Table 9.
Differences in the probability (percentage points) of social and mental-health outcomes at ages 16–18 among girls conceived by medically assisted reproduction (MAR) versus those conceived naturally. Between-family and within-family analyses of girls born in Finland 1995−2000
| MAR | Between− family sample | Within− family sample | ||
|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 4 | |
| Grade point average (GPA) a | 23.93*** (1.434) | 4.132*** (1.309) | − 2.387 (5.009) | − 8.403 (5.399) |
| Academic track (ref. vocational) | 10.46*** (0.624) | 1.010* (0.575) | − 0.337 (1.978) | − 2.320 (2.137) |
| School dropout | − 0.276 (0.213) | 0.384* (0.213) | 1.317* (0.689) | 1.425* (0.746) |
| NEET status at age 18 | − 0.857*** (0.232) | 0.223 (0.231) | 0.339 (0.741) | 0.149 (0.802) |
| Early home-leaving | − 7.586*** (0.464) | − 3.001*** (0.451) | − 0.469 (1.396) | − 0.972 (1.511) |
| Antidepressant use | − 0.0670 (0.328) | 0.607* (0.329) | − 0.614 (1.268) | − 1.393 (1.371) |
| Mental disorders (any) | − 0.187 (0.361) | 0.829** (0.361) | 1.457 (1.338) | − 0.004 (1.447) |
| Internalizing | − 0.0112 (0.293) | 0.588** (0.294) | 1.348 (1.102) | 0.114 (1.192) |
| Externalizing | − 0.147 (0.140) | 0.260* (0.140) | 0.631 (0.501) | 0.223 (0.542) |
| Developmental | 0.177 (0.126) | 0.285** (0.127) | − 0.976* (0.512) | − 1.230** (0.554) |
| Other mental disorders | 0.232 (0.242) | 0.579** (0.244) | 1.219 (0.971) | 0.843 (1.049) |
| High-risk health behaviour | − 0.257 (0.168) | 0.196 (0.169) | 0.0108 (0.648) | − 0.323 (0.703) |
| Adjustments: | ||||
| Child's sex | x | x | x | x |
| Birth year | x | x | ||
| Birth order | x | x | ||
| Maternal age at birth | x | |||
| Maternal smoking | x | x | ||
| Parental education at birth | x | |||
| Family structure at birth | x | |||
| Income decile at birth | x | x | ||
| Hospital district at birth | x | |||
Standard Errors in parentheses. Sample sizes vary across outcomes but are held constant for each outcome across estimations for each type of analysis. Sample sizes of each outcome are the same as reported in Table 3. Sibling fixed effects used for the within-family models. Significance levels: ***p < 0.001, **p < 0.01, *p < 0.05
aThe coefficient (multiplied by 100) for GPA represents a linear estimation, not a probability