Like a sudden forest fire with no traceable origin, depression often flares up for no apparent reason. Sometimes, though, one can identify a catalyst—the lightning bolt that delivered the spark. On its own no single misfortune can fully explain why and how someone develops depression, and depression sometimes arises and lingers largely irrespective of events or circumstances outside the mind. But some painful experiences—such as the death of a loved one, divorce and abrupt unemployment—can trigger individual episodes of depression, especially the very first incidence.

For a long time psychiatrists and psychologists have lumped such triggers together under rather vague umbrella terms, including "severe psychosocial stressors" and "stressful life events." In recent years, however, a few researchers have looked more carefully at the different kinds of events that provoke a depressive episode. The evidence they have collected so far argues for a more nuanced understanding of how stress interacts with individual susceptibility to depression, how quickly depression follows different types of stress, and how best to treat depression in these various situations.

The current guidebook for psychiatrists, the Diagnostic and Statistical Manual of Mental Disorders IV (DSM-IV), defines a major depressive episode as at least five characteristic symptoms of depression persisting two weeks or longer. Symptoms include low mood and energy, insomnia, feelings of worthlessness, diminished pleasure in daily activities and weight change; to meet the criteria for a diagnosis, the symptoms must interfere with one's work or social life.

The taxonomy of events that precipitate a major depressive episode is vast and diverse. Some people become depressed after learning that they have a serious illness, after a natural disaster destroys their homes or when they fail to achieve important goals. Depression is also prevalent among those who have survived rape and war. The most common trigger of depression is loss, which takes many different forms, including economic misfortune, unexpected unemployment and the loss of cherished possessions. According to large surveys, around 44 percent of depressive episodes are preceded by "interpersonal loss," such as the death of a loved one, divorce, the end of a romantic relationship or the fact that a close friend has moved to another part of the country. In other words, a severed connection with another person probably triggers more depression than any other kind of painful experience.

An event that catalyzes a depressive episode does not have to be catastrophic—sometimes what seems like mild stress or a minor loss to most people is enough to plunge someone into murky misery that refuses to fade. It all depends on an individual's vulnerability to depression, which is determined by a complex interaction of many different factors, including: sources of stress in one's life; family history of mental illness; cognitive style—that is, the patterns of thought unique to an individual; and psychosocial factors, such as adversity in early childhood and the presence or absence of caring relatives and friends. Someone with low vulnerability and no previous depressive episodes may survive a devastating hurricane or emerge from a period of grief following the death of a sibling having never experienced true depression. In contrast, someone at high risk of depression with little social support might fall into the depths of despair for months on end after a budding romance wilts and withers.

Kenneth Kendler of Virginia Commonwealth University and other researchers have argued that people at high risk for depression are "prekindled"—it might not take much to spark their first depressive episode and, from then on, they are increasingly susceptible to spontaneous bouts of depression not triggered by any specific event. A bigger spark—or a greater number of small sparks—is needed to kindle depression in people whose risk is lower, and any relapses are more likely to be linked to a particular loss or stressful experience, rather than flaring up on their own.

An example of how stress interacts with individual susceptibility to depression comes from recent research by George Slavich of the University of California, Los Angeles, and his colleagues. Among 100 people who had been diagnosed with major depressive disorder, those who had experienced greater adversity in their childhoods and who had a longer history of depression were more likely to have had episodes of depression triggered by relatively minor forms of loss. Past experiences had lowered their threshold for depression or, as Kendler might say, prekindled their minds—the embers of depression were still warm. Slavich and his colleagues speculate that people who lose important relationships early on—through the death of a parent, for example—may become especially sensitive to even small losses in the future, especially interpersonal losses.

Slavich has also found that the deliberate rejection of one person by another—a form of interpersonal loss known as "targeted rejection"—is a particularly powerful catalyst of depression. In one study, he and his colleagues interviewed 27 people who had been diagnosed with major depressive disorder. Twelve of 16 participants (75 percent) who had experienced targeted rejection developed depression within 30 days; only three of 11 interviewees (27 percent) who had not been actively rejected became depressed that quickly. Overall, the onset of depression was three times faster following targeted rejection than other forms of loss. The researchers note that rejection by one person often involves subtler exclusion from many others, a phenomenon they term rejection reverberation: if your boss fires you, you will probably lose contact with many of your fellow employees; if your partner unilaterally ends a romantic relationship, you may lose some mutual friends.

New understanding of how quickly interpersonal loss can trigger depression in certain vulnerable people is reflected in a recent change to the DSM. The current version, the DSM-IV, stipulates that someone who has recently lost a loved one should not be diagnosed with a major depressive episode unless his or her depressive symptoms persist longer than two months. If the symptoms have not lasted that long, the reasoning goes, then the person is most likely grieving—a typical and often transient response—rather than suffering from depression. The DSM-5, to be published this May, eliminates this caveat, allowing someone to be diagnosed with depression two weeks after the death of a loved one. It includes some footnotes distinguishing depression from grief, however. Whereas depression is usually constant, grief is more likely to ebb and flow in waves and it does not usually invoke the feelings of worthlessness and low self-esteem that are so characteristic of depression. Grievers long to be reunited with someone they loved; the depressed often believe that they are unlovable.

The DSM revision emerged from new studies confirming that bereavement is one of the most devastating forms of interpersonal loss and sometimes triggers a genuine depressive episode alongside grieving. Studies have also shown that people who are both grieving and depressed benefit from therapy and antidepressants.

In one experiment conducted in California, for example, 22 adults who had lost their spouses in the previous six to eight weeks and subsequently met the DSM-IV criteria for a major depressive episode volunteered to take a daily dose of bupropion, a commonly prescribed antidepressant also known by its brand name, Wellbutrin. Grief and depression responded differently to the treatment, suggesting that the volunteers were indeed experiencing both simultaneously: most people showed lessening symptoms of depression, but not of grief. In fact, some people found that as the fog of depression lifted they were able to grieve properly and confront the fact that their spouse was gone. Sidney Zisook of U.C. San Diego, who conducted the study with his colleagues, cautions that it is a small and uncontrolled trial, which means that the treated volunteers were not directly compared with a similar group of people who did not receive medication. The study was also sponsored by a pharmaceutical company, GlaxoSmithKline, for which Zisook has been a consultant.

When a patient shows symptoms of depression soon after a loss—whether the death of a spouse or a failed romance—clinicians face a dilemma: They must determine whether the patient is heading toward or has already developed true depression or, instead, whether the patient is passing through a phase of typical grief. Weighing factors such as changes in self-esteem and family history of mental illness can help clinicians make an informed evaluation in many cases, but some situations are more ambiguous. Psychiatry has no universal litmus test for depression. Richard Friedman of Weill Cornell Medical College and other psychiatrists have argued that the changes in the DSM-5 will discourage clinicians from carefully considering the difference between grief and depression, thereby encouraging overmedication. In line with DSM-IV, they advocate a period of watchful waiting for at least two months after a patient has lost a loved one before diagnosing depression. "There's nothing to be lost by waiting," Friedman says. "There's probably a tiny fraction of people at high risk for quickly developing depression after bereavement and they will have a known history of depression."

Zisook acknowledges that axing the bereavement exclusion in DSM-5 will help only a small segment of the population. When treating a vulnerable patient with a history of depression, however, a clinician may need to act quickly to prevent bereavement from triggering another major depressive episode. That is what the DSM-5 allows. Zisook also thinks that the DSM-IV confused many clinicians by implying that grief does not last longer than two months. To the contrary, Zisook says, grief can last a lifetime. He further points out that antidepressants are not the only or necessarily the best option for people who are both depressed and grieving. Talk therapy and cognitive behavioral therapy—which involves recognizing and changing maladaptive thought patterns—work as well.

More effective treatments will likely require a much clearer understanding of exactly what happens in the brain and body during depression. As with the majority of mental disorders, depression's underlying biology remains cloudy—but projects such as the National Institute of Mental Health's Research Domain Criteria are making impressive progress. For now, although we may not be able to govern all the genetic and social factors that kindle our minds, we can learn how to contain the sparks life flings at us. "At the end of the day, it's not stressful events alone that result in depression," Slavich says. "It's about the differences in how our brains construe those types of events. All the stress we experience gets translated into the types of biological and cognitive processes that precipitate depression. Some people ruminate about them and others don't. Some people may never develop depression, no matter how badly they are rejected. That's the silver lining—although we can't always control whether someone dies or whether our girlfriend breaks up with us, we can try to control how we think about it and deal with it."