|
||||||||
|
EPIDEMIOLOGY Taking the Long View of DepressionFifty-year Study Reveals Rise in the Illness Among Younger Women You would not find Stirling County on a map of Atlantic Canada, but it is a real place. It was given this protective pseudonym by Alexander Leighton, HSPH professor emeritus of social psychiatry, back in 1948 when he first picked the location for an unprecedented longitudinal study of mental illness. Since the first interviewers took the field in 1952, Stirling County has mirrored nearly all the social changes that have transformed daily life in metropolitan centers across North America. Stirling County has become more suburban and less rural, the local economy less industrial and, for that matter, less local. It has seen living standards rise, educational opportunities widen, and health care delivery expand. It has also seen more crime, more drug abuse, more media saturation, frailer families, and weaker religious values. What Stirling County has not seen is a general increase of depression.
The longitudinal study on mental illness begun by Jane Murphy and Alexander Leighton 50 years ago is showing some surprising results today. Photo by Liza Green, HMS Media Services The rate of depression in the general population remained virtually constant at about 5 percent from 1952 to 1992, according to a flurry of new papers from the Stirling County Study published in recent months by Jane Murphy, HMS professor of psychiatry at Massachusetts General Hospital; Nan Laird, HSPH professor of biostatistics; Richard Monson, HSPH professor of epidemiology; Art Sobol, HSPH programmer analyst; and Leighton. As reported in the March Archives of General Psychiatry and the May Psychological Medicine, what changed over 40 years is the demographics of depression. Compared to 1952, younger women are now at twice the risk for the disorder. These new results from the study, based on the third harvest of data ending in 1996, fly in the face of other epidemiological studies of community mental health that reported significant increases based on one-time surveys asking subjects to recall their lifetime experiences with depression. The Stirling County Study findings also contradict the conventional wisdom that modern life is so overwhelming that we live in an age of depression, according to Murphy, who has been the project director since 1975. "What we're able to report is that the overall prevalence rate is steady at 5 percent, but the way it's distributed in the population is different between 1970 and 1992, and it's largely influenced by this twofold increase among women under 45." Murphy continued, "Now when I'm asked how our findings fit with what other people are talking about in terms of an increase in depression, well, it's difficult to explain, but it both refutes and corroborates. They refute it in terms of saying it doesn't look as if there's an overall increase. But the increase among young women under 45 fits very well with the other studies because they have suggested that something began to happen among people born after World War II." The overall rate remained constant because the sharp increase in depression among younger women was offset by smaller reductions in the rates among younger men, older men, and older women. Banking DataLongitudinal studies create large cohorts of subjects who can be followed over time and re-interviewed. Thus the original Stirling County sample of 1952 was tracked down as a cohort at the time a new sample was identified in 1970. Both of these cohorts were re-interviewed as a third sample was added in 1992.The result is a richly detailed data trove that can be re-analyzed in new ways to get at novel questions. This round of papers, Murphy explained, scrutinized the depression rate in two separate ways—by prevalence and by incidence. The prevalence study was based on the three samples, but the incidence study was based on following up the first two samples. Tracking down these older cohorts is time-consuming. But for an incidence study, the focus is on people who are well at the beginning of the period of interest. "Incidence is theoretically a more important rate, but there are fewer opportunities to calculate it because it takes all this tracking over time," Murphy said. The annual incidence rate for depression (new cases) was about four per 1,000 for both cohorts, which is quite consistent with the overall prevalence rate of roughly 5 percent, given that many depressions are chronic, said Murphy. There was, however, a contrast between the incidence and prevalence trends. There was no change over time in the distribution of incidence by age and gender as there was for prevalence. The incidence study was based solely on the two earlier cohorts in which none of the women in 1952 and only a few in 1970 were born after 1945. "This difference seems to highlight an important feature of our findings," Murphy said. "Only the younger women in the new sample of 1992 had been born so recently, and they are the ones whose rate markedly increased. "This suggests that a portion of the prevalence rate is under social influences," said Murphy. "We have tried to emphasize that depression is a very complex disorder. There's no doubt from evidence other than in our study that there is a genetic component, but it's by no means a genetic disorder. Depression has never shown anything like a genetic heavy loading. So some portion of it would relate to experience, and with this suggestion that women born after the Second World War are at increased risk, it makes you think that maybe the proportion of depressions that are socially related is on the rise." Depression's EffectDepression is a debilitating, persistent, and deadly disease. It is especially dangerous for men, said Murphy, and the Stirling County data bear that out. Excluding alcohol abusers, the men identified as depressed in 1952 had twice the expected mortality rate over the next 16 years. By 1970, 83 percent of them were either dead or had been chronically or recurrently depressed with persistent impairment. Depressed women fared better, perhaps because they were twice as likely as men to seek help."Our next task is to look at everything we have that might provide some clues about why younger women still in their childbearing years would have this high rate. We have economic information we can look at, marital history, childbearing history. We have information that goes right back to the first 1952 study on attitudes—whether a woman should work outside the home, for example. The most immediate thing that's different from 1952 to 1970 to 1992 is the number of women who are contributing to the household income by regular well-recognized work. You would think that would be a blessing in some regards, but it may be a disadvantage in others. Whether we're going to be able to tease any of that out, I don't know, but we're extremely anxious to see what the results will be." In his commentary on Stirling County, Psychological Medicine editor E.S. Paykel hailed the study as "one of the classics of epidemiology." He also noted that "these studies are very difficult to carry out, requiring remarkable persistence and devotion." Leighton and Murphy, who are husband and wife, have shown both. They brought the Stirling County Study with them from Cornell when they came to Harvard in 1966. Since 1982, the study has been based at MGH. At 92, Leighton still remains an active consultant on the project. It was Murphy who actually made the first "cold call" in 1952, driving up to a randomly selected farmhouse in Stirling County and knocking on the door. Her task was to convince a total stranger that she was for real, that confidentiality would be ironclad, and that answering an hour and 45 minutes of questions would advance our knowledge of mental health. That's how the Stirling County Study has proceeded ever since. —John Fleischman |
