(Updated Post) Reviewing the evidence on donor-conceived psychological outcomes
Comparing the psychological outcomes of donor and non-donor conceived people: A systematic review. (Talbot, 2024)
Talbot, C., Hodson, N., Rose, J., & Bewley, S. (2024). Comparing the psychological outcomes of donor and non-donor conceived people: A systematic review. BJOG: An International Journal of Obstetrics & Gynaecology, 131(13), 1747–1759. https://doi.org/10.1111/1471-0528.17892
Editor’s Note: This was one of the first papers I reviewed. As part of ongoing review of all of my content (my approach has evolved a lot, yall!), and based on feedback, I revisited this paper. I corresponded with the lead author regarding my concerns about analysis and citation discrepancies. The author indicated they had been made aware of issues with the citations recently and acknowledged the other feedback. This summary will be further updated if a revised paper is published.
Geographic Region: Primarily high-income anglophone countries: United Kingdom (18 studies), United States (12), Australia (6), Netherlands (4), with additional studies from France, Sweden, and multi-country samples.
Research Question: What are the psychological outcomes donor-conceived people throughout their life course and how to they compare with those of non-donor-conceived people?
Methods: Systematic review inventory of both quantitative and qualitative studies (e.g., cross-sectional, case-control, and longitudinal cohort studies). Inclusion criteria were five or more participants, peer review, and assessment of any donor conception psychological outcomes. The inventory included 40 cross-sectional articles (34 quantitative, five qualitative, one mixed methods), three case-control, and seven longitudinal cohorts (one qualitative). The articles employed 66 different psychological outcome measures, most commonly the Strengths and Difficulties Questionnaire (11 papers), Child Behaviour Checklist (10 papers), Golombok Rust Inventory of Marital State (8 papers), Trait Anxiety Index (9 papers), and Parenting Stress Index (8 papers). Outcomes were derived from parent-, teacher-, child-, and professional-rated tools. Quality assessment used Joanna Briggs Institute critical appraisal tools specific to each study design.
Editor’s Note: The authors say they’re reviewing for psychological outcomes but never clearly define what that means. They end up including a wide range of psychosocial and developmental measures, including neurodevelopmental conditions (autism spectrum disorder, ADHD), behaviors (aggression, substance use, parenting), relationships (parent-child warmth), and social/academic functioning. The categorization of autism as ‘worse psychological outcomes’ is particularly bothersome because it’s scientifically inaccurate: donor conception can’t cause neurodevelopmental conditions.
While this paper follows proper systematic review procedures, the actual analysis and synthesis are a bit messy. The authors couldn’t combine studies statistically because they were too different from each other. That’s totally fair. Instead, they sorted quantitative findings into three buckets: "better," "same," or "worse" for donor-conceived people. Without clear decision rules for how findings get sorted, bias and miscategorization become likely. When I crosswalked the in-paper tables, the supplemental table, and a number of the papers cited to compare orginal findings to the authors’ summaries, I disagreed with how some findings were framed and categorized, and found a number of flat-out errors. For example, the authors list six papers that had "worse" outcomes for DCP, but two of the findings in the table actually show DCP doing better than comparison groups and two misrepresent the study findings (e.g., the table states one study showed 'higher rates of conduct disorders' in donor-conceived females but the actual finding was that while the score for conduct problems - subscale of SDQ - in female DCP was higher than expected, the overall scores were not significantly different from the Australian general population and were well within normal range). Only three of the seven included qualitative papers center donor-conceived people's own perspectives (totaling approximately 38 participants). Two report parent perspectives and one examines how young donor-conceived children understand their family composition.
Sample: The inventory includes data from 4,666 donor-conceived participants and 3,790 non-donor-conceived participants across 50 articles, with mean study size of 93 participants (range: 10-528). Publication years spanned 1993-2023, with 70% published during or after 2010. Participant age distribution showed 26 articles (52%) included only children under age 12, nine (18%) focused on adolescents aged 13-17, and 14 (28%) examined adults over age 18, with one spanning ages 10-25.
Editor’s Note: The authors note that "two cohort studies followed up at more than two time points.” In reality, at least 22 of the 50 articles (44%) draw from five longitudinal studies, meaning they represent the same families measured at different ages rather than independent groups of people. Specifically, eight articles come from the U.S. National Longitudinal Lesbian Family Study, 12 papers come from three separate Golombok longitudinal cohorts, and two Kovacs papers are from the same cohort. When the same families appear multiple times across different papers, counting each paper as a separate study inflates the apparent size and diversity of the evidence base. What looks like broad agreement across 50 independent studies is actually much narrower.
Key Findings:
Editor’s Note: Donor conception itself is just a method of family formation. It doesn’t directly cause either anxiety and attachment difficulties or high self-esteem and strong family bonds. So when research finds differences in well-being, we need to ask: what is the actual mechanism of harm or protection? Is harm caused by being deceived? By social stigma? By family dynamics around infertility? And when donor-conceived people do well, what protects them? Open communication? Supportive families? Social status? The critical question isn’t whether donor-conceived people differ from others, but what specific, modifiable factors shape their experiences. Without identifying these mechanisms, findings can only be filtered through cultural narratives about ‘real’ families rather than what actually helps people thrive.
This systematic inventory provides a reference point for the field by documenting what research exists through early 2024. It maps the research landscape rather than providing a definitive synthesis of psychological and well-being outcomes. With more careful abstraction and a different synthesis framework (perhaps outcome domain and age), we’d be able to better visualize patterns in findings and identify research gaps. The takeaways below are what I’m comfortable extracting from the paper.
Across broad measures of mental health and well-being, the majority of donor-conceived children and adults show outcomes comparable to those of comparison groups and fall within normal ranges.
Some studies found better outcomes for specific subpopulations of donor-conceived children when compared to non-donor-conceived individuals.
A handful of studies found worse outcomes for specific subpopulations of donor-conceived children and adults when compared to non-donor-conceived individuals. In some cases, the outcomes are “worse on paper” and still in the normal range.
Editor’s Note: I found the discussion a bit peculiar. The authors make recommendations that contradict what they found. For example, they state donor-conceived children are more likely to have behavioral problems and then recommend parenting programs, which doesn’t at all reflect the actual findings. A quick google search revealed that one author has a professional interest in parenting interventions. Coincidence?
Limitations: In addition to everything else described, systematic review limitations include varying quality across included studies and challenges in comparing studies with different methodologies and sample sizes. The review also noted research samples typically over-represented families who disclosed donor conception to children and predominantly relied on self-selected volunteers who likely represented more engaged and motivated community members.
Applications: Future research should develop recruitment strategies that reach non-disclosing families, families disconnected from donor conception communities, and more diverse populations. Systematic reviews should move beyond simply cataloging findings to providing interpretive synthesis that helps readers understand patterns, contradictions, and gaps in the literature. Future research should consider using more uniform measures of psychological and mental health outcomes. Reviews should avoid overgeneralizing findings from narrow samples to all DCP.
Editor’s Note: Peer reviewers and editors have an important role to play in improving the quality of donor conception research. People are making decisions based on this research. That raises the stakes for getting it right. Please help authors give us the quality papers we deserve.
Funding Source: No specific funding was received for this systematic review. The authors were funded by National Institute for Health and Care Research (UK) clinical academic training grants.
Lead Author: Charlotte Talbot is affiliated with University of Birmingham, United Kingdom, and works at Queen Elizabeth Hospital Birmingham. Talbot was funded by a National Institute for Health and Care Research clinical academic training grant during this research. No personal connection to donor conception was indicated.
Regulatory Context: Varies globally
Related Posts
So glad to have you 'translating' these academic papers for us. Thank you!