Science: Family Longevity

David Brown
Washington Post Science Writer
Tuesday, January 16, 2007; 11:00 AM

A new study has measured the "fitness cost" of human reproduction. This is the price that parents pay in their own health and longevity for the privilege of having their genes live on in future generations.

Washington Post science writer David Brown discussed the findings 11 a.m. ET on Tuesday, Jan. 16.

A transcript follows.

Read more about this in the article by David Brown: Children, Parents Drive Each Other to Early Graves (Jan. 15.)


David Brown: Welcome readers and demographers! (There seem to be a lot of the latter on this chat). I will try to answer some of your questions about the story in yesterday's paper about the "fitness cost of human reproduction". The paper appreated in the Proceedings of the National Academy of Sciences and analyzed data from Utah. I warn you, however, that the expertise of many of you exceeds me on these matters, so I may not be able to be terribly helpful on some questions. So now may we please to begin . . .


Syosset, N.Y.: The abstract in PNAS does not indicate that any consideration of social-economic status was included. Even today, with modern access to technology, there is a clear correlation between number of children and social-economic status, and also a strong correlation with that status and longevity for both men and women. This is a more likely explanation of the effect on men, as well as women, than the conjectures mentioned. - Lester, MD

David Brown: There were obviously a number of semi-fine points I couldn't address in the story, and this is one of them. The researchers did consider the effects of socioeconomic status for a subgroup of their population, namely about 5,000 couples (out of the total of nearly 22,000) in which the father died in 1904 or later, that being the period in which occupation is listed on the death certificate. There is a metric that converts occupation to income and SES that I was assured by one of the researchers is considered valid. In this subgroup, low SES was associated with higher mortality, which assured them that in fact they were measuring low SES, but the overall gradient of more children leading to shorter lives, with the effect much more evident in women than in men, was still observed, as did all the other major conclusions.


Woodbridge, Va.: This study was interesting but did they adjust the findings for the average life expectancy current at the time of the sampling? In other words, how can you compare the infant mortality and maternal and paternal mortality based on today's current life expectancy?

David Brown: I guess I don't quite get this question. Life expectancy was presumably rising during the period covered in this study (couples marrying between 1860 and 1895). That trend would have been experienced by everyone, regardless of family size. The study addresses relative risks of the sexes by family size, and the risk to children of mother loss, late birth order, and the presence of many siblings. None of this really has to do with today's life expectancy.


Hanover, Pa.: Good Morning Mr. Brown. I have not taken the time to read the study so please forgive my willful ignorance, but my question involves the INVOLVEMENT of extended family members. Doesn't the loss of grandparents and/or other wise family members from the nuclear family model set us up for fission? Thank you for your time and consideration.

David Brown: Again, there was obviously great heterogeneity between families, including the important issue of the presence of other relatives in the households that you suggest. The researchers did not look at how families that were nuclear-only fared compared to ones with maiden aunts, unmarried uncles, grandparents, etc. to help out. They did note, however, that couples marrying between 1875 and 1895 were more likely to be native Utahns, which suggests that the younger half of the study population (couples married from 1860 to 1874) were probably pioneers, and that many of them probably left parents back home wherever they had left from.


Potomac, Md.: Was age at death - not just years surviving after last child - correlated with number of children? (This may have been examined in the study, but was not reported in The Post.) Consider the following: Since women having fewer children would be expected to end their child-bearing years early, they would be expected to survive for more years after their last child's birth than would women who had more babies and ended their bearing years later in life. The first group would create more separation between end of bearing years and predicted mortality than the second group. Statistically, if every woman in the study died at age 45, the data would still support the report's hypothesis but neglect to examine this fact.

David Brown: The paper did not discuss age at death and number of children. It did find that the survival of women after age 50 (which was long after the last birth in most cases, that occurring at age 39 on average) was affected by the number of children they had, with women who had borne lots of children having shorter post-50 survival than women who had borne fewer. So this appears to be a pretty potent effect.


Fort Washington, Md.: Since this data examines a time period before major breakthroughs in public health and the concept of microorganisms (germs), how relative is this data to 2007 parents and people planning to be parents? For example, the Washington Post did a series noting that the leading cause of death of pregnant women was murder. How were the factors such as the flu epidemic in the early 1900's considered? Thank you!!

David Brown: As you suggest, these findings are not relevant to parents of 2007. What makes these data interesting is that they are from an era when little, if any, contraception existed, and most (although not all) of the medical, sanitation, and public health advances that are responsible for our incredible longevity didn't exist.


Laurel: A couple points about possible biases in the study: 1. Obviously, the conditions that existed from 1860-95 might be different from today. For instance, the number of children to which a woman gave birth may not reflect "reproductive choice" the way we currently understand it. 2. Fertility is usually correlated with the expectation that the children will survive to majority. Modern Americans can almost always expect that their children who live to two years without obvious threatening conditions will survive to adulthood, so we have the children we want an then stop. Where survivability is only 50 percent, couples have 6-8 children hoping that AT LEAST 2 will reproduce, but on average more will. 3. Having 12 babies by age 39 means a woman is almost constantly pregnant. Not recovering fully from the last childbirth can be more damaging than the births or children themselves.

David Brown: A couple of points here I might comment on. I think the assumption on the part of the researchers is that there was little reproductive choice in these women. (Of course, fertility can vary greatly, so even in the absence of contraception there can be a great range of family size). About 65 percent of the couples were Mormon, so there may have been somewhat more of an imperative to have large families in this group than in the frontier population as a whole. Around the world and in all cultures, the number of children a woman bears goes down as childhood mortality falls, child survival rises, economic status rises, and educational status of girls and women rises. This is the so-called "demographic transition" that the industrialized countries of the West went through in the late 19th and early 20th centuries, but that some countries elsewhere in the world are still in the middle of.


Laurel, Md.: On Saturday there was a report on Page D2 of the latest outbreak of _E. coli_. All references to the bacterium were italicized in the body of the article. However, the first reference did not present the generic name in full, there was no italicization in the headline, and the specific name was incorrectly capitalized in the headline. Why pays the fines for violating the rules of binomial nomenclature?

David Brown: I'll answer this only because I agree with you and sent a memo to one of the assistant managing editors of such matters during the big Escherichia coli [itals] a couple of months ago. We either seem to have a problem following our own style, or it has changed and I don't know about it.


SoCal: My wife and I are having our third child, already having a 4 and 2 year old, I can surely tell you that my kids are going to put me in an early grave! Please tell me the toddler years end! In all seriousness, having children is a God-given responsibility that teaches one -- the parent -- that their own selfish needs are secondary to that of another human being. While obviously children are the result of procreation, people willingly embrace child rearing due to the fact that only by giving ourselves away do we find true happiness. In any scientific endeavor, there is a human side to the equation as well. Science can explain only so much, the spiritual often finishes the question.

David Brown: Sounds good.


Annandale, Va.: I am curious; was there any correlation between age of the parents and survival rate of the children?

David Brown: That question was not addressed in the study. The data did show, however, that the older a woman was at her first delivery, the more her mortality rose in her post-reproductive years.

I also want to add something that a previous questioner brought up. The data also showed that short birth intervals--i.e., the state of being almost constantly pregnant--was also associated with lower survival in the post-reproductive years, with the association again being much stronger for women than men. So no surprise there either.


Savage, Md.: What is the full bibliographic reference to the PNAS article, and why did you omit it from your story?

David Brown: The paper was published online by PNAS on December 27th, I believe. I got an early edition pdf and can't provide the full citation, I'm afraid.


Falls Church, Va.: Were the variances in death rates according to number of children statistically significant? I assume so, or else this study is much less important, but the article didn't expressly say so.

David Brown: Every finding I mentioned in the paper was statistically significant. Something that I didn't have a chance to get into in the story was the issue of the so-called "grandmother effect". That is the idea that there is an evolutionary advantage for women to survive long after their reproductive years are over in order to aid in the upbringing and survival of their offspring's offspring, that is their grandchildren. A study of Finnish and Canadian populations from the 1700s and 1800s published a few years ago found evidence of a grandmother effect. This study found that women who died prematurely had fewer grandchildren, but that was primarily because they had fewer children, not because they weren't able to be grandmothers. When women who had the same number of children but either survived into post-menopausal years or did not were compared, there was a non-statistically significant trend for the surviving women to have more grandchildren. So there was a hint at the grandmother effect, although it did not reach the limit of statistical significance.


Silver Spring, Md.: Any indication of what the mothers died from? Increased child-bearing is supposed to be linked to lower breast cancer incidence. If that is true then women with more children should have been dying from something else.

David Brown: Nothing was mentioned on cause of death. I agree that would be very interesting to investigate, although I assume that for most of that period there were either no death certificates or ones that probably din't contain that information. But that's just a guess.


Falls Church, Va.: Were the deaths of the younger children in large families attributable to nutritional deficiencies; i.e, to the family being unable to support so many children? In other words, they can't afford as much food to go around once the 10th, 11th, and 12th children arrive, and the youngest children are hit the hardest by this lack of nutrition, since they're the weakest.

David Brown: There was nothing mentioned about the cause of death in the younger children. One imagines that it was multifactorial, with nutritional deficiency, possibly the need to go enter the labor force early, less adult supervision, possibly more dilute parent-child bonds, all played a part.


David Brown: Okay, I guess that's it. Thanks for reading and tuning in. I enjoyed it.


DC: Here's the citation from the PNAS website:

Dustin J. Penn and Ken R. Smith

Differential fitness costs of reproduction between the sexes

PNAS 2007 104: 553-558; published online before print December 27 2006, 10.1073/pnas.0609301103 -Abstract] -Full Text] -Figures Only] -PDF] -Supporting Tables]

David Brown: thanks for this.


DC: We really need to know causes of death to make much sense of these findings. Are there any databases available that would provide that information for a followup study? Is there any additional work being done by these authors on this? For example, infectious diseases were among important causes of death in these years; would more kids mean more exposure risk?

David Brown: Good question; there may be size-dependent risks in big families. I do not know if such data bases exist, or if any of the Utah data bases contain them. You might want to try to reach one of the researchers.



Editor's Note: moderators retain editorial control over Live Online discussions and choose the most relevant questions for guests and hosts; guests and hosts can decline to answer questions. is not responsible for any content posted by third parties.

© 2007 The Washington Post Company