Alexandra Molloy
Like fine wine, everything improves with age. Well, maybe not everything but when it comes to longitudinal cohort studies in epidemiology, the old adage rings true. These kinds of studies--in which a population is examined over time to determine the presence or absence of a particular health outcome--are unparalleled in their ability to ferret out specific information about the distribution of diseases and the risk factors associated with their onset and progression. The greater the time period, the more information they impart.
Case in point is the Stirling County Study (scs), which for more than 50 years has provided insight into the realm of mental illness, with a particular focus on the more common disorders of depression and anxiety. Started in 1948 by Dr. Alexander Leighton, professor emeritus of social psychiatry at the Harvard School of Public Health, scs may well be the longest running epidemiologic study to date, meticulously surveying the inhabitants of a rural region in Atlantic Canada about their personal mental health status over three distinct phases. With a host of recently published results, the study is still going strong, much like Leighton himself, who at 92 remains an active consultant to the project he conceived all those years ago. "So it takes a long life to do a long, long study," laughs Dr. Jane Murphy, who began working on SCS as a graduate student when it was launched and took over its direction after Leighton's retirement in 1975. "And two generations help, in a way. I'm the study's second generation, but I've also been a part of it for so long that we both really look upon it as our life's work."
Not only has SCS become a lifetime investment for its two principal researchers but also an investment of a lifetime for the field of mental health. Few studies can hold a candle to its longevity or its impact on the understanding of mental illness in the general "household" population as opposed to those seeking treatment. "One of the main reasons for my starting a psychiatric epidemiology study was the sense of needing to know more about well people so that we could improve the precision with which we distinguish them from the diagnostically ill," says Leighton. "In view of the dearth of biological markers for psychiatric disorders, the need is compelling." To this end, Leighton designed SCS to survey a socioeconomically and culturally diverse cohort using a knock-on-door approach to eliminate the selection bias seen in studies that examine mental disorders in the clinical setting. After an initial period of prep work, interviewers--mostly trained graduate students--first appeared on the doorsteps of Stirling County in 1952, faced with the daunting task of getting the residents to open up their homes and their "heads" in the name of science. To their relief, they found the local citizenry willing subjects; more than a thousand people took part in the initial phase.
Over time, thousands more in Stirling County would follow suit. In 1970, the original cohort was reinterviewed and a new sample added to the study, both of which would be tracked down again in 1992 as yet a third new group was tacked on. This unique structure has allowed SCS researchers to look at depression and anxiety over time as well as cross-sectionally as each new sample was added. "That's an unusual design for a longitudinal study," notes Murphy, professor of epidemiology at the School with affiliations at Harvard Medical School and Massachusetts General Hospital, "but it's one I have fought diligently to keep. It just makes more sense if you add a new representative sample, because everyone you follow ages and then they're no longer representative. If you're interested in both individuals and populations, you really need to combine the two."
The very things that make the study exceptional, however, have posed special
challenges for its investigators. For one, time can be an adversary as well
as a friend. With the years have come changes in scientific methodology and
difficulties in finding previous study subjects. Even the study itself has changed
home base three times, from Cornell University, to the Harvard School of Public
Health, to its current abode at Massachusetts General Hospital. "Having
a study team stay together for more than 50 years and be able to get funding
for 50 years, it's unheard of," says Richard Monson, professor of epidemiology
at the School, who has been with the project for 15 years himself. But even
these challenges can provide perspective. Notes Nan Laird, professor of biostatistics
and relative newcomer to the SCS team, "It's so interesting to see how
the study has evolved. It's really helped me understand how methods of gathering
data in psychiatric illness have developed--because Alec and Jane were pioneers
in these methods--and how they have changed over time. It becomes clear why
it's so complicated to do the kind of work that they do."
Not surprisingly, things have also changed in Stirling County since back in
'52. Nestled in the craggy Canadian coastline, this rural region of about 20,000
residents has borne witness to the social trends that typify the North American
experience more generally. Its standard of living has risen, its health care
delivery improved, and its primary industries declined. Reliance on family and
religion is down, and crime and drug use is up. "Everything we can think
of across North America has happened there," notes Murphy. "I don't
think we could have ever foreseen that, but it does make it an all the more
valuable laboratory."
One thing that hasn't changed in this "laboratory," however, is its overall rate of depression. Bucking the popular belief that the prevalence of the disorder is on the rise, SCS researchers have found that depression in Stirling County from 1952 to 1992 remained relatively stable at about 5 percent. What may explain this disconnect between perception and reality is the changing demographics of the illness; while the rate of depression has remained constant over time, how it's been distributed among the population has not. From 1970 to 1992, a two-fold increase occurred in the depression rate among women under age 45, offset by concomitant declines among older women and men of all ages to keep the overall rate the same. What makes young women born after World War II at greater risk for depressive illness remains an open question. One possible explanation is that employment may be a double-edged sword for young women; because women joined the work force as never before during this time period, a connection may exist between the two trends.
But women aren't the only segment of the population faced with the consequences of depression. One significant outcome of SCS was to document the catastrophic effects depression has on men. Following residents diagnosed as depressed for 16 years, the study found that men had twice the expected mortality rate, and, by 1970, 83 percent of the depressed men identified in 1952 had died or remained chronically or recurrently depressed with persistent impairment. Their depressed female counterparts, in contrast, fared far better, with 43 percent dead or chronically/recurrently depressed over the same period. The researchers suspect that willingness to seek treatment might be the reason for the difference. "A study like ours doesn't prove that treatment was responsible for the better outcome, but the figures are pretty compelling," notes Murphy. "And people not getting treatment, people having poor outcomes--that makes for a significant public health problem, doesn't it?"
That Murphy and Leighton often speak of their study in the context of public health seems only natural. Not only is SCS a longitudinal cohort study, one of the mainstays of public health research, but its findings lend support to the public health approach embodied by the well-known HSPH/WHO Global Burden of Disease (GBD) report, which created a revolution in thinking about disease burden by including disability as well as death in the overall picture. "Taken together with the fact that psychiatric disorders have been found to be far more common than most people believed when psychiatric epidemiology got started," says Leighton, "many of the disorders greatly reduce the quality of life, some of them cut life short, and a major portion go untreated. In short, psychiatric disorders are a major public health concern." GBD predicts that by 2020 depression will be second only to ischemic heart disease as the leading contributor to global disease burden. "The Global Burden of Disease found that if you took into account disability, psychiatric disorders take up much more burden than they ever did before," says Murphy, "which is not due to an increase in the incidence of depression; it's more due to the control of diseases that have a very high impact on mortality: the diarrheal diseases, the respiratory diseases, and an increase in the population of age groups that are at risk for depression. So it isn't just a raw increase; it's like so many things in public health--a juggling act of many factors."
It's a juggling act with which Murphy is all too familiar in her own research, and she credits the success of SCS to the joint efforts of the many people who have helped her keep the balls in the air. "It is a humbling experience to have this much cooperation over the years," she reflects. Each individual has had a unique role to play, from the members of the current SCS team itself--Murphy, Leighton, the epidemiologist Monson, the biostatistician Laird, and programmer analyst Arthur Sobol, a.k.a. "the SAS guru"--right on down to the multitude of graduate students knocking on Stirling County's doors or applying the data to a range of research analyses on subjects like aging and Alzheimer's. But it is a group mentality that pervades the project, a sense of camaraderie that can only come with time and experience and which makes the SCS team meeting each month, as Monson puts it, "an unqualified pleasure." Muses Laird, "Often when you're a biostatistician, you get called in when people think, 'Well, this is a statistical issue,' whereas in the Stirling County Study, you're a part of the whole process, and I think that makes a big difference." Perhaps the biggest difference comes from the cooperative efforts of the Stirling County residents themselves, who have consistently and unabashedly provided the study with insights into the workings of their minds. But in case you were planning to pay a visit to this amicable Canadian region, be forewarned that you won't find "Stirling County" on any map. Because of the sensitivity of the subject matter, Leighton long ago devised this pseudonym for the area to protect the privacy of its residents. But while the people of Stirling County will remain forever anonymous, their contributions to understanding mental illness are as good as gold.
Harvard Public Health Review Winter 2002/text version
This page is maintained by Development Communications in the Office of Resource
Development.
Updated January 2005
To contact us with suggestions, comments, and questions, please e-mail: editor@hsph.harvard.edu
Copyright, 2005, President and Fellows of Harvard College