Search for mechanisms of exceptional human longevity Dr. Natalia S. Gavrilova, Ph.D. Dr. Leonid A. Gavrilov, Ph.D. Center on Aging NORC and The University of Chicago Chicago, USA

Approach To study “success stories” in long-term avoidance of fatal diseases (survival to 100 years) and factors correlated with this remarkable survival success

Centenarians represent the fastest growing age group in the industrialized countries Yet, factors predicting exceptional longevity and its time trends remain to be fully understood In this study we explored the new opportunities provided by the ongoing revolution in information technology, computer science and Internet expansion to explore early-childhood predictors of exceptional longevity Jeanne Calment ( )

Revolution in Information Technology What does it mean for longevity studies? Millions of official census, birth, marriage, death and other records are available online now!

Predictors of Exceptional Longevity

Study 1 How centenarians are different from their shorter-lived sibling?

Within-Family Study of Exceptional Longevity Cases Centenarians born in U.S. in Controls – Their own siblings Method: Conditional logistic regression Advantage: Allows researchers to eliminate confounding effects of between- family variation

Design of the Study

A typical image of ‘centenarian’ family in 1900 census

First-born siblings are more likely to become centenarians (odds = 1.8) Conditional (fixed-effects) logistic regression N=950, Prob > chi2= VariableOdds ratio95% CIP-value First-born status Male sex <0.001

Birth Order and Odds to Become a Centenarian

Can the birth-order effect be a result of selective child mortality, thus not applicable to adults? Approach: To compare centenarians with those siblings only who survived to adulthood (age 20)

First-born adult siblings (20+years) are more likely to become centenarians (odds = 1.95) Conditional (fixed-effects) logistic regression N=797, Prob > chi2= VariableOdds ratio95% CIP-value First-born status Male sex <0.001

Are young fathers responsible for birth order effect? Conditional (fixed-effects) logistic regression N=950, Prob > chi2= VariableOdds ratio95% CIP-value Born to young father Male sex <0.001

Birth order is more important than paternal age for chances to become a centenarian Conditional (fixed-effects) logistic regression N=950, Prob > chi2= VariableOdds ratio95% CIP-value First-born status Born to young father Male sex <0.001

Are young mothers responsible for the birth order effect? Conditional (fixed-effects) logistic regression N=950, Prob > chi2= VariableOdds ratio95% CIP-value Born to young mother Male sex <0.001

Maternal Age at Person’s Birth and Odds to Become a Centenarian

Birth order effect explained: Being born to young mother! Conditional (fixed-effects) logistic regression N=950, Prob > chi2= VariableOdds ratio95% CIP-value First-born status Born to young mother Male sex <0.001

Even at age 75 it still helps to be born to young mother (age <25) (odds = 1.9) Conditional (fixed-effects) logistic regression N=557, Prob > chi2= VariableOdds ratio95% CIP-value Born to young mother Male sex <0.001

Being born to Young Mother Helps Laboratory Mice to Live Longer Source: Tarin et al., Delayed Motherhood Decreases Life Expectancy of Mouse Offspring. Biology of Reproduction :

Possible explanation These findings are consistent with the 'best eggs are used first' hypothesis suggesting that earlier formed oocytes are of better quality, and go to fertilization cycles earlier in maternal life.

Study 2 How centenarians are different from their shorter-lived peers when compared at young adult age?

Physical Characteristics at Young Age and Survival to 100 A study of height and build of centenarians when they were young using WWI civil draft registration cards

Height – What to Expect 1. Height seems to be a good indicator of nutritional status and infectious disease history in the past. 2. Historical studies showed a negative correlation between height and mortality. 3. Hence we may expect that centenarians were taller than average

Build – What to Expect 1. Slender build may suggest a poor nutrition during childhood. We may expect that centenarians were less likely to be slender when young. 2. On the other hand, biological studies suggest that rapid growth may be harmful and somewhat delayed maturation may be beneficial for longevity.

Small Dogs Live Longer Miller RA. Kleemeier Award Lecture: Are there genes for aging? J Gerontol Biol Sci 54A:B297–B307, 1999.

Small Mice Live Longer Source: Miller et al., The Journals of Gerontology Series A: Biological Sciences and Medical Sciences 55:B455-B461

Design of the Study

Data Sources 1. Social Security Administration Death Master File 2. WWI civil draft registration cards (completed for almost 100 percent men born between 1873 and 1900)

WWI Civilian Draft Registration In 1917 and 1918, approximately 24 million men born between 1873 and 1900 completed draft registration cards. President Wilson proposed the American draft and characterized it as necessary to make "shirkers" play their part in the war. This argument won over key swing votes in Congress.

WWI Draft Registration Registration was done in three parts, each designed to form a pool of men for three different military draft lotteries. During each registration, church bells, horns, or other noise makers sounded to signal the 7:00 or 7:30 opening of registration, while businesses, schools, and saloons closed to accommodate the event.

Registration Day Parade

Information Available in the Draft Registration Card age, date of birth, race, citizenship permanent home address occupation, employer's name height (3 categories), build (3 categories), eye color, hair color, disability

Draft Registration Card: An Example

Study Design Cases: men centenarians born in 1887 (randomly selected from the SSA Death Master File) and linked to the WWI civil draft records. Out of 120 selected men, 19 were not eligible for draft. The linkage success for remaining 101 records was 75% (76 records) Controls: men matched on birth year, race and county of WWI civil draft registration

Height and Survival to 100

Body Build and Survival to 100

Multivariate Analysis Conditional multiple logistic regression model for matched case-control studies to investigate the relationship between an outcome of being a case (extreme longevity) and a set of prognostic factors (height, build, occupation, marital status, number of children, immigration status) Statistical package Stata-10, command clogit

Results of multivariate study VariableOdds Ratio P-value Medium height vs short and tall height Slender and medium build vs stout build 2.63*0.025 Farming2.20*0.016 Married vs unmarried Native born vs foreign b

Results of multivariate study Significant predictors only VariableOdds Ratio P-value ‘Slender’ body build reference: stout build ‘Medium’ body build reference: stout build Farming

Other physical characteristics VariableOdds Ratio P-value Blue eye color ‘Short’ body height reference: tall height ‘Medium’ body height reference: tall height Other variables include body build and farming

Having children by age 30 and survival to age 100 Conditional (fixed-effects) logistic regression N=171. Reference level: no children VariableOdds ratio95% CIP-value 1-3 children children

Conclusion The study of height and build among men born in 1887 suggests that rapid growth and overweight at young adult age (30 years) might be harmful for attaining longevity

Conclusion The study of height and body build among men born in 1887 suggests that obesity at young adult age (30 years) has strong long-lasting effect in preventing longevity

Other Conclusions Both farming and having large number of children (4+) at age 30 significantly increased the chances of exceptional longevity by %. The effects of immigration status, marital status, and body height on longevity were less important, and they were statistically insignificant in the studied data set.

Acknowledgments This study was made possible thanks to: generous support from the National Institute on Aging and the Society of Actuaries

For More Information and Updates Please Visit Our Scientific and Educational Website on Human Longevity: And Please Post Your Comments at our Scientific Discussion Blog:

Multivariate Analysis: Conditional logistic regression For 1:1 matched study, the conditional likelihood is given by: Where x i1 and x i0 are vectors representing the prognostic factors for the case and control, respectively, of the ith matched set.

Final Conclusion The shortest conclusion was suggested in the title of the New York Times article about this study

New Striking Findings: Invitation for discussion and brain-storming! The favorable "Young Mother Effect" is particularly strong when parents have particularly large differences in their lifespan

Odds Ratio to live to 100 years if born to young mother as a function of maternal and paternal lifespans (tertiles) MOTHER FATHER Shorter-livedMedium-livedLonger-lived Shorter-lived * Medium-lived 3.49* Longer-lived 11.62* * p<0.05

Month of Birth and the Likelihood to Become a Centenarian among WWI draft participants Method: Conditional logistic regression for odds to become a centenarian. Controls – men matched on birth year, race and county of draft registration. Adjusted for occupation, height and build. 76 observations

Month of Birth and the Likelihood to Become a Centenarian for Adult Siblings Method: Conditional logistic regression for odds to become a centenarian, using siblings as within- family control. 787 observations

Question Families were quite large in the past, particularly those covered by genealogical records (large family size bias). Is the "young mother effect" robust to the family size, and is it observed in smaller families too? Or is it confined to extremely large families only? Approach: To split data in two equal parts by median family size (9 children) and re-analyze the data in each group separately.

Results In smaller families (less than 9 children) the effect of young mother is even larger: Odds ratio = 2.23, P=0.004; 95%CI = Compare to larger families (more than 9 children): Odds ratio = 1.72, P=0.11; 95%CI = Conclusion: "Young mother effect" is not confined to extremely large family size