Introduction

Top Abstract Introduction Methods Results Discussion References

The link between socioeconomic status and survival is enigmatic. Many studies have shown that high occupational achievement is related to better health, yet the underlying mechanisms remain disputed.^1-3 Writers on this topic can use behavioural or biological theories, or both, to explain the findings.⁴ Behavioural theories assume that lifestyles of people in the upper social classes contribute to good health. Indeed, smoking is a notable cause of sickness and is somewhat more common among poor people than rich people.⁵ Biological theories assume that internal processes related to success reduce susceptibility to disease. ^{6 7} For example, stress that prevails in states of relative deprivation may cause the immune system to malfunction.⁸

Academy awardsnominees and winners Three wins Billy Wilder: The Lost Weekend (1945), Sunset Boulevard (1950), The Apartment (1960) Charles Brackett: The Lost Weekend (1945), Sunset Boulevard (1950), Titanic (1953) Five or more nominations At least one win Woody Allen: Hannah and Her Sisters (1986), Annie Hall (1977) Billy Wilder: (see above) John Huston: The Treasure of the Sierra Madre (1948) Charles Brackett: (see above) Ben Hecht: The Underworld (1928), The Scoundrel (1935) Carl Foreman: The Bridge on the River Kwai (1957) Oliver Stone: Midnight Express (1978) Robert Benton: Kramer vs Kramer (1979), Places in the Heart (1984) Francis Ford Coppola: Patton (1970), The Godfather (1972), The Godfather II (1974) Michael Wilson: A Place in the Sun (1951), The Bridge on the River Kwai (1957) Joseph L Mankiewicz: A Letter to Three Wives (1949), All About Eve (1950) Richard Brooks: Elmer Gantry (1960) Robert Riskin: It Happened One Night (1934) No wins Federico Fellini Ingmar Bergman Stanley Kubrick

The link between occupational achievement and survival is further complicated by common misunderstandings. Firstly, the association is not due solely to a reverse causality artefact. Aside from a few diseasesfor example, schizophreniamost studies indicate that good health results from, rather than leads to, high achievement. ^{9 10} Secondly, the association is not limited to the poor only. Instead, several studies suggest that a link exists at average levels of achievement.¹¹ Thirdly, the association is unlikely to have a single explanation.¹² Furthermore, many determinants of survivalfor example, command over resources, perceptions of social hierarchy, inequities in the material world are compatible with both behavioural and biological theories.¹³

We wondered whether the lives of screenwriters might provide some insight into the relative contributions that behavioural and biological mechanisms make to the link between achievement and survival. We selected screenwriters because they labour in anonymity, yet their work is renowned. For example, millions of people know the movie Lawrence of Arabia, yet few would recognise the authors of the screenplay (Robert Bolt and Michael Wilson). Unlike other occupations, screenwriters can obtain high stature without a special upbringing or daily accountability. Without such guides for behaviour, successful screenwriters might gain little survival advantage according to a behavioural theory, yet they might gain great survival advantage according to a biological theory. Our study tests the long term survival of highly successful screenwriters.

	Methods

Top Abstract Introduction Methods Results Discussion References

Academy of Motion Picture Arts and SciencesThe Academy of Motion Picture Arts and Sciences currently has about 6000 members and is grouped into 13 branchesfor example, the writers branch has about 300 members. The annual awards process is complex (see www.oscar.org for details). In December of each year the academy compiles a list of eligible films; each writer of these films is eligible to be nominated for a screenwriting award. In the following January the list is sent to academy members, and members of the writers branch are invited to nominate five films in each of two screenplay categories. In February the nominations are tabulated, the top five in each category are identified, and all academy members subsequently vote for each category. The academy award goes to the writers of the screenplay with the most votes.

Selection of writersWe identified all screenwriters nominated for an academy award for writing at any time during their career. Specific categories for original screenplay and for adapted screenplay were included; distinctions have been given different names over the years. The Academy of Motion Picture Arts and Sciences supplied a listing of all screenwriters nominated for writing. The period of selection spanned the time from inception to the most recent awards (1929-2001), amounting to 73 consecutive annual award ceremonies. Many films had more than one author, and we included every person who received writing credits. Some winners had multiple wins during their career; we recorded each one, along with the total number of films and total number of nominations of every screenwriter.

Births and deathsWe collected data on each person's date of birth, and date of death if applicable, from the internet through four databases: the AMG all movie guide (www.allmovie.com), the internet movie database (www.imdb.com), the Los Angeles public library obituary index (http://dbase1.lapl.org/pages/rip.htm), and the social security death index (www.ancestry.com). Each source aims to provide up to date information and undergoes public scrutiny. Data were checked by consulting 14 written publications, and conflicts were resolved by accepting printed sources over the internet.^14-27 We obtained additional data on births and deaths by inquiry to the national film information service and by contacting agents representing the screenwriters.

Individual characteristicsPersonal details about individual screenwriters were retrieved by using similar methods to the above, with the following exceptions. We determined whether the screenwriter had received any formal education beyond high school by consulting the single most comprehensive source, the AMG all movie guide. The start and end dates of a career are often ambiguous, so we accepted the first and final years as listed in the guide's filmography of movie contributions for each person. Screenwriters write on varied topics, so each writer's genre was classified according to what was listed first by the AMG all movie guide. Similarly, film reviews are not necessarily an accurate measure of a film's quality, yet the all movie guide's five star ratings were accepted.

Statistical analysisOur primary analysis compared the mortality of screenwriters who won an award with the mortality of screenwriters who were nominated but did not win. We plotted survival by using the Kaplan-Meier method and estimated life expectancy as the area under the curve. We based statistical significance on the log rank test. These methods are identical to methods used previously.²⁸ Multivariate analyses used the proportional hazards model to adjust for year of birth, sex, education (documented or not), film genre (drama or not), total films, total four star films, total nominations, age at first film, and age at first nomination. We used time dependent covariate analyses to evaluate writers who achieved victory years after their first nomination. ^{29 30} All tests were conducted using two tailed analyses.

	Results

Top Abstract Introduction Methods Results Discussion References

Overall, 850 screenwriters were nominated for an academy award; of these, 189 won at least once (see www.oscar.org) and 661 did not win. Winners and nominees had similar demographic characteristics (table 1). The median age at first film and first nomination was 32 and 41 years, respectively. For the winners, the median age at first victory was 41 years; most (171/185, 92%) achieved victory before 55 years of age and a few (24/189, 13%) had multiple victories. The most frequently nominated screenwriter was Woody Allen, who accumulated 13 nominations and two wins. Most wrote as teams, including Billy Wilder, who had 12 nominations and three wins. He won two of these awards with Charles Brackett. Most writers did not act or direct. The film that won an academy award for best picture was usually nominated for a screenwriting award also (66/73).

Billy Wilder (1906-) (left) and Charles Brackett (1892-1969) (right) both won three academy awards for screenwriting. They are pictured here with Gloria Swanson, who starred in Sunset Boulevard, for which Wilder and Brackett won an academy award in 1950

We assessed the length of each screenwriter's career as the interval from their first to their latest film. The degree of occupational achievement varied substantially (table 1). On average, winners had careers that lasted 14% longer than careers of nominees (27.7 v 24.2 years, P=0.004), they worked on 34% more total films than nominees (23.2 v 17.3, P<0.001), and wrote 58% more four star films (4.8 v 3.1, P<0.001). Education was documented for 58 winners and 143 nominees; of these, most had attended some academic courses beyond high school (55/58 v 116/143, 95% v 81%, P=0.013). Drama was the most common writing genre, accounting for the majority of screenwriters and having a slight tendency to be more frequent in winners than nominees (131/189 v 408/661, 69% v 62%, P=0.056).

A total of 428 screenwriters had died by 6 April 2001, reflecting a median follow up from birth of 68 years. Winners had shorter lives than nominees (figure next page) and the difference in life expectancy was 3.6 years (74.1 v 77.7 years, P=0.004). Analyses based only on men showed a 4.0 year difference in life expectancy between winners and nominees (73.6 v 77.6 years, P=0.002). The overall difference persisted after screenwriters who died before age 55 or were first nominated after age 55 were excluded (76.2 v 79.8 years, P=0.003). Analyses based on survival after first film, rather than after birth, yielded a difference of 2.4 years (37.1 v 39.5 years, P=0.008). Analyses based on survival after first nomination, rather than after birth, yielded a difference of 3.2 years (45.4 v 48.6 years, P=0.002).

The generally higher mortality for winners compared with nominees was equivalent to about a 37% relative increase in death rates (95% confidence interval 10% to 70%). Adjusting for demographic factors yielded similar results (table 2). Analyses using a time dependent step function, in which winners were counted as nominees until the time of their first victory, yielded a relative increase in death rates of 48% (19% to 84%). Analyses that confined the group of winners to screenwriters who won at first nomination, and classified all others as nominees, yielded a relative increase in death rates of 40% (10% to 79%). Analyses that excluded all screenwriters with multiple victories yielded a relative increase in death rates of 39% (11% to 75%).

Screenwriters with long careers often received more nominations than did screenwriters with short careers. The winners accumulated a total of 396 nominations, of which 181 were defeats. The nominees accumulated a total of 827 nominations. The number of nominations in each group was equivalent to about 0.14 per year of career. We found no association between number of defeats and increased death rates, either for winners (3, 13 to 9) or nominees (8, 21 to 8). Analyses of both groups together showed that each victory was associated with a 22% (3% to 44%) increase in death rates, whereas each nomination not followed by a win offered no significant increase in death rates (7%, 15% to 2%).

We focused further analyses on the winners to better understand their shorter survival. Winners who worked intensely, intensity being measured as films per year of career, had a 67% (19% to 134%) higher death rate than winners who worked less intensely, measured as the additional hazard per film per year. For example, screenwriters who averaged less than one film each year in their career lived about 4.5 years longer than screenwriters who averaged one or more films each year in their career (75.7 v 71.2 years, P=0.035). The increased risk associated with work intensity was not unique to winners but was also observed when nominees averaging less than one film per year were compared with nominees averaging one or more films per year (79.0 v 73.7, P=0.003).

Other analyses assessed how missing data might affect the robustness of the results. Overall, the proportion of screenwriters not known to be dead was smaller for the winners than for the nominees (77/189 v 345/661, 41% v 52%, P=0.006). However, analyses based on screenwriters known to be dead showed a 20% (5% to 47%) increase in death rates among the winners. In addition, 55 screenwriters were missing valid birth dates; more dates were missing for nominees than winners (table 1). For life expectancy to be equal for winners and nominees, however, all these missing screenwriters would need to have died at an average age of 15. 

View larger version (14K): [in this window] [in a new window]   Survival of winners and nominees of academy awards for screenwriting. The graph shows the percentage of each group alive, plotted by using the Kaplan-Meier technique. Primary statistical analysis is based on a log rank test comparing winners to nominees (n=185, deaths=112 and n=610, deaths=316, respectively)

	Discussion

Top Abstract Introduction Methods Results Discussion References

Winning an academy award for screenwriting is associated with a loss in life expectancy. The apparent decrease in survival was about three years, could not be explained by simple demographics, and was evident even though victory leads to a major gain in earnings.³¹ A career with many nominations but no awards was not associated with decreased survival, even though such a career indicates a high level of skill. The increased mortality is not easily attributed to occupation, talent, social hierarchy, neomaterial conditions, reverse causality, or measurement error. Rather, the results highlight the importance of occupational activity on susceptibility to disease and indicates that higher status does not always confer greater longevity.

ExplanationsSeveral explanations might account for why the link between success and survival does not extend to screenwriters who win academy awards. For example, screenwriters are not forced to preserve their image by avoiding disgraceful behaviour, maintaining exemplary conduct, keeping physically fit, working regular hours, sleeping each night, or following the ideals of lifestyle. They are not surrounded by people who have a vested interest in the writer's reputation and who can enforce high standards. Outstanding screenwriters, furthermore, win early in their lives yet gain no special control (e.g. influence at work and in the community), ability to avoid stress, access to care, or celebrity privileges. Untangling the possible explanations is difficult because screenwriters are as diverse as their manuscripts.

LimitationsThe main limitation of our research is the missing data on birth, death, and the intervening years. Missing birth data is not a major bias in our analysis because the screenwriters could not have died at an age earlier than the age of any screenwriter who has ever been nominated for an academy award. Likewise, missing data on deaths is not a major bias because differences in survival were still observed in screenwriters known to be dead and because most missing birth dates were for nominees known to be alive. Missing data on time before nomination could be a major bias if suffering in early life leads to outstanding writing in later life. We believe, however, that the priority for future research is to discover how winners and nominees behave after the award ceremony is over.

Behavioural factorsWe suggest that behavioural factors account for the reduced life expectancy of winners, yet biological factors may still be relevant. That is, perhaps a biological factor is linked to both shortened survival and greater talent. If so, the benefits of winning might be masked, the gains from increased status hidden, and our inference misleading. However, such speculation is problematic because the determinants of talent are also prevalent in screenwriters with multiple nominations (but no wins). Such speculation also implies that it is possible to predict a winning manuscript before the film is produced. Finally, such speculation would not explain why writing is so different from other talents and is the first occupation where success leads to a large increase in all cause mortality.³²

Writing versus actingThe results of this study are even more intriguing when compared with results on actors and actresses who have won academy awards.²⁸ In essence, winning an academy award for acting is associated with a large increase in life expectancy, whereas winning an academy award for screenwriting is associated with a large decrease in life expectancy. Both findings are hard to attribute to chance and the simplest explanation for the discrepancy could be the obviousnamely, that the life of an actor is different from the life of a writer. One unifying mechanism for the discrepant health effects may relate to the amount of daily monitoring and control, because actors lead lives of celebrity whereas writers lead lives of anonymity. More research is needed.

Our study is not about writers only, any more than the Whitehall study is about civil servants only.³³ The difference in life expectancy that we found was similar to the societal losses due to heart disease. ^{34 35} This difference occurred even though winners had longer careers, more films, and more nominations (all of which are measures of occupational achievement) than nominees who did not win. The implication is that greater success may lead to worse health in some groups. More generally, biological theories on the link between accomplishment and longevity are unlikely to be the only factors for all people. Behaviour might be a powerful factor that can modulate and even reverse the link between status and survival.

	Acknowledgments

We thank Susan Campbell for data entry, Robert Tibshirani and John-Paul Szalai for statistical insights, and Chaim Bell, David Juurlink, Cam Mustard, Lee Ross, Miriam Shuchman, and Matthew Stanbrook for commenting on drafts of this manuscript.

Contributors: DAR and SMS conceived the idea and designed the study, analysed and interpreted the data, drafted and revised the article, and gave final approval. SMS collected the data. DAR is the guarantor.

	Footnotes

Editorial by Davey Smith

Funding: DAR was supported by a career scientist award from the Ontario Ministry of Health and the de Souza Chair of the University of Toronto. SMS was supported by the Jane and Howard Jones Bursary of the University of Toronto. These funding sources had no role in the design, conduct, or reporting of this study.

Competing interests: None declared.

	References

Top Abstract Introduction Methods Results Discussion References

1.	Marmot MG, Shipley MJ, Rose G. Inequalities in deathspecific explanations of a general pattern. Lancet 1984; 1: 1003-1006[CrossRef][Medline].
2.	Marmot MG, Kogevinas M, Elston MA. Social/economic status and disease. Annu Rev Public Health 1987; 8: 111-135[CrossRef][Medline].
3.	Alter DA, Naylor CD, Austin P, Tu JV. Effects of socioeconomic status on access to invasive cardiac procedures and on mortality after acute myocardial infarction. N Engl J Med 1999; 341: 1359-1367[Abstract/Free Full Text].
4.	Adler NE, Ostrove JM. Socioeconomic status and health: what we know and what we don't. Ann N Y Acad Sci 1999; 896: 3-15[Abstract/Free Full Text].
5.	Lantz PM, House JS, Lepkowski JM, Williams DR, Mero RP, Chen J. Socioeconomic factors, health behaviors, and mortality. JAMA 1998; 279: 1703-1708[Abstract/Free Full Text].
6.	Evans RG, Barer ML, Marmor TR, eds. Why are some people healthy and others not? London: Aldine de Gruyter, 1994.
7.	Cohen SD, Tyrrell DA, Smith A. Psychological stress and susceptibility to the common cold. N Engl J Med 1991; 325: 606-612[Abstract].
8.	Sapolsky RM. Stress in the wild. Sci Am 1990; 262: 116-123[Medline].
9.	Myers J, Bean LL. A decade later: a follow-up of social class and mental illness. New York: Wiley, 1968.
10.	Syme SL. Social and economic disparities in health: thoughts about intervention. Milbank Q 1998; 76: 493-505[CrossRef][Medline].
11.	Garber AM. Pursuing the links between socioeconomic factors and health: critique, policy implications, and directions for future research. In: Bunker JP, Gomby DS, Kehrer BH, eds. Pathways to health: the role of social factors. Palo Alto: Henry J Kaiser Foundation, 1989:271-315.
12.	Krieger N, Williams DR, Moss NE. Measuring social class in US public health research: concepts, methodologies, and guidelines. Annu Rev Public Health 1997; 18: 341-378[CrossRef][Medline].
13.	Lynch JW, Davey Smith G, Kaplan GA, House JS. Income inequality and mortality: importance to health of individual income, psychosocial environment, or material conditions. BMJ 2000; 320: 1200-1204[Free Full Text].
14.	Katz E. The film encyclopedia. 3rd ed. New York: HarperCollins Publishers, 1998.
15.	Walker J, ed. Halliwell's filmgoers companion. 12th ed. New York: HarperCollins Publishers, 1997.
16.	Stewart WT, McClure A, Jones KD. International film necrology. New York: Garland Publishing, 1981.
17.	Thomas N, ed. International dictionary of films and filmmakers: writers and production artists. 2nd ed. Chicago: St James Press, 1990.
18.	Vincendeau G, ed. Encyclopedia of European cinema. New York: Facts on File, 1995.
19.	Sadoul G. Dictionary of filmmakers. Berkeley: University of California Press, 1972.
20.	Thompson D. A biographical dictionary of film. 3rd ed. New York: Alfred A Knopf, 1994.
21.	Manvell R. The international encyclopedia of film. London: Rainbird Reference Books, 1972.
22.	Bawden LA. The Oxford companion of film. New York: Oxford University Press, 1976.
23.	Pickard R. A companion to the movies: from 1903 to the present day. London: Lutterworth Press, 1972.
24.	Langman L. A guide to American screenwriters: the sound era 1929-1982. New York: Garland, 1984.
25.	Corliss R. Talking pictures: screenwriters in American cinema 1927-1973. Woodstock, NY: Overlook Press, 1974.
26.	Andrews G. The film handbook. Boston: GK Hall Publishers, 1990.
27.	Doyle BH. The ultimate directory of film technicians. Lanham, MD: Scarecrow Press, 1999.
28.	Redelmeier DA, Singh SM. Survival of academy award winning actors and actresses. Ann Intern Med 2001; 134: 955-962[Abstract/Free Full Text].
29.	Gail MH. Does cardiac transplantation prolong life? Ann Intern Med 1972; 76: 815-817.
30.	Cox DR, Oakes D. Time dependent covariates. In: Cox DR, Oakes D, eds. Analysis of survival data. London: Chapman & Hall, 1984:112-141.
31.	Lewis M. Stumping for Oscar. Time Magazine 2001 Mar 25:37.
32.	Quint JV, Cody BR. Preeminence and mortality: longevity of prominent men. Am J Public Health 1970; 60: 1118-1124.
33.	Marmot MG, Shipley MJ. Do socioeconomic differences in mortality persist after retirement? 25 Year follow up of civil servants from the first Whitehall study. BMJ 1996; 313: 1177-1180[Abstract/Free Full Text].
34.	Olshansky SJ, Carnes BA, Cassel C. In search of Methuselah: estimating the upper limits to human longevity. Science 1990; 250: 634-640[Abstract/Free Full Text].
35.	Detsky AS, Redelmeier DA. Measuring health outcomesputting gains into perspective. N Engl J Med 1998; 339: 402-404[Free Full Text].