Week 5 - Comments to readings

Week 5 - Comments to readings

by Kate Chirikova -

1.     O’Flaherty et al. 2015

I very much enjoyed reading this paper! I think it’s a great example of how theory and appropriate analytical design can lead to a meaningful empirical work. When authors discussed the lifecourse approach and all the interconnected factors that might affect health in the long-term – like transition between multiple social roles affected by available resources which are in turn affected by role-performing behaviors and cumulative advantage – I was skeptical of how one could capture all these mechanisms in relation to delayed health outcomes. But I was impressed by the creative methods that were used to derive meaningful family lifecourse groupings (which was the main exposure variable). I haven’t heard of a sequence analysis before, and it sounds somewhat similar to cluster analysis in machine learning, but I would like to look more into it.

As for the conclusions, I was surprised to see that for women only a disrupted marital history and a high level of fertility were found to be linked to poorer health outcomes. I would expect more factors (and combination of factors) playing a role in the delayed health outcomes. I wonder if it is a real thing or a consequence of those analytical steps that were undertaken during data analysis.

2.     Gupta et al. 1997

Very disturbing evidence from India. This article is dated 1997. I wonder how much has changed since then. Is there a decrease in sex bias among young children and at birth? Or has a larger shift happened towards pre-natal sex regulation (i.e. abortion) vs. post-natal (i.e. infanticide)?


Week 5 - Comments to readings

by Carolyn Hughes -
I agree: re: The O'Flaherty paper. And I hadn't heard of sequence analysis either, so I'm not sure I'm fully following what they did. I agree that the differences in the factors associated with poorer health outcomes for men vs women were surprising. Like you said, it would be interesting to know why this is the case, and what proportion is due to differences between men and women in life expectancy and overall health later in life (I'm assuming it's minimal and was controlled for in the study, but not sure). It would be interesting to see how these life interruptions impact care-seeking behaviors, nutrition, and other factors that then can influence health later in life, and to what extent; the paper mentions it toward the beginning, but the analysis only includes parental smoking (from what I could see in the tables).

In Table 5, I thought it was odd that "No family formation" was negatively associated with physical health in Model 4 (-2.9), but then was positively associated in Model 5 (2.4); it seems odd that the association would "flip" like that, especially since none of the other associations seemed to change much, and I didn't see a mention of it in the results or discussion. Perhaps it's some kind of artifact or a typo (missing "-" sign?)?

Re: Week 5 - Comments to readings

by Leah Koenig -
Hi both,

I echo a lot of the comments you made about the O'Flaherty paper and was also interested in understanding the differences between men and women. Some of the explanations that came to mind were differences in health-seeking behaviors (as Carolyn raised), social connectedness, and financial stress.

I was also interested in the classifications of their exposure - specifically that the Late Family Formation group characterized by "marriage in the 30s followed by low to moderate fertility commencing in the mid-30s". Given that the mean age at first childbearing in Australia in 2019 was 29.4, I wonder if the authors would have characterized the groupings with different age bounds if they were using more recent data.

Re: Week 5 - Comments to readings

by Richard Hu -
It's heartbreaking to read about what sex bias in India in the 80s and 90s, and I'm sure that a preference for boys still exists today, although I'm also curious how the rates of abortion and infanticide have changed since Das Gupta's study. I know little about the situation in India but know a bit more about the East Asian countries - I wonder if the bias in India is also due to (among other things) the cultural expectation that sons are responsible for caring for their parents in old age? It was also interesting (but shocking) to read that the education level of women had little effect on the sex bias, and implied infanticide was occurring at such high rates despite there being no mandate (that I know of) akin to the One Child Policy in China, and therefore were probably entirely economical in decision.