Research on adoptees and suicide

Content warning: I discuss suicide and mental health in this post.

Over the past few years I have seen a huge increase in discussions among adoptees and adoption professionals about suicide among adoptees. In the adoptee community, this has been a sad reality we have known about for a long time, but it has taken quite some time for it to be acknowledged. I think this is because adoption professionals and adoptive parents do not want to believe that trauma is at the root of adoption or that suicide may be connected to adoption.

The findings on adoptees and suicide have been mostly mixed (those finding no risk, others finding risk), although in recent years there have been more studies that found higher rates of suicide attempt. I’ll discuss the recent literature, but first I wanted to summarize the older research.

One of the interesting segments of the research on adoptees and suicide is the research focused on the biological aspect of suicide among adoptees; that is, whether the adoptee’s biological parents/family had history of psychiatric disorders or suicide. In other words, researchers thought that the genetic aspect of suicide was a larger factor than the environmental aspects (such as the experience of adoption in general or the role of the adoptive family environment). This research appears to grow out of the interest in the role of genetics in general on suicide and mental health. Outside of the U.S., a lot of the research on adoptees and suicide come from countries that have nationalized health data¹; this level of population data allows for a greater ability to analyze the population across a country and with enough numbers for comparison (i.e. adoptees and non-adoptees), something that the U.S. is unable to do because we do not have national health data collected by the government. As a result, the best we can do in the U.S. is to attempt to get a large and diverse sample using randomized sampling, and this is difficult and costly. This is probably why the authors of the study I describe below drew from a study that had a sample of families that included adoptees and their non-adopted siblings.

Most of the adoptees I see sharing the statistics of adoptee suicide attempts is based on the study published in the journal Pediatrics in 2013 by Margaret A. Keyes, Stephen M. Malone, Anu Sharma, William G. Iacono, and Matt McGue,² so I wanted to focus on this article in particular because I’ve been seeing some misrepresentations of the findings. I’m going to use the framework of questions to ask when reading adoption research (posted on my @harlows_monkey Instagram account and included in this article).

Who is the author and what is their world view? I typically like to see if I can find out more about the researcher’s positionality. What do I mean by positionality? I mean, is the researcher an adoptee, an adoptive parent, related to an adoptee/adoptive parent (i.e. sibling, partner), or someone with no connection to adoption? This is important to me because I’ve found that many academics and researchers are also adoptive parents, and this informs why they choose to research adoption, informs their questions, methodology, and how they interpret their findings. This isn’t wrong, it just is – our lived experiences and world views naturally inform how we would approach research. Adoptee researchers also are influenced by our experiences. Those who have no personal lived experience also have their own biases and opinions that can come through in the research. Sometimes, however, researchers have an agenda and this can show up in their research so again, it’s always helpful to know more about the researcher. I was not able to determine the positionality of all of these researchers in the Keyes et al. study but at least one is an adoptive parent. I could not determine if any of the authors are adoptees.

Who is in the sample? This study’s particular sample included 672 adoptees and 515 non-adoptees recruited from a larger sample of adoptive and non-adoptive families (called the SIBS data). The original sample of adoptive families came from 3 adoption agencies in Minnesota and the children were all between 11 and 21 years at the time the first round of data was collected (1998-2004). These families were invited to participate in the current study. One thing to note is the high participation rate of Korean adoptees in the original sample – 74% of the adoptee sample were intercountry adoptees and 90% of these were from South Korea. Since the original recruitment of the families all came from Minnesota, and Minnesota has such a high rate of Korean adoption, it’s helpful to note this aspect of the sample. The authors did not provide a description of the number of Korean adoptees in the follow up study, so we can’t say if this was true of the current study but given the high number of participants from the original study, it is likely to be similar. I’ll say more about that later when I discuss the analysis.

What is the unit of analysis? The unit of analysis is the adoptees and the non-adoptees.

What tools for assessment were used? Several of the tools for assessment were completed by non-adoptees/siblings.

• Both parents and their children were asked about their child’s suicide attempt.
• The children were assessed using the DSM-IV (which has now been updated to DSM-V), the standard clinical diagnostic tool used by mental health clinicians. I have critiques about this assessment tool, but that’s a discussion for another day.
• Parents were also asked a few questions related to mental health.
• Teachers completed two questionnaires that asked about the child’s behaviors.
• Both parents and children completed Parental Environment Questionnaire scales.
• The Multidimensional Personality Questionnaire was used but the article did not say if the children completed this part or if the parents did on their behalf.
• Parents were asked about their child’s academic engagement.

What is the methodology? These researchers used a quantitative approach. There was no discussion if the researchers used or analyzed any open-ended comments.

What are the research questions? The researchers had three questions:

• Does adoption status represent a risk of suicide attempt among the sibling participants?
• Are known risk factors for suicidal behavior elevated in study participants who have reported suicide attempts compared with those who have not reported attempts?
• Do these factors mediate the potential relationship between adoption status and suicide attempt?

What are the limitations? The authors stated that limitations include the lack of a more diverse sample, that adoptees were represented from only some adoption agencies (3 agencies in Minnesota), and did not include foster care adoptees or adoptees that were placed in private independent arrangements. They also included the lack of information on prior history, biological family history, and more detailed information about nature of the suicide attempts as limitations.

Some of the findings I find noteworthy:

• 56 out of 1165 children were reported to have attempted suicide, representing 4.7% of the full sample
  • 5.4% of male adoptees and 8.6% of female adoptees were reported to have attempted suicide
• The adoptees were ~4 times more likely to report attempt suicide than their non-adopted peers
• This finding was consistent with Hjern et al.’s study (see source #1 below)
• Age at placement, ethnicity, or type of adoption (domestic, intercountry) was not correlated to suicide risk
• There is not a table showing demographics, it is written out. Therefore, I was not able to look at the descriptive statistics related to age at placement, ethnicity, or type of adoption and we are only able to go by what is written.

Summary: It is important to consider this study in its context. I think what is valuable to me is that the findings are similar to the Swedish cohort studies, which had the ability to have larger numbers in their sample and comparison to a larger non-adopted sample. Both found roughly a ~4 times greater odds for suicide attempt among adoptees.

While the authors did say a limitation was the lack of diversity, and noting they did not find a difference between the intercountry adoptee sample and the domestic adoptee sample, I find it really interesting that so much of the adoptee sample were Korean adoptees.

I think it’s also important to note that the numbers of those who reported suicide are fairly small and are not large enough to do a statistical power analysis – meaning that their findings that there are no differences between intercountry adoptee sample and domestic adoptee sample might be limited. I would have liked to have seen a table with those findings.

I hope that this has been a helpful analysis of this important article. I wrote it because in part, it’s a topic that is much on the minds of the adoption community, and also because I’ve seen some misrepresentation of the results. It’s really important to understand the difference between causation and correlation. We cannot say that adoption causes suicide attempt or death. We can only say there are correlations between adopted people and suicide attempt/death. There is no way we can do a study that would prove that adoption causes suicide attempts/death because we cannot control for all of the factors that might also be a reason for adoptees to attempt suicide. Also, the study does not find that adoptees are ~4 times more likely to die by suicide, which I have seen being reported. As a last note, I hope this helps the adoption community recognize that much more research on this topic is needed.

Sources

¹ Hjern, A., Lindblad, F., & Vinnerljung, B. (2002). Suicide, psychiatric illness, and social maladjustment in intercountry adoptees in Sweden: a cohort study. The lancet, 360(9331), 443-448. Also, Von Borczyskowski, A., Hjern, A., Lindblad, F., & Vinnerljung, B. (2006). Suicidal behaviour in national and international adult adoptees. Social psychiatry and psychiatric epidemiology, 41(2), 95-102.

² Keyes, M. A., Malone, S. M., Sharma, A., Iacono, W. G., & McGue, M. (2013). Risk of suicide attempt in adopted and nonadopted offspring. Pediatrics, 132(4), 639-646.