Table 1.
Overview of traits for which polygenic scores were constructed.
| Trait | N | Overlap/exclusions | Rationale | GWAS Reference |
|---|---|---|---|---|
| Cerebral Palsy (FinnGen phenotype G6_CP: ICD-10 code G80 and ICD-8 or 9 code 343) | 3465 | Directly tests polygenic contribution to CP. CP inclusion criteria is similar to MyCode. | Kurki et al.17,a | |
| Age at walking onset | 70,560 | Captures motor development, a core domain affected in CP | Gui et al.29 | |
| Birth Weight | 298,142 | A person with CP is seven times more likely to be born small for their gestational age than the population average | Warrington et al.26 | |
| Gestational Duration | 151,987 | Excluding 23andMe, Inc., data | A person with CP is four times more likely to have had a preterm birth than the population average | Solé-Navais et al.24 |
| Stroke | 278,028 | For Australian polygenic scores | At least 30% of neonates affected by perinatal stroke will develop CP | Mishra et al.22 |
| 147,590 | For MyCode polygenic scores (used because MyCode data are included in Mishra et al.22) | Malik et al.30 | ||
| Autism spectrum disorders | 44,367 | Comorbidity with 8%–30% of people living with CP | Grove et al.21 | |
| Epilepsy | 65,124 | Comorbidity with approximately 30% of people living with CP | Stevelink et al.25 | |
| Educational Attainment | 757,277 | Excluded QIMR data (Australian polygenic scores) | Proxy for cognitive function/intellectual disability, which affects around 50% of people with CP | Okbay et al.23 |
| 750,721 | Excluded MyCode data (MyCode polygenic scores) | |||
| All excluded 23andMe, Inc., data |
For each trait, details of the genome-wide association study (GWAS) used are displayed, including sample size (N), reported as effective sample size for binary traits and as a total sample size for quantitative traits. Effective sample size was calculated using 4/(1/[number of cases] + 1/[number of controls]). The overlap/exclusions column describes steps taken to avoid sample overlap between the GWAS discovery data and the Australian and MyCode cohorts, thereby preventing bias and ensuring independent predictions. Data from 23andMe, Inc. were excluded due to lack of approval for use and this is also indicated where applicable in the Overlap/exclusions column. The Rationale column provides information for why each trait was included, and the GWAS Reference column cites the corresponding publication for the GWAS used. For all traits, the GWAS summary statistics from analysis including Europeans only were used.
CP GWAS summary statistics provided by FinnGen; FinnGen release 12 meta-analysed with Europeans from the UK Biobank (GWAS meta-analysis results: https://public-metaresults-fg-ukbb.finngen.fi/pheno/G6_CP and methods: https://finngen.gitbook.io/documentation/methods/meta-analysis).