Hand-me-down suicide: Suicide & Birth Order
One of the lesser known suicide facts: younger siblings have higher suicide rates
Over the course of the past decade, increasing research has pointed to an interesting conclusion: birth order is predictive of suicide and suicidal behaviours.
The best that I can tell, birth order has been statistically linked to suicide at the turn of the millenium. Prior to 2000, the closest I can find is a very small university cohort in which Beck inventories were given to 167 college students, finding no link. In 2001, Funahashi et al reported a difference in birth order for schizophrenic patients who died by suicide and those who did not die by suicide. In their report, being a “middle child” was related to suicide risk, and this is after controlling for multiple confounders (p=0.01).
In 2007, Kirkcaldy et al published an analysis of >2,000 youth who were admitted to psychiatric hospitals and found that birth order was indeed related to suicidality. Suicidal thinking was again more common in middle children, who had a 77% (95% CI 35% to 132%) higher odds of having suicidal thinking.
Probably the two “landmark studies” came, of course, from the incredible data we can get out of Scandinavian countries due to their incredible registries.
A 2013 study of a 30-year cohort of of Norwegian siblings were observed for 15 years, and adjustments made for age, sex, date of birth, maternal age revealed suicide risk elevation:
First Born (Reference)
Second Born (+60%, 95% confidence interval +37% to +86%)
Third Born (+112%, 95% confidence interval +62% to +176%)
Fourth Born or higher (+205%, 95% confidence interval +104% to +358%)
Yes you read that right. +200% elevation in fourth born siblings. In fact, each “subsequent birth order” resulted in a +46% (29%-66%) increase in suicide risk. This wasn’t as well-controlled as a 2014 study of a 48-year (!) cohort of Swedish siblings who were observed for 22 years (!), adjusting for age, sex, maternal age, marital status, socioeconomic status, and socioeconomic rank in the sibling group. They found:
First Born (Reference)
Second Born (+17%, 95% confidence interval +5% to +30%)
Third Born (+43%, 95% confidence interval +20% to +69%)
Fourth Born or higher (+63%, 95% confidence interval +27% to +108%)
Paradoxically, both studies found that while younger mothers were less likely to have children who died by suicide, when birth order was controlled for this reversed: older maternal age as associated lower suicide rates when birth order was considered.
Outside of the Scandinavian countries, we see evidence supporting the above.
A British cohort of children born in 1958 was unfortunately underpowered, but they followed for 52 years and found: First Born (Reference); Second Born (+38%, 95% confidence interval -38% to +206%); Third Born (+84%, 95% confidence interval -27% to +368%); and Fourth Born or higher (+127%, 95% confidence interval -10% to +475%)
Obviously, those confidence intervals tell us the picture: this study was remarkably underpowered to look specifically at birth order (only 44 suicides).
A 2014 Taiwanese study established compared to first-born children:
Second Born: males +5% (NS); females +55% (p<0.05)
Third Born: males: +35% (NS); females +90% (p<0.01)
Fourth Born: males: +44% (p<0.05); females +170% (p<0.001)
By this point, it became more useful to examine what new studies demonstrated as to potential mechanisms and subgroups. A 2016 study in Finland saw a birth order effect for Finnish speakers (REF, +26%, +40%, +87%) but not Swedish speakers (REF, +26%, -5%, -10%), suggesting cultural rather than biological factors. A Swiss study of adolescents found being an only child (was associated with higher suicide rates for 10 to 14 year olds, but not 15 to 18-year-olds. Middle children were more likely to die of suicide in the older but not younger adolescent group.
I’ve plotted the studies here. There is substantial heterogeneity!
A very cool recent study used a longitudinal study of parents and children to investigate what was associated with the birth order phenomenon. While they found that maternal depression and paternal absence to contribute slightly to the phenomenon, the bulk of the mediation analysis still left it attributed to birth order alone, meaning they still could not determine the source of the effect. A 2019 Lancet metanalysis found a similar number of studies, and the funnel plot indicated very little chance of publication bias.
What are the suggested mechanisms?
Early life circumstances can affect risk throughout life.
Older children may bully younger ones.
Possible weaker attachment between later-born children and parents.
Parents may have limited time and emotional resources for later-born kids
Parental resources might be unequally distributed among siblings.
Parental investment could decline as family size increases.
Higher birth order children might have reduced access to parental attention and supervision.
This known association is under-reported and counter-intuitive. Younger siblings are often seen as coddled and protected, and only children are seen as “problems.” I am dubious that there is a strong biological effect - the 2016 Finnish study showing that Swedish speakers vs Finnish speakers obliterated that effect suggests to me that cultural and environmental factors play the largest role here. There’s good news in that: environmental factors are modifiable!