Brown's team comes from a project directed by Ellen Frank, a psychologist at the University of Pittsburgh.

"These findings could mean that there are endogenous depressions that, once started, just keep perking along without significant environmental stress," holds Myrna M. Weissman, director of clinical and genetic epidemiology at the New York State Psychiatric Institute in New York City "There may be a kindling process that occurs in endogenous depression."

Kindling refers to the tendency of some brain areas to react to repeated low-level electrical stimulation by progressively boosting electrical discharges and thus lowering seizure thresholds. Similarly, an initial bout of endogenous depression may somehow lessen substantially the amount of stress needed to produce a recurrence, Weissman suggests.

The British researchers studied 127 women hospitalized for their first episode of depression. Participants received diagnoses of either endogenous depression -- which includes pervasive sadness and hopelessness, loss of interest in all activities, and physical symptoms such as weight loss and sleep problems -- or reactive depression, which features a rock-bottom mood but usually no physical changes. Interviews with each woman probed for stressful events and difficulties over the past 6 months, which interviewers and other team members rated for severity

Volunteers underwent a second interview about 4 years later, when most had suffered a subsequent bout of depression.

Data were combined with those for two earlier groups of depressed patients studied in the same way, as well as for women in a London-area community sample who responded to life-stress interviews.

A total of 60 percent of first-time depressed patients in both endogenous and reactive groups reported a severely stressful experience in the preceding 6 months. But for those who suffered a subsequent depression, 70 percent of the reactive group had encountered severe stress shortly before their episode, compared to one-third of those in the endogenous group and in the community sample.

Frank's group, which administered the same interview to 70 women and 20 men who had suffered numerous bouts of depression, finds a comparable excess of recent severe stress in reactive cases. Highly threatening events appear to lead to recurrences of reactive depression within several weeks, they contend. When people diagnosed with endogenous depression cited such experiences, they usually had occurred several months before.

Some overlap of symptoms occurs between endogenous and reactive depression, they note, which may partly explain why the groups do not differ even more on prior exposure to severe stress. Moreover, interviews probably missed some stressful events that promoted depression.

Brown's group theorizes that recurrences of endogenous depression may be provoked by threatening events that took place between 6 months and 1 year earlier.