Triumphs of Experience: The Men of the Harvard Grant Study

by Vaillant, George E.

  I puzzled over Miller. Psychiatrists were now seeing PTSD in many of the Vietnam vets, and Miller had been in a war too. But I quickly dismissed this notion. He had reported “combat fatigue” once during the war; had he perhaps exaggerated a bit? The Woods traits attributed to him in college had been Cultural, Ideational, and Creative and intuitive, and after all, he was a drama professor. In his interview with Monks, he had said that he had not seen sustained combat, and by his own report many of the Grant men had more severe and more prolonged combat histories than he.

  The years passed. I formulated various hypotheses, not all of them flattering, for Miller’s elusive behavior. In 2009, I had to reread his chart for this book. There I discovered a scrap of paper entered by Clark Heath, who had traveled to Washington in 1950 to review the men’s military medical records. Fifty-five years after World War II ended, I learned what had happened to Art Miller in the army. Here is his record from an Italian field hospital, beginning on June 13, 1944:

  Patient saw 3 or 4 days of combat, remembers killing three Germans. The last he remembers is attacking up hill with men falling, nearby land blasts, then woke up here two days ago. He has no idea as to what happened in the intervening time. On admission, he was acutely disturbed, kept his fists clenched, and threw himself about, calling “Shells! Bombs! I’m afraid!” and no contact could be established.

  There was no change on transfer here. He was restless, disturbed, over-responds to minor stimuli, crawls under covers and into fetal position at sounds of planes and there is no response obtained on questioning.

  After two electro-convulsions, amnesia and agitation entirely cleared. There is some noise sensitivity, battle dreams, and fear of combat remains. He is a mildly shy, sensitive, personality type with a strong sense of duty. . . . From an emotional standpoint he will be of no use whatsoever under combat. . . . I feel that he should be reclassified to limited service, permanent.

  July 3, 1945: Reclassified. Normal now.

  July 5, 1945: Certificate of honorable discharge. Character excellent.

  It had been easy to stand in judgment. I could point to Miller’s noncompliant and passive-aggressive behavior. I could note that he had run away from his family and his country, and that his earned income was as low as any man’s in the Study.

  Or I could—and in 2010 I finally did—understand Art Miller’s whole life as a creative example of posttraumatic growth. Many playwrights (Edward Albee and Eugene O’Neill among them) have endured traumatic childhoods or major depressions, and spun gold from this wretched straw. It’s hard to imagine that Art Miller’s students scorned him as a dropout or a victim; on the contrary, they must have greatly enjoyed the drama that this published scholar brought to their small-town school. But it took forty-five years for me to see that truth. And I saw it not by looking back on what Miller had or hadn’t done, but by discovering the prospective entry that one very caring physician had made years before, when the trauma was still fresh and contemporary.

  A SELF-PORTRAIT

  Since it is I who guided the Study for more than half of its existence, it is fitting that I share some relevant aspects of my own development and the potential biases that grow out of them. I was born in New York City to academic WASP parents and educated at Phillips Exeter Academy and Harvard College, where I majored in history and literature and was an editor of the Lampoon. I went on to Harvard Medical School with the internalized injunction that teaching and service were good and that business and private practice were bad. And perhaps I put a more than necessarily puritanical spin on this dichotomy, because I remember mocking my professors behind their backs for preferring research to clinical care. I have always considered the New York Times to be the source of truth where news is concerned, and politically I usually vote Democratic. Between 1960 and 2009 I lived mostly in Cambridge, Massachusetts. I was happily married during most of those years, despite more than one divorce. As I write this book I am recently remarried once again, living in southern California, and learning (reluctantly) to broaden my political perspectives.

  As a psychiatrist, and like many of my colleagues, I pretend to belong to no “school,” but it won’t have taken long for my readers to realize that I am trained as a psychoanalyst and am a staunch admirer of Adolf Meyer and Erik Erikson. I began psychoanalytic training not because I thought psychoanalysis “cured” people, but because my best clinical teachers were psychoanalysts. All my life I have wanted to be a teacher, and I have always tried to teach my readers how to look for what transpires invisibly beneath their awareness—in their defenses, in their development, and in their hearts. Less obvious, but also important, is the fact that I worked for two years in a Skinnerian laboratory. There I came to believe that the experimental method far surpasses intuition as a means of uncovering truth, and that by charting behavior over time, patterns may be revealed that would interest the staunchest of psychoanalysts. Despite the strictures of my medical student conscience, I have spent more time in “selfish” research pursuits than in “noble” clinical care.

  I should probably note that my interest in Alcoholics Anonymous (see Chapter 9) does not have to do with being in recovery myself. Here again I was following Willie Sutton’s example. To run the Grant and Glueck Studies I needed a Harvard appointment. The only job opening at the time was a co-directorship of the Alcohol Treatment Center at Cambridge Hospital. As a condition of employment at the Treatment Center I went to AA meetings once a month for ten years. I came to scoff but stayed to admire, and as a result of my enthusiasm in 1999 I was appointed a Class A (that is, non-alcoholic) trustee of AA, and the six years that I spent in that capacity were among the best in my life.

  I have always loved big questions. At ten I wrote my sixth-grade term paper on the origin of the universe. I’ve already told how mesmerized I was, two years after my father’s death, when our family received a courtesy copy of his twenty-fifth Harvard reunion book. I marveled at how those raw college caterpillars had evolved into mature forty-seven-year-old men, and was first struck by the profound implications of adult development.

  Until I was eighteen, I intended to become an astrophysicist. By nineteen, however, I had read my roommate’s copy of Robert White’s Lives in Progress, perhaps the first prospectively designed textbook on adult development, and decided to go to medical school instead.33 In my training and in the years that followed it, I was fascinated by people’s capacity to recover from apparent catastrophe and continue to develop over the course of their lives. I became a dedicated believer in the power of long-term prospective study to answer psychiatric riddles, and I read avidly about the world’s great longitudinal studies. They entranced me; they were telescopes, all right, but the kind that could focus on human lives, not just inanimate stars. I was equally entranced by B. F. Skinner’s idea of the cumulative record, a sort of behavioral cousin of the EKG that draws a “picture” of people’s doings as they change over time, allowing us to visualize those changes in a single gestalt. In 1966, while still an assistant professor of psychiatry at Tufts Medical School, I joined the Grant Study and was immediately hooked.

  My dream was to make the Grant Study, at the time little known, as important as the longitudinal studies that had so inspired me. One of the few memories I have of my training psychoanalysis was holding up my brand-new key to the Study files and crowing to my analyst, “I have acquired the key to Fort Knox!” But even I did not anticipate that I would still be working on the Study full-time forty-five years later.

  With financial support from an NIMH Research Scientist Development Award, I worked on the Grant Study for five years before moving my appointment as Associate Professor of Psychiatry from Tufts to Harvard, and succeeding Charles McArthur as director of the Study.

  At the time that I took over, Harvard was still not convinced of either the power or the importance of the Study, and was seriously considering destroying the sensitive files that so enthralled me. (They were especially sensitive just then because the
year before, protesting students had broken the windows of the Health Services and occupied Harvard’s administrative offices.)

  In 1973, I gave a dinner as the new director. I invited various Harvard potentates and Radcliffe president Matina Horner, the sole woman. I described the Study’s plight, and while I was still resisting entreaties by the Harvard authorities to reduce the case records to soulless microfiche, Horner established at Radcliffe the Henry A. Murray Center for the Study of Lives—a center dedicated to preserving longitudinal studies in their original form. One of its first and most exciting acquisitions was the records of the Harvard Study of Adult Development. Over the years (1976–2003) of the Murray Center’s existence, its director, Ann Colby, was an inspiration to all who knew her, especially me.

  The focus of the Study shifted again as my interests pushed it away from sociology and toward epidemiology and psychodynamics. I was especially fascinated by involuntary coping mechanisms and by the relationship between psychological stress and physical symptoms. I am increasingly convinced, especially since passing seventy myself, that psychiatry and psychology need to become more aware of people’s positive emotions and experiences of spirituality. I believe too that a predilection for love and compassion are hard-wired in mammals. One early reviewer of this book said to me, “George, your view of adult development is sooooo 1970s.” But you know what? The seventies weren’t so bad. Those were the years of Bernice Neugarten, Robert Kegan, Jane Loevinger, Emmy Werner, and Jerome Kagan, and I do not believe that their work has so far been surpassed. And anyway, like all the Study directors, like all the men whose life voyages we accompanied, and like the Grant Study itself, I was—I am—a creature of my time. And from age thirty-three to age seventy-eight my participation in the Study has been an abiding joy.

  INTO THE FUTURE

  One of the points I’ll be making throughout this book is that our close relationships when we are young make an enormous difference in the quality of our lives. Yet there have been few studies of intimate relationships near the end of life, let alone of how these relationships affect physical and mental health. In 2003, Robert Waldinger began to study the marriages of the Grant Study men, now in their tenth decades, and that meant inviting their wives to join the Study. Waldinger, associate professor at Harvard Medical School and a psychoanalyst, has since succeeded me as director. His fresh view of the project is now giving the Study a microscopic (as opposed to my telescopic) look at the daily lives of single and married older adults.

  However, the National Institute of Mental Health has not been willing to fund such studies into the twenty-first century. Scientific fashion has once again abandoned the social for the biological. If Willie Sutton wanted grant money to investigate adult lives today, he’d have to do brain studies. Since 2005, Waldinger has obtained funding for neuroimaging studies from three sources: the Harvard Neurodis-covery Center, the Fidelity Foundation, and once again the National Institute on Aging. And for the third time since it stopped funding the Study in 1945, the W. T. Grant Foundation has come to the rescue with additional funds.

  These vicissitudes of funding may be nerve-racking, but they are not an unmitigated curse. Over the last five years, they have encouraged the Study to add biological and neuroscientific data to nearly seventy-five years of information about men’s lives. Most surviving Grant Study men are allowing the Study to collect their DNA, conduct sensitive tests of their intellectual capacity, and carry out neuroimaging of their brain structure and function to create an unprecedented and irreplaceable resource that someday may help us understand the links between brain and behavior in human aging. The combination of life histories and neuroscience data will allow us to investigate fundamental and pressing questions about the aging process. For example, how do genes and environmental factors—such as alcohol use or exposure to traumatic stress—interact to determine who maintains an active and vigorous mind and body into their nineties and who does not? Are there lifestyle choices and behaviors that can buffer the human brain against the ravages of old age?

  As the population grows older, these questions become more urgent for families, for healthcare providers, and for public policymakers. As I’ve noted before, there are advantages as well as disadvantages to studying an elite sample. One of the advantages is that the unusually healthy Grant Study ninety-year-olds offer some insight into what life will look like in thirty years for the average baby boomer. The seventy-five-year-old Study may look dated to modern (or very young) investigators. But it is well to recall Santayana’s warning: “Those who cannot remember the past are condemned to repeat it.”

  4

  HOW CHILDHOOD AND ADOLESCENCE AFFECT OLD AGE

  O joy! that in our embers Is something that doth live, That nature yet remembers What was so fugitive!

  —WILLIAM WORDSWORTH

  ONE ASPECT OF OUR PASTS that we tend to repeat unremembered are the experiences we absorb in childhood about other people and the world they embody. Novelist Joseph Conrad laid the stakes out somberly: “Woe to the man whose heart has not learned while young to hope, to love, to put its trust in life.”1 If the child really is father to the man, then it’s reasonable to ask: how? This is not a simple question, obviously, but the Grant Study has been able to deconstruct it a bit, allowing us to explore the complex of childhood circumstances that shapes our sense of self and our expectations. These in turn shape the relationships that we make, and the social surround that will enrich, or not, our later years. The Study has affirmed the truth of Conrad’s warning, yet it also offers some heartening news in counterbalance, and I’ve listed this in Chapter 2 as the fifth of the great lessons of the Grant Study: What goes right in childhood predicts the future far better than what goes wrong.

  Another mitigating factor, although an elusive one, are certain “sleeper effects” of childhood, which the Study has also illuminated. These may be deep early attachments that have been lost from view through chance, tragedy, or forgetfulness, but are brought to memory’s light again many decades later. Recovery of these lost loves can be profoundly healing, as we will see shortly in another chapter in the life story of Godfrey Camille. But there are less benign sleepers as well—genes for alcoholism, major depression, or Alzheimer’s, for example, that are present from the outset, but do not wreak their havoc until later in life. The sleeper phenomenon points up a corollary to Lesson Five: it is not any one thing for good or ill—social advantage, abusive parents, physical weakness—that determines the way children adapt to life, but the quality of their total experience.

  As we’ve seen already and will see again, long-term perspectives may upset established assumptions about cause and effect, especially when these have been built on the instantaneous observations of cross-sectional research. This is what makes longitudinal studies at once so powerful and so disconcerting. In cross-sectional studies, for example, two of the most powerful correlates of successful aging are income and social class.2 But wealth and status are neither the only nor the best measures by which to judge successful aging. Furthermore, our longitudinal evidence makes clear that when mid- and late-life riches do accompany successful old age, often they are not the result of financial privilege in childhood or even of other attributes popularly considered conducive to worldly success, such as looks or an extroverted personality.3 They are indirect fruits of a childhood experienced or recalled as warm and intimate, for it is such childhoods that give children their best shot at learning to put their trust in life. Many measures of success throughout life are predicted less reliably by early financial and social advantage than by a loved and loving childhood.

  You’ll remember that among the College men, even achievement of military rank correlated more highly with a warm childhood than with social class, athleticism, or intelligence. For the Glueck Inner City men, whose parents were not wealthy, admired fathers, loving mothers, and warm friendships were prominent among the predictors of high income.4 A father on welfare, or even multiple problems in the family
of origin, predicted future income and social class less robustly than a warm family.5 The availability of warm relationships (of which parents are ideally but not necessarily a main source) was the overriding factor.

  Erik Erikson was an early student of the way children acquire Conrad’s learning of the heart. An artist/psychoanalyst, he spent many years working on the first major prospective study of child development, the Berkeley Growth Studies. Erikson believed that an infant’s first task is to develop trust and hope; a toddler’s to develop autonomy; and a five-year-old’s to become comfortable with initiative.6 I will elaborate Erikson’s schema of adult development in Chapter 5; for now, however, I will concentrate on those three early necessities and what they mean for later life.

  ASSESSMENT OF CHILDHOOD ENVIRONMENT

  To start with, here’s some statistical context. In 1970 we set out to investigate which aspects of the College men’s childhoods would predict successful mastery of hope, trust, and the resulting confidence on which autonomy and initiative depend.

  In assessing childhood quality, we observed several rules to minimize bias. First, ratings were based not only upon psychiatric interviews obtained when the College men were at Harvard, but also upon interviews at that time with the men’s parents, especially their mothers.7 It is a limitation of the Grant Study that the men’s childhoods had to be assessed retrospectively, when they were around nineteen. But it is a compensatory strength that the retrospection was very extensive—raters had access not only to the multiple-choice questions and brief essays characteristic of retrospective investigations, but to ten hours of interviews by skilled psychiatrists as well as the parent interviews.