Cross-sectional study compares self-reported mental health outcomes between donor sperm-conceived and spontaneously conceived adults
Self-reported mental health status of donor sperm-conceived adults. (Adams, 2022)
Adams, D. H., Gerace, A., Davies, M. J., & de Lacey, S. (2022). Self-reported mental health status of donor sperm-conceived adults. Journal of Developmental Origins of Health and Disease, 13, 220–230. https://doi.org/10.1017/S2040174421000210
Geographic Region: Global with most respondents coming from Australia, the Netherlands, the United States, the United Kingdom, and Belgium
Research Question: Do donor sperm-conceived adults self-report different mental health outcomes (e.g., formal diagnoses, personal experience measures, and standardized depression/anxiety/stress scores) compared to spontaneously conceived adults?
Design: Cross-sectional online survey fielded between December 2017 and March 2018. The survey covered demographic characteristics, birth and gestational history, general health and lifestyle, diagnosed mental health conditions, self-reported mental health experiences, and the Depression Anxiety Stress Scale-21 (DASS-21), which is a validated instrument measuring current levels of depression, anxiety, and stress over the prior week.
Sample: 1,149 adults (272 donor sperm-conceived and 877 spontaneously conceived). The mean age was approximately 32–33 years in both groups. Over 80% in each group identified as female, and education levels were higher in the donor-conceived group (32% holding postgraduate degrees). Donor-conceived participants were recruited primarily through donor conception Facebook groups and organizations, supplemented by a survey recruitment platform and snowball sampling. Spontaneously conceived participants were recruited through Flinders University advertising, a survey recruitment platform, and snowball sampling. The two groups were matched on mean age, height, sex ratio, alcohol consumption, current smoking status, exercise levels, maternal smoking during pregnancy, and whether participants had received fertility treatment themselves. The study does not report race or ethnicity, socioeconomic status beyond education level, or the age at which participants learned of their donor conception. Donor type (known, anonymous, identity-release), family structure, and disclosure circumstances are also not captured.
Editor’s Note: It's worth being precise about what this study is positioned to detect. The recruitment strategy of drawing donor-conceived participants almost entirely from support communities means the study is structurally more likely to surface distress, regardless of the broader population's actual experience. As such, the study uncovers self-reported mental health patterns in a community-engaged, self-selected subset of donor-conceived adults rather than the mental health profile of donor-conceived adults as a population.
Regarding methods, when you run dozens of statistical tests at once, some will look like meaningful findings just by chance. The authors applied the Benjamini–Hochberg procedure to adjust the results to account for this. Several results that initially cleared the threshold for statistical significance didn’t survive the correction. “Significant” is a statistical term that means the result was unlikely enough to be a fluke that we treat it as real, but a statistically significant difference can still be small, and a large difference can fail to reach significance if the sample is too small to detect it reliably.
In the discussion section, the authors used the Developmental Origins of Health and Disease research framework as a possible explanation for their findings. The framework proposes that early biological exposures (nutrition, infection, stress hormones, perinatal conditions) shape long-term health outcomes, often through epigenetic or developmental mechanisms. This is one plausible explanatory pathway, but it exists alongside other plausible explanations that the study cannot rule out: the psychological effects of secrecy, late or traumatic disclosure/discovery, the experience of searching for a donor or encountering donor siblings, or psychological weight of confronting how the fertility industry itself operated (e.g., discovering that one's conception involved anonymity by design, limited or falsified donor information, profit-driven recruitment practices).
This sample's age distribution is not fully reported (only mean and standard deviation), so the range of disclosure eras represented cannot be determined. Given the wide standard deviations (10.3 and 12.5 years), respondents likely span several decades of birth years and correspondingly different disclosure and donor-conception regulatory contexts. The mean age was approximately 32–33 years in both groups, corresponding to a mean birth year of roughly 1984–1985, but this should not be read as describing a narrow or typical disclosure-era cohort.
Key Findings
The largest difference observed was in identity formation difficulty: 52% of donor sperm-conceived adults reported it vs. 14% of spontaneously conceived adults. This is also the only finding that remained statistically significant when the analysis was restricted to Australian respondents only (54% vs. 14%). The authors did not define what “identity formation difficulty” means, used a single yes/no survey item, and did not distinguish between different dimensions of identity (genealogical, psychological, relational, narrative).
Donor sperm-conceived adults were more likely to self-report a diagnosis of attention deficit disorder or ADHD (10% vs. 4%) and an autism or autism spectrum disorder diagnosis (5% vs. 2%). These findings remained statistically significant after correction.
Donor sperm-conceived adults reported a higher rate of diagnoses falling outside the listed categories (13% vs. 7%); the most common free-text conditions named were borderline personality disorder, OCD, and PTSD, though none of these individual categories reached statistical significance after correction.
Donor sperm-conceived adults were more likely to report a depression diagnosis (40% vs. 31%), but this difference was not significant after the authors did the statistical correction.
Donor sperm-conceived adults were more likely to report having had panic attacks (54% vs. 43%), learning difficulties (17% vs. 7%), alcohol or drug dependency (12% vs. 6%), and recurrent nightmares (26% vs. 18%). These findings maintained statistical significance after correction.
Donor sperm-conceived adults were more likely to have seen a mental health professional at all (70% vs. 50%), including psychologists (47% vs. 33%) and psychiatrists (22% vs. 16%). Whether this reflects greater distress, greater help-seeking, or both cannot be determined from these data. These findings maintained statistical significance after correction.
On the DASS-21, donor sperm-conceived adults scored significantly higher on the stress subscale over the past week (mean 13.43 vs. 11.65), but both of these scores fall within the "normal" range on the DASS-21 scoring system. Depression and anxiety subscale scores did not differ significantly between groups. The overall DASS-21 total score difference was also not statistically significant.
The subgroup analyses were underpowered, and no firm conclusions can be drawn from them.
Limitations: The authors stated that recruitment through donor conception Facebook groups and advocacy organizations meant donor-conceived participants were self-selected from among those already engaged with and potentially distressed about their conception. Those who do know their donor conception status and seek out community support are not representative of the broader population. The heavy female skew (over 80% in each group) limits generalizability to men. The study captures no information about disclosure timing or other characteristics. Given the study’s own observation that early disclosure is associated with “better outcomes”, this omission is significant. No race or ethnicity data are reported, and most donor conception research to date is drawn from samples that are disproportionately white and educated. Family structure is not captured, so subpopulations have meaningfully different experiences that are collapsed into a single group. Donor type is unknown, so anonymous, identity-release, and known donor conception are pooled. The study makes no distinction between donor-conceived adults who are actively searching for or in contact with their donor and those who are not. The absence of age-bracket data is itself a limitation that prevents distinguishing whether outcomes are driven by a specific historical secrecy regime or are more broadly distributed across donor-conception practice over time.
Applications: Donor-conceived adults who recognize themselves in these findings, particularly regarding identity difficulties, stress, or having sought mental health support, should know they are not statistically unusual in this study's sample. At the same time, this is not a population-representative study, so it shouldn't be read as "this is what donor conception does to people." Future studies should treat disclosure timing and quality, family structure, donor type, and contact status as primary variables and use theoretically grounded, validated, multi-item measures for identity. Recruitment strategies that reach beyond advocacy and support communities would meaningfully improve the generalizability of findings in this area.
Funding Source: Australian Government Research Training Program Scholarship
Lead Author: Damian Adams is a post-doctoral researcher in the College of Nursing and Health Sciences at Flinders University, South Australia, whose work focuses on physical and mental health outcomes and welfare rights of donor-conceived people, grounded in the Developmental Origins of Health and Disease framework. Adams is donor-conceived, has publicly identified as an advocate for the rights of donor-conceived people, has lobbied for legislative reforms in South Australia and Victoria, has provided expert testimony to Australian state and federal government inquiries and to the United Nations, and has published opinion and advocacy pieces alongside his peer-reviewed research.
Regulatory Context: The study is international in scope, drawing participants primarily from Australia, the Netherlands, the United States, the United Kingdom, and Belgium. Regulatory frameworks differ substantially across these jurisdictions.
Related Posts