Appendix Appendix Figures Appendix Figure 1: Correlations between maternal smoking during pregnancy and outcomes in offspring and parts of the correlations explained by different sources of variance Note: These are the results from the full models, without model fitting steps performed. Due to computational issues the model for the association between SDP and preterm birth, as well as with being born small for gestational age, was fitted in three steps; first SDP and outcomes were fitted separately, then the results from model fitting were used in the cross-phenotype analyses. Appendix Figure 2: Path diagram representations for each variance source separately. A covariance C covariance M covariance P covariance E covariance Note: gm, genetic relation of mothers. go, genetic relation of siblings in offspring generation. Sub-index 11 refers to offspring 1 in nuclear family 1, sub-index 21 refers to offspring 1 in nuclear family 2, and subindex 22 refers to offspring 2 in nuclear family 2. Appendix Tables Appendix Table 1: ACMPE univariate modelled parameters [parameter estimate (standard error)] Maternal phenotype Maternal smoking during pregnancya Offspring phenotype Birth weight Preterm birth Born small for gestational age Low academic achievement General cognitive ability Criminal convictions Violent criminal convictions Drug/alcohol misuse ππ ππ ππ ππ ππ .831 (.007) .171 (.021) .197 (.010) .396 (.004) ππ ππ ππ ππ ππ .392 (.005) .221 (.003) .147 (.002) .000 (.014) .288 (.003) .517 (.039) .492 (.015) .276 (.014) .001 (.078) .643 .658 (.035) .398 (.022) .270 (.015) .002 (.068) .579 .927 (.023) .107 (.041) .349 (.010) .057 (.046) .067 1.058 (.028) .816 (.052) .000 (.114) .218 (.197) 1.258 (.014) .627 (.023) .140 (.051) .208 (.027) .000 (.117) .737 .793 (.050) .129 (.139) .254 (.055) .001 (.390) .539 .290 .591 (.023) .241 (.020) .159 (.023) .110 (.014) .745 Note: The e-parameters for binary phenotypes are defined as remaining variance after estimation of other parameters and are therefore void of standard errors. a Calculated using the full 1983-2009 cohort. Appendix Table 2: ACMPE full model results [parameter estimate (standard error)] Within SDP Within outcome Between phenotypes ππ ππ ππ ππ ππ ππ ππ ππ ππ ππ ππ¨ ππͺ ππ΄ ππ· ππ¬ 0.823 (0.003) 0.828 (0.008) 0.828 (0.008) 0.262 (0.003) 0.183 (0.015) 0.183 (0.015) 0.213 (0.003) 0.202 (0.017) 0.202 (0.017) 0.354 (0.003) 0.394 (0.003) 0.394 (0.003) 0.288 0.394 (0.003) 0.514 (0.048) 0.660 (0.015) 0.207 (0.003) 0.494 (0.019) 0.396 (0.015) 0.147 (0.001) 0.277 (0.015) 0.270 (0.011) 0.035 (0.003) 0.002 (0.075) 0.002 (0.075) 0.300 (0.003) 0.644 -0.046 (0.003) NA -0.773 (0.001) NA -0.679 (0.003) NA -1.000 (0.005) NA -0.244 (0.001) NA 0.578 NA NA NA NA NA Low academic achievement General cognitive ability 0.810 (0.000) 0.344 (0.000) 0.267 (0.000) 0.296 (0.000) 0.258 0.893 (0.000) 0.154 (0.000) 0.366 (0.000) 0.105 (0.000) 0.185 0.598 (0.001) 0.024 (0.006) 0.623 (0.001) 0.494 (0.009) -0.018 (0.001) 0.813 (0.022) 0.263 (0.020) 0.307 (0.019) 0.313 (0.016) 0.278 1.076 (0.033) 0.819 (0.023) 0.097 (0.019) 0.212 (0.012) 1.244 (0.016) -0.536 (0.010) -0.071 (0.014) -0.327 (0.008) -0.980 (0.010) -0.030 (0.010) Criminality 0.802 (0.029) 0.782 (0.017) 0.835 (0.001) 0.351 (0.030) 0.370 (0.020) 0.293 (0.001) 0.290 (0.038) 0.314 (0.019) 0.260 (0.001) 0.291 (0.011) 0.297 (0.012) 0.296 (0.001) 0.254 0.573 (0.024) 0.795 (0.061) 0.600 (0.001) 0.203 (0.024) 0.000 (0.155) 0.256 (0.001) 0.222 (0.018) 0.272 (0.042) 0.146 (0.000) 0.076 (0.030) 0.057 (0.073) 0.088 (0.000) 0.759 1.000 (0.099) 0.654 (0.228) 0.591 (0.001) -0.693 (0.090) 0.795 (0.147) 0.190 (0.000) 0.009 (0.120) 0.421 (0.481) 0.743 (0.001) 0.511 (0.498) 0.522 (0.816) 0.182 (0.000) 0.000 (0.016) 0.024 (0.039) 0.000 (0.000) Birth weight Preterm birtha Born small for gestational agea Violent criminality Drug misuse 0.290 0.290 0.255 0.250 0.538 0.738 Abbreviations: SDP, maternal smoking during pregnancy. NA, not applicable. Note: Standard errors may be misleading due to constraints put on the optimizer when fitting the model. The e-parameters for binary phenotypes are defined as remaining variance after estimation of other parameters and are therefore void of standard errors. a Due to computational issues SDP and outcome were fitted separately. Appendix Table 3: ACMPE best-fitting model results [parameter estimate (standard error)] Within SDP Within outcome Between phenotypes ππ ππ ππ ππ ππ ππ ππ ππ ππ ππ ππ¨ ππͺ ππ΄ ππ· ππ¬ 0.823 (0.003) 0.828 (0.008) 0.828 (0.008) 0.262 (0.003) 0.183 (0.015) 0.183 (0.015) 0.213 (0.003) 0.202 (0.017) 0.202 (0.017) 0.354 (0.003) 0.394 (0.003) 0.394 (0.003) 0.288 0.394 (0.003) 0.514 (1.314) 0.660 (0.006) 0.207 (0.003) 0.494 (0.003) 0.396 (0.004) 0.147 (0.001) 0.277 (0.002) 0.270 (0.005) 0.035 (0.003) 0 0.300 (0.003) 0.644 0.578 -0.773 (0.001) 0.212 (0.170) 1.000 (1.414) -0.679 (0.003) 0.385 (0.248) -0.138 (3.247) -1.000 (0.005) 0 0 -0.046 (0.003) 0.045 (0.131) 0.536 (1.321) -0.244 (0.001) 0.048 (0.012) 0.172 (0.706) Low academic achievementb General cognitive ability 0.810 0.344 0.267 0.296 0.258 0.893 0.154 0.366 0.105 0.185 0.598 0 0.623 0.494 -0.018 0.820 (0.047) 0.283 (0.071) 0.295 (0.070) 0.289 (0.027) 0.278 1.114 (0.026) 0.825 (0.020) 0 0 1.228 (0.014) -0.555 (0.048) -0.301 (0.093) 0 0 -0.034 (0.020) Criminality 0.808 (0.006) 0.802 (0.028) 0.850 (0.005) 0.344 (0.006) 0.349 (0.030) 0.290 (0.005) 0.282 (0.006) 0.286 (0.036) 0.229 (0.005) 0.293 (0.006) 0.297 (0.010) 0.281 (0.005) 0.254 0.585 (0.006) 0.899 (0.018) 0.730 (0.005) 0.192 (0.006) 0 0.227 (0.020) 0 0 0.755 0 0.438 -0.687 (0.017) 0 0 0 0 0 0 0 0 0.683 1.000 (0.002) 0.691 (0.027) 0.635 (0.005) 0 0 0 -0.001 (0.013) 0.026 (0.019) -0.002 (0.005) Birth weight Preterm birtha Born small for gestational agea Violent criminality Drug misuse 0.290 0.290 0.255 0.249 0 Abbreviations: SDP, maternal smoking during pregnancy. Note: Standard errors may be misleading due to constraints put on the optimizer when fitting the model. The e-parameters for binary phenotypes are defined as remaining variance after estimation of other parameters and are therefore void of standard errors. All values significant, on a 5%level, except when a “0” is stated, where they are non-significant in a likelihood ratio test and have been removed from model. e-parameters are not subject to significance tests. ππΈ is included in final model regardless of significance. a Due to computational issues the model was fitted in three steps; first SDP and outcome were fitted separately, then the results from model fitting were used in the cross-phenotype analyses. b No valid standard errors could be obtained from the optimization. 8 Appendix Table 4a: Likelihood ratio tests for smoking during pregnancy and birth weight Model name Number of estimated parameters -2 log likelihood ACMPE – ACMPE ACME – ACME ACPE – ACPE AMPE – AMPE CMPE – CMPE ACMPE – CMPE ACMPE – ACPE ACMPE – ACME ACMPE – AMPE ACMPE – ACMP 40 38 38 38 38 39 39 39 39 39 4108980.3 4109054.1 4109917.0 4109821.6 4109241.9 4108984.5 4108998.2 4109054.3 4109070.6 4111042.7 P-value; test again previous model with more estimated parameters NA <0.001 <0.001 <0.001 <0.001 0.039 <0.001 <0.001 <0.001 <0.001 Abbreviations: NA, not applicable. Note: “Best-fitting model” marked in bold letters. Each model is tested against previous model with more estimated parameters. The model name implies which parameters that are included, it is stated as ACMPE – ACMPE, where the first five letters tells us which of the variance parameters in birth weight that are included (i.e., ππ2 , ππ2 , ππ2 , ππ2 , and ππ2 ), and the five last letters tells us which of the cross-phenotype parameters that are included (i.e., ππ΄ , ππΆ , ππ , ππ , and ππΈ ). All of the variance parameters in maternal smoking during pregnancy are included in all models. Appendix Table 4b: Likelihood ratio tests for preterm birth Model name Number of estimated parameters -2 log likelihood ACMPE ACME ACE AME CME 17 16 15 15 15 714467.6 714467.6 714554.3 714644.1 714501.1 P-value; test again previous model with more estimated parameters NA 1 <0.001 <0.001 <0.001 Abbreviations: NA, not applicable. Note: “Best-fitting model” marked in bold letters. Due to computational issues only the outcome phenotype was tested for significance. Each model is tested against previous model with more estimated parameters. The model name implies which parameters that are included, it is stated as ACMPE, where the letters tells us which of the variance parameters in preterm birth that are included (i.e., ππ2 , ππ2 , ππ2 , ππ2 , and ππ2 ). Appendix Table 4c: Likelihood ratio tests for being born small for gestational age Model name Number of estimated parameters -2 log likelihood P-value; test again previous model with more estimated 9 ACMPE ACME ACE AME CME 17 16 15 15 15 808894.7 808894.7 808986.0 808975.9 808994.6 parameters NA 1 <0.001 <0.001 <0.001 Abbreviations: NA, not applicable. Note: “Best-fitting model” marked in bold letters. Due to computational issues only the outcome phenotype was tested for significance. Each model is tested against previous model with more estimated parameters. The model name implies which parameters that are included, it is stated as ACMPE, where the letters tells us which of the variance parameters in being born small for gestational age that are included (i.e., ππ2 , ππ2 , ππ2 , ππ2 , and ππ2 ). Appendix Table 4d: Likelihood ratio tests for smoking during pregnancy and low academic achievement Model name Number of estimated parameters -2 log likelihood ACMPE – ACMPE ACME – ACME AMPE – AMPE ACPE – ACPE CMPE – CMPE ACMPE – AMPE ACMPE – AME ACMPE – MPE ACMPE – APE ACMPE – AMP 39 37 37 37 37 38 37 37 37 37 1298228.9 1298279.4 1298270.4 1298563.5 1298296.0 1298229.9 1298283.1 1298425.3 1298295.3 1298229.9 P-value; test again previous model with more estimated parameters NA <0.001 <0.001 <0.001 <0.001 0.900 <0.001 <0.001 <0.001 1 Abbreviations: NA, not applicable. Note: “Best-fitting model” marked in bold letters. Each model is tested against previous model with more estimated parameters. The model name implies which parameters that are included, it is stated as ACMPE – ACMPE, where the first five letters tells us which of the variance parameters in low academic achievement that are included (i.e., ππ2 , ππ2 , ππ2 , ππ2 , and ππ2 ), and the five last letters tells us which of the cross-phenotype parameters that are included (i.e., ππ΄ , ππΆ , ππ , ππ , and ππΈ ). All of the variance parameters in maternal smoking during pregnancy are included in all models. Appendix Table 4e: Likelihood ratio tests for smoking during pregnancy and general cognitive functioning Model name Number of estimated parameters -2 log likelihood ACMPE – ACMPE 38 1154273.6 P-value; test again previous model with more estimated parameters NA 10 ACPE – ACPE ACE – ACE AE – AE CE – CE ACE – AE ACE – CE ACE – AC 36 34 32 32 33 33 33 1154275.8 1154278.3 1154790.9 1154927.5 1154339.4 1154626.1 1154321.9 0.327 0.289 <0.001 <0.001 <0.001 <0.001 <0.001 Abbreviations: NA, not applicable. Note: “Best-fitting model” marked in bold letters. Each model is tested against previous model with more estimated parameters. The model name implies which parameters that are included, it is stated as ACMPE – ACMPE, where the first five letters tells us which of the variance parameters in general cognitive functioning that are included (i.e., ππ2 , ππ2 , ππ2 , ππ2 , and ππ2 ), and the five last letters tells us which of the cross-phenotype parameters that are included (i.e., ππ΄ , ππΆ , ππ , ππ , and ππΈ ). All of the variance parameters in maternal smoking during pregnancy are included in all models. Appendix Table 4f: Likelihood ratio tests for smoking during pregnancy and criminal convictions Model name Number of estimated parameters -2 log likelihood ACMPE – ACMPE ACME - ACME AME – AME ACE – ACE CME – CME ACME – ACE ACME – AE ACME – CE ACME – AC 39 37 35 35 35 36 35 35 35 847607.1 847607.9 847877.4 847675.9 847640.0 847607.9 847619.3 847976.4 847607.9 P-value; test again previous model with more estimated parameters NA 0.692 <0.001 <0.001 <0.001 1 <0.001 <0.001 1 Abbreviations: NA, not applicable. Note: “Best-fitting model” marked in bold letters. Each model is tested against previous model with more estimated parameters. The model name implies which parameters that are included, it is stated as ACMPE – ACMPE, where the first five letters tells us which of the variance parameters in criminal convictions that are included (i.e., ππ2 , ππ2 , ππ2 , ππ2 , and ππ2 ), and the five last letters tells us which of the cross-phenotype parameters that are included (i.e., ππ΄ , ππΆ , ππ , ππ , and ππΈ ). All of the variance parameters in maternal smoking during pregnancy are included in all models. Appendix Table 4g: Likelihood ratio tests for smoking during pregnancy and violent criminal convictions Model name Number of estimated parameters -2 log likelihood ACMPE – ACMPE AMPE – AMPE AME – AME AE – AE E –E 39 37 35 33 31 624094.2 624094.2 624098.1 624101.1 625980.3 P-value; test again previous model with more estimated parameters NA 1 0.142 0.224 <0.001 11 AE –E AE – A 32 32 625594.2 624101.1 <0.001 0.975 Abbreviations: NA, not applicable. Note: “Best-fitting model” marked in bold letters. Each model is tested against previous model with more estimated parameters. The model name implies which parameters that are included, it is stated as ACMPE – ACMPE, where the first five letters tells us which of the variance parameters in violent criminal convictions that are included (i.e., ππ2 , ππ2 , ππ2 , ππ2 , and ππ2 ), and the five last letters tells us which of the cross-phenotype parameters that are included (i.e., ππ΄ , ππΆ , ππ , ππ , and ππΈ ). All of the variance parameters in maternal smoking during pregnancy are included in all models. Appendix Table 4h: Likelihood ratio tests for smoking during pregnancy and drug/alcohol misuse Model name Number of estimated parameters -2 log likelihood ACMPE – ACMPE ACME – ACME ACE – ACE AME – AME AE - AE AE – E AE – A 39 37 35 35 33 32 32 573109.2 573109.2 573115.3 573111.9 573112.3 574303.7 573112.3 P-value; test again previous model with more estimated parameters NA 1 0.049 0.262 0.827 <0.001 0.803 Abbreviations: NA, not applicable. Note: “Best-fitting model” marked in bold letters. Each model is tested against previous model with more estimated parameters. The model name implies which parameters that are included, it is stated as ACMPE – ACMPE, where the first five letters tells us which of the variance parameters in drug/alcohol misuse that are included (i.e., ππ2 , ππ2 , ππ2 , ππ2 , and ππ2 ), and the five last letters tells us which of the crossphenotype parameters that are included (i.e., ππ΄ , ππΆ , ππ , ππ , and ππΈ ). All of the variance parameters in maternal smoking during pregnancy are included in all models. 12 Appendix Table 5. Covariance between random effects, within and across phenotypes, for different relations as stated in terms of offspring Within subject A π¨πΊπ«π· πππ ππ 2 π¨πΊπ«π· πππ π¨πΆπΌπ» πππ 1 ππ π 2 π π π΄ Full sibs Half sibs Mothers MZ twins Mothers DZ twins Mothers full sibs Mothers maternal half sibs Mothers paternal half sibs π¨πΆπΌπ» πππ π¨πΊπ«π· πππ′ π¨πΆπΌπ» πππ′ π¨πΊπ«π· πππ′ π¨πΆπΌπ» πππ′ π¨πΊπ«π· ππ′⋅ π¨πΆπΌπ» ππ′⋅ π¨πΊπ«π· ππ′⋅ π¨πΆπΌπ» ππ′⋅ π¨πΊπ«π· ππ′⋅ π¨πΆπΌπ» ππ′⋅ π¨πΊπ«π· ππ′⋅ π¨πΆπΌπ» ππ′⋅ π¨πΊπ«π· ππ′ ⋅ 1 ππ π 2 π π π΄ ππ2 ππ 2 ππ 2 1 ππ π 8 π π π΄ 1 ππ π 8 π π π΄ 1 2 π 16 π 1 2 π 4 π 1 ππ π 4 π π π΄ 1 ππ π 4 π π π΄ 1 2 π 8 π 1 2 π 4 π 1 ππ π 4 π π π΄ 1 ππ π 4 π π π΄ 1 2 π 8 π 1 2 π 2 π 1 ππ π 2 π π π΄ 1 ππ π 2 π π π΄ 1 2 π 4 π 1 2 π 2 π 1 ππ π 2 π π π΄ 1 ππ π 2 π π π΄ 1 2 π 4 π ππ 2 1 ππ π 2 π π π΄ 1 ππ π 2 π π π΄ 1 2 π 2 π πͺπΊπ«π· ππ′ ⋅ πͺπΆπΌπ» ππ′⋅ πͺπΊπ«π· ππ′ ⋅ πͺπΆπΌπ» ππ′⋅ πͺπΊπ«π· ππ′ ⋅ πͺπΆπΌπ» ππ′⋅ πͺπΊπ«π· ππ′ ⋅ πͺπΆπΌπ» ππ′⋅ π¨πΆπΌπ» ππ′⋅ 1 ππ π 8 π π π΄ 1 1 2 ππ π π 8 π π π΄ 16 π πͺπΊπ«π· πππ πͺπΆπΌπ» πππ πͺπΊπ«π· πππ′ πͺπΆπΌπ» πππ′ πͺπΊπ«π· πππ′ πͺπΆπΌπ» πππ′ πͺπΊπ«π· πππ ππ 2 ππ ππ ππΆ ππ 2 ππ ππ ππΆ ππ 2 ππ ππ ππΆ 0 0 0 0 0 0 0 0 0 0 πͺπΆπΌπ» πππ ππ ππ ππΆ ππ2 ππ ππ ππΆ ππ2 ππ ππ ππΆ ππ2 0 0 0 0 0 0 0 0 0 0 M π΄πΊπ«π· πππ π΄πΆπΌπ» πππ π΄πΊπ«π· πππ′ π΄πΆπΌπ» πππ′ π΄πΊπ«π· πππ′ π΄πΆπΌπ» πππ′ π΄πΊπ«π· ππ′ ⋅ π΄πΆπΌπ» ππ′⋅ π΄πΊπ«π· ππ′ ⋅ π΄πΆπΌπ» ππ′⋅ π΄πΊπ«π· ππ′ ⋅ π΄πΆπΌπ» ππ′⋅ π΄πΊπ«π· ππ′ ⋅ π΄πΊπ«π· ππ′⋅ π΄πΆπΌπ» ππ′⋅ ππ 2 ππ ππ ππ ππ 2 ππ ππ ππ ππ 2 ππ ππ ππ ππ 2 ππ ππ ππ ππ 2 ππ ππ ππ ππ 2 ππ ππ ππ ππ 2 ππ2 ππ ππ ππ ππ2 ππ ππ ππ ππ2 ππ ππ ππ ππ2 ππ ππ ππ ππ2 ππ ππ ππ ππ2 ππ ππ ππ ππ2 π·πΊπ«π· πππ π·πΆπΌπ» πππ π·πΊπ«π· πππ′ π·πΆπΌπ» πππ′ π·πΊπ«π· πππ′ π·πΆπΌπ» πππ′ π·πΊπ«π· ππ′⋅ π·πΆπΌπ» ππ′ ⋅ π·πΊπ«π· ππ′⋅ π·πΆπΌπ» ππ′ ⋅ π·πΊπ«π· ππ′⋅ π·πΆπΌπ» ππ′ ⋅ π·πΊπ«π· ππ′⋅ π·πΆπΌπ» ππ′ ⋅ ππ 2 ππ ππ ππ ππ 2 ππ ππ ππ 0 0 0 0 0 0 0 0 0 0 0 0 ππ2 ππ ππ ππ ππ2 0 0 0 0 0 0 0 0 0 0 0 0 π¬πΊπ«π· πππ π¬πΆπΌπ» πππ π¬πΊπ«π· πππ′ π¬πΆπΌπ» πππ′ ππ 2 ππ ππ ππΈ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ππ ππ ππΈ ππ2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 C π΄πΊπ«π· πππ ππ ππ ππ π΄πΆπΌπ» πππ P π·πΊπ«π· πππ ππ ππ ππ π·πΆπΌπ» πππ E π¬πΊπ«π· πππ π¬πΆπΌπ» πππ π¬πΊπ«π· πππ′ π¬πΆπΌπ» πππ′ π¬πΊπ«π· ππ′ ⋅ π¬πΆπΌπ» ππ′ ⋅ π¬πΊπ«π· ππ′ ⋅ π¬πΆπΌπ» ππ′ ⋅ π¬πΊπ«π· ππ′ ⋅ π¬πΆπΌπ» ππ′ ⋅ π¬πΊπ«π· ππ′ ⋅ π΄πΆπΌπ» ππ′⋅ πͺπΊπ«π· ππ′⋅ ππ ππ ππ 0 π¬πΆπΌπ» ππ′ ⋅ 0 π·πΊπ«π· ππ′ ⋅ π¬πΊπ«π· ππ′⋅ πͺπΆπΌπ» ππ′⋅ 0 0 π·πΆπΌπ» ππ′⋅ π¬πΆπΌπ» ππ′⋅ Abbreviations: SDP, smoking during pregnancy – the exposure. OUT, outcome. A, additive genetic effects. C, shared environment between siblings in offspring generation. M, shared environment between siblings in parental generation. P, paternal/spouse effect. E, unique environment in offspring generation. The π ′ and π ′ subscript are the compliments of π and π, respectively. The “⋅” is used to indicate any of possible values at subscript. 13 Appendix Table 6a: Observed means of birth weight in different exposure-combination groups Birth weight SDP status First pregnancy Any value 0 1 Any value Any value 0 1 0 1 Second pregnancy Any value Any value Any value 0 1 0 0 1 1 Mean birth weight in grams and number of individuals First Second N pregnancy pregnancy 3461 3625 1,360,988 3486 3651 1,129,312 3344 3498 231,674 3482 3654 1,173,284 3334 3443 187,702 3487 3654 1,098,770 3411 3643 74,514 3446 3516 30,542 3312 3429 157,160 Appendix Table 6b: Observed prevalences of pre-term birth in different exposure-combination groups Pre-term birth SDP status First pregnancy Any value 0 1 Any value Any value 0 1 0 1 Second pregnancy Any value Any value Any value 0 1 0 0 1 1 Prevalences and number of individuals First Second N pregnancy pregnancy 5.63% 3.61% 1,360,988 5.53% 3.41% 1,129,312 6.14% 4.62% 231,674 5.54% 3.39% 1,173,284 6.23% 4.99% 187,702 5.52% 3.38% 1,098,770 5.80% 3.56% 74,514 5.88% 4.33% 30,542 6.30% 5.12% 157,160 Appendix Table 6c: Observed prevalences of being born small for gestational age in different exposure-combination groups Born small for gestational age SDP status Prevalences and number of individuals 14 First pregnancy Any value 0 1 Any value Any value 0 1 0 1 Second pregnancy Any value Any value Any value 0 1 0 0 1 1 First pregnancy 7.95% 6.87% 13.22% 7.07% 13.46% 6.82% 10.85% 8.91% 14.35% Second pregnancy 4.40% 3.62% 8.18% 3.60% 9.37% 3.52% 4.75% 7.16% 9.80% N 1,360,988 1,129,312 231,674 1,173,284 187,702 1,098,770 74,514 30,542 157,160 Appendix Table 6d: Observed prevalences of low academic achievement in different exposurecombination groups Low academic achievement SDP status First pregnancy Any value 0 1 Any value Any value 0 1 0 1 Second pregnancy Any value Any value Any value 0 1 0 0 1 1 Prevalences and number of individuals First Second N pregnancy pregnancy 7.80% 9.30% 453,806 5.73% 6.87% 341,492 14.08% 16.66% 112,314 5.76% 6.93% 356,552 15.26% 17.97% 97,254 5.44% 6.51% 327,628 9.38% 11.68% 28,924 12.58% 15.41% 13,864 15.71% 18.39% 83,390 Appendix Table 6e: Observed means of general cognitive ability in different exposure-combination groups General cognitive ability SDP status First pregnancy Any value 0 1 Second pregnancy Any value Any value Any value Means (on a nine-point scale) and number of individuals First Second N pregnancy pregnancy 5.44 5.08 31,614 5.59 5.22 23,836 4.99 4.64 7,778 15 Any value Any value 0 1 0 1 0 1 0 0 1 1 5.58 4.94 5.62 5.17 5.05 4.92 5.21 4.59 5.25 4.80 4.68 4.57 25,064 6,550 22,928 2,136 908 5,642 Appendix Table 6f: Observed prevalences of criminal convictions in different exposure-combination groups Crime SDP status First pregnancy Any value 0 1 Any value Any value 0 1 0 1 Second pregnancy Any value Any value Any value 0 1 0 0 1 1 Prevalences and number of individuals First Second N pregnancy pregnancy 9.14% 10.20% 197,130 7.46% 8.37% 143,608 13.64% 15.10% 53,522 7.55% 8.51% 150,956 14.35% 15.74% 46,174 7.25% 8.16% 137,532 10.58% 12.02% 13,424 12.21% 13.17% 6,076 14.67% 16.13% 40,098 16 Appendix Table 6g: Observed prevalences of violent criminal convictions in different exposurecombination groups Violent crime SDP status First pregnancy Any value 0 1 Any value Any value 0 1 0 1 Second pregnancy Any value Any value Any value 0 1 0 0 1 1 Prevalences and number of individuals First Second N pregnancy pregnancy 1.76% 2.06% 196,726 1.17% 1.32% 143,370 3.36% 4.06% 53,356 1.20% 1.37% 150,704 3.59% 4.34% 46,022 1.09% 1.25% 137,310 2.33% 2.57% 13,394 2.84% 2.90% 6,060 3.71% 4.56% 39,962 Appendix Table 6h: Observed prevalences of drug/alcohol misuse in different exposure-combination groups Drug/alcohol misuse SDP status First pregnancy Any value 0 1 Any value Any value 0 1 0 1 Second pregnancy Any value Any value Any value 0 1 0 0 1 1 Prevalences and number of individuals First Second N pregnancy pregnancy 4.69% 5.88% 94,470 3.61% 4.70% 68,242 7.50% 8.97% 26,228 3.69% 4.72% 71,848 7.88% 9.57% 22,622 3.49% 4.53% 65,386 5.66% 6.62% 6,462 6.37% 8.40% 2,856 8.09% 9.73% 19,766 Appendix Table 7: Expected/observed means or proportions of mothers concordant and discordant for smoking during pregnancy for each outcome 17 Birth weight Pre-term birth Born small for gestational age Low academic achievement General cognitive ability Criminal conviction Violent criminal conviction Drug/alcohol misuse Observed concordant not SDP Observed concordant SDP Observed discordant in SDP 3514 4.80% 7.87% Expected discordant SDP, additivea 3471 5.08% 8.62% Expected discordant SDP, multiplicativea NA 5.04% 7.96% 3571 4.45% 5.17% 3371 5.71% 12.08% 5.98% 17.05% 11.65% 11.51% 10.26% 5.43 4.75 4.95 5.09 NA 7.71% 15.40% 11.73% 11.55% 10.98% 1.17% 4.13% 2.58% 2.65% 2.21% 4.01% 8.91% 6.53% 6.46% 6.01 % Abbreviation: NA, not applicable. a Expected discordant SDP refers to assuming that the effect of SDP has a linear effect on both siblings outcomes, i.e. discordant pairs have half the effect of the concordant smoking pairs effect relative to the concordant non-smoking pairs. 18 Appendix Table 8: Test of carry-over and contagion effects. Outcome Birth-weight Pre-term birth Born small for gestational age Low academic achievement General cognitive ability Crime Violent crime Drug/alcohol misuse Effect estimates (standard error) SDP1 on OUT2 SDP2 on OUT1 -152.5 (1.7) 0.32 (0.02) 0.86 (0.01) -148.4 (1.9) 0.13 (0.02) 0.72 (0.01) Interaction effect; p-value 0.110 <0.001 <0.001 1.00 (0.02) 1.08 (0.02) <0.001 -0.59 (0.03) -0.64 (0.04) 0.282 0.67 (0.02) 1.15 (0.05) 0.69 (0.03) 0.72 (0.02) 1.12 (0.05) 0.80 (0.05) 0.098 0.624 0.064 19 Appendix A: Description of the ACMPE model Each phenotype, either exposure or outcome, is assumed to be caused by the variance sources A, C, M, P, E and possible some covariates. Let i (=1, 2,…, N; N=number of extended families) be extended family ππ·π number, j (=1, 2) be nuclear family number, and k (=1, 2) be offspring number. Let π¦πππ be maternal πππ ππ·π smoking during pregnancy and π¦πππ be the outcome and ππππ and ππππ πππ be covariates, for individual k within nuclear family j and extended family i. Let ππ·ππππ and ππππππ be the residuals; the unexplained ππ·π πππ ππ·π ππ·π ππ·π variance in π¦πππ and π¦πππ when adjusted for covariates ππππ and ππππ πππ . Furthermore, let π΄πππ , πΆπππ , ππ·π ππ·π ππ·π ππππ , ππππ , and πΈπππ be random effects representing the previously defined variance sources for SDP πππ πππ πππ πππ (the exposure), and π΄πππ πππ , πΆπππ , ππππ , ππππ , and πΈπππ be random effects representing the variance sources for the outcome. These random effects are assumed normally distributed with mean zero. Finally, let π·ππ·π and π·πππ be regression coefficients for the covariates on the phenotypic values respectively. For brevity the model is here described using the unity link, as used for continuous variables; however for binary variables the liability threshold model is used (similar to using a probit link), the model generalizes to this situation. The observed phenotypes within an individual are defined by the equation ππ·π ππ·π π¦πππ ππππ π·ππ·π + ππ·ππππ [ πππ ] = [ πππ ], π¦πππ ππππ π·πππ + ππππππ where the residuals ππ·ππππ and ππππππ are decomposed into random effects: [ π ππ·ππππ ππ·π ππ·π ππ·π ππ·π ππ·π π΄πππ + πΆπππ + ππππ + ππππ + πΈπππ ] = [ π πππ πππ πππ πππ ]. ππππππ π΄πππ πππ + πΆπππ + ππππ + ππππ + πΈπππ Intercepts may depend on extended and nuclear family types, if so then indicators of family types are included in ππππ , and the intercepts regression coefficients are included in the π·-vectors. It should be noted that these parameters are estimated simultaneously with the random effects. We use extended families consisting of two nuclear families, and each nuclear family consists of one or two offspring of the same mother. Thus an extended family may consist of 4 individuals, each with two phenotypes: π ππ·π ππ·π ππ·π ππ·π ππ·π ππ·ππ11 π΄π11 + πΆπ11 + ππ11 + ππ11 + πΈπ11 π πππ πππ πππ πππ ππππ11 π΄πππ π11 + πΆπ11 + ππ11 + ππ11 + πΈπ11 π ππ·π ππ·π ππ·π ππ·π ππ·π ππ·ππ12 π΄π12 + πΆπ12 + ππ12 + ππ12 + πΈπ12 π πππ πππ πππ πππ ππππ12 π΄πππ π12 + πΆπ12 + ππ12 + ππ12 + πΈπ12 = π ππ·π ππ·π ππ·π ππ·π ππ·π . ππ·ππ21 π΄π21 + πΆπ21 + ππ21 + ππ21 + πΈπ21 π πππ πππ πππ πππ ππππ21 π΄πππ π21 + πΆπ21 + ππ21 + ππ21 + πΈπ21 π ππ·π ππ·π ππ·π ππ·π ππ·π ππ·ππ22 π΄π22 + πΆπ22 + ππ22 + ππ22 + πΈπ22 π πππ πππ πππ πππ [ππππ22 ] [π΄πππ π22 + πΆπ22 + ππ22 + ππ22 + πΈπ22 ] 20 The covariance between subjects, both within and across phenotypes, is dependent on their relation, both genetic and environmental. In Appendix Table 5 the correlation between random variables π΄, πΆ, ππ·π π, π, and πΈ within and across phenotypes are tabulated. Here ππ 2 is the variance of π΄πππ , ππ2 ditto for ππ·π 2 πππ 2 π΄πππ πππ , ππ is the variance of πΆπππ , ππ ditto for πΆπππ , et cetera. The parameter ππ΄ captures the genetic overlap between exposure and outcome, ππΆ captures the overlap for environmental πΆ part, et cetera. Appendix Table 5 shows covariance within individual and across relatives, both within each phenotype and across phenotypes. It can be seen that environments are assumed shared across phenotypes, and therefore across generations, by simple prolongation of how it looks within each generation. Note that all covariances between variance sources, both within and across phenotypes, are assumed to be zero, ππ·π ππ·π ππ·π πππ πππ πππ e.g. πΆππ£(π΄πππ , πΆπππ , ππππ , ππππ ) ≡ 0. The SEM is shown in ) ≡ 0, πΆππ£(πΆπππ ) ≡ 0, and πΆππ£(ππππ part as path diagrams in Appendix Figure 2 where an extended family structure of a family with one offspring in “Nuclear family 1” and a family with two offspring in “Nuclear family 2” are depicted. Each variance source is shown separately; however, they are simultaneously estimated. In the SEM an expected covariance matrix is specified, this covariance matrix is specific to any combination of siblings in nuclear families with any combination of mothers who are siblings in extended families. Since we allow for one to four individuals in the offspring generation (one: mothers without siblings with one offspring; four: two mothers who are siblings, each mother with two offspring) the extended families may be accompanied by an expected covariance matrix of dimensions ranging from 2x2 to 8x8. We can derive all covariances between phenotypes in one subject with that of another in the same extended family using Appendix Table 5 (or Appendix Figure 2), both between and within phenotypes; Within an individual, note that off-diagonals are symmetric: πππ(ππ·ππππ ) πΆππ£(ππ·ππππ , ππππππ ) [ ]= πΆππ£(ππππππ , ππ·ππππ ) πππ(ππππππ ) ππ 2 + ππ 2 + ππ 2 + ππ 2 + ππ 2 [1 ]. ππ ππ ππ΄ + ππ ππ ππΆ + ππ ππ ππ + ππ ππ ππ + ππ ππ ππΈ ππ2 + ππ2 + ππ2 + ππ2 + ππ2 2 Between full siblings, the π ′ sub-index is the complement of π: πΆππ£(ππ·ππππ , ππ·ππππ ′ ) πΆππ£(ππ·ππππ , ππππππ ′ ) [ ]= πΆππ£(ππππππ , ππ·ππππ ′ ) πΆππ£(ππππππ , ππππππ ′ ) [1 ππ 2 + ππ 2 + ππ 2 + ππ 2 π π π + ππ ππ ππΆ + ππ ππ ππ + ππ ππ ππ 2 π π π΄ 1 2 π 2 π + ππ2 + ππ2 + ππ2 ]. Between half siblings: ππ 2 + ππ 2 + ππ 2 πΆππ£(ππ·ππππ , ππ·ππππ ′ ) πΆππ£(ππ·ππππ , ππππππ ′ ) [ ] = [1 π π π + ππ ππ ππΆ + ππ ππ ππ πΆππ£(ππππππ , ππ·ππππ ′ ) πΆππ£(ππππππ , ππππππ ′ ) 2 π π π΄ 1 2 π 4 π + ππ2 + ππ2 ]. Between cousins where mothers are MZ twins, the π ′ sub-index is the complement of π, and “⋅” is used to indicate any of possible values at sub-index: 21 πΆππ£(ππ·ππππ , ππ·πππ′ ⋅ ) ππ 2 + ππ 2 πΆππ£(ππ·ππππ , πππππ′ ⋅ ) [ ] = [1 π π π + ππ ππ ππ πΆππ£(ππππππ , ππ·πππ′ ⋅ ) πΆππ£(ππππππ , πππππ′ ⋅ ) 2 π π π΄ 1 2 π 4 π + ππ2 1 2 π 8 π + ππ2 ]. Between cousins where mothers are DZ twins or full siblings: 1 2 π 2 π + ππ 2 [ ] = [1 πΆππ£(ππππππ , ππ·πππ′ ⋅ ) πΆππ£(ππππππ , πππππ′ ⋅ ) ππ ππ ππ΄ + ππ ππ ππ πΆππ£(ππ·ππππ , ππ·πππ′ ⋅ ) πΆππ£(ππ·ππππ , πππππ′ ⋅ ) 4 ]. Between cousins where mothers are maternal half siblings: 1 2 π 4 π + ππ 2 [ ] = [1 πΆππ£(ππππππ , ππ·πππ′ ⋅ ) πΆππ£(ππππππ , πππππ′ ⋅ ) ππ ππ ππ΄ + ππ ππ ππ πΆππ£(ππ·ππππ , ππ·πππ′ ⋅ ) πΆππ£(ππ·ππππ , πππππ′ ⋅ ) 8 1 2 π 16 π + ππ2 ]. Between cousins where mothers are paternal half siblings: πΆππ£(ππ·ππππ , ππ·πππ′ ⋅ ) 1 2 π 4 π πΆππ£(ππ·ππππ , πππππ′ ⋅ ) [ ] = [1 πΆππ£(ππππππ , ππ·πππ′ ⋅ ) πΆππ£(ππππππ , πππππ′ ⋅ ) ππ ππ ππ΄ 8 1 2 ]. π 16 π From this we may assemble any constellation of families included in the model. As an example, let extended family 1 be a mother with one offspring, let extended family 2 be a mother with two full sibling offspring, and let the mothers be maternal half siblings; the expected covariance matrix looks like ππ·ππ11 ππππ11 ππ·ππ21 πΆππ£ ππππ21 ππ·ππ22 ([ππππ22 ]) πππ(ππ·ππ11 ) πΆππ£(ππππ11 , ππ·ππ11 ) πππ(ππππ11 ) πΆππ£(ππ·ππ21 , ππ·ππ11 ) πΆππ£(ππ·ππ21 , ππππ11 ) = πΆππ£(ππππ21 , ππ·ππ11 ) πΆππ£(ππππ21 , ππππ11 ) πΆππ£(ππ·ππ22 , ππ·ππ11 ) πΆππ£(ππ·ππ22 , ππππ11 ) [πΆππ£(ππππ22 , ππ·ππ11 ) πΆππ£(ππππ22 , ππππ11 ) πππ(ππ·ππ21 ) πΆππ£(ππππ21 , ππ·ππ21 ) πππ(ππππ21 ) πΆππ£(ππ·ππ22 , ππ·ππ21 ) πΆππ£(ππ·ππ22 , ππππ21 ) πππ(ππ·ππ22 ) πΆππ£(ππππ22 , ππ·ππ21 ) πΆππ£(ππππ22 , ππππ21 ) πΆππ£(ππππ22 , ππ·ππ22 ) πππ(ππππ22 )] 22 ππ 2 + ππ 2 + ππ 2 + ππ 2 + ππ 2 1 π π π + ππ ππ ππΆ + ππ ππ ππ + ππ ππ ππ + ππ ππ ππΈ 2 π ππ΄ 1 2 π + ππ 2 4 π = 1 π π π + ππ ππ ππ 8 π ππ΄ 1 2 π + ππ 2 4 π 1 π π π + ππ ππ ππ [ 8 π ππ΄ … … … [ … ππ2 + ππ2 + ππ2 + ππ2 + ππ2 1 π π π + ππ ππ ππ 8 π ππ΄ 1 2 π + ππ2 16 π 1 π π π + ππ ππ ππ 8 π ππ΄ 1 2 π + ππ2 16 π ππ 2 + ππ 2 + ππ 2 + ππ 2 + ππ 2 1 π π π + ππ ππ ππΆ + ππ ππ ππ + ππ ππ ππ + ππ ππ ππΈ 2 π ππ΄ ππ 2 + ππ 2 + ππ 2 + ππ 2 … 1 π π π + ππ ππ ππΆ + ππ ππ ππ + ππ ππ ππ 2 π ππ΄ ] ππ2 + ππ2 + ππ2 + ππ2 + ππ2 1 π π π + ππ ππ ππΆ + ππ ππ ππ + ππ ππ ππ 2 π ππ΄ 1 2 π + ππ2 + ππ2 + ππ2 2 π ππ 2 + ππ 2 + ππ 2 + ππ 2 + ππ 2 1 π π π + ππ ππ ππΆ + ππ ππ ππ + ππ ππ ππ + ππ ππ ππΈ 2 π ππ΄ . ππ2 + ππ2 + ππ2 + ππ2 In the ACMPE model there are a number of parameters which may be estimated: ο· The intercepts and covariate regression parameters; π·ππ·π and π·πππ . ο· The variance of random parameters A, C, M, P, and E for exposure and outcome; ππ 2 , ππ 2 , ππ 2 , ππ 2 , ππ 2 , ππ2 , ππ2 , ππ2 , ππ2 , and ππ2 . ο· … The cross-phenotype correlation parameters; ππ΄ , ππΆ , ππ , ππ , and ππΈ . . For each variable, both exposure and outcomes, we may use the above defined variance, e.g. πππ(ππππππ ) = ππ2 + ππ2 + ππ2 + ππ2 + ππ2 , to estimate the fraction of variance explained by each variance source. For example, the narrow sense heritability for an outcome, i.e. the fraction of variance explained by additive genetic factors, may be expressed as narrow sense heritability = ππ2 . πππ(ππππππ ) Above we derived that the covariance between SDP and outcome within an individual is 1 πΆππ£(ππ·ππππ , ππππππ ) = ππ ππ ππ΄ + ππ ππ ππΆ + ππ ππ ππ + ππ ππ ππ + ππ ππ ππΈ . 2 The amount of explained covariance between phenotypes, within individual, per variance source is the particular source divided by the total amount of covariance, e.g. + ππ2 ] 23 1 ππ ππ ππ΄ 2 explained covariance due to additive genetics = πΆππ£(ππ·π πππ ,ππππππ ) . 24 Appendix B: Generalizability of mothers who change smoking status between pregnancies; carry-over and contagion effects Within sibling comparison analyses may be regarded as having an estimated effect that is closer to a causal effect compared to ordinary between individual analyses. Nevertheless, as all statistical models, comparisons within siblings rely on a number of assumptions, discussed in detail in D’Onofrio et al. (2013). We here address three assumptions which we believe are most important for our study. Generalizability. Are women who change their smoking behavior between two pregnancies different? The within sibling comparisons utilize siblings discordant for exposure to smoking while in utero. If mothers who change their smoking status between pregnancies are very different (i.e., in terms of how smoking during pregnancy (SDP) affects the outcome) from those who do not, the results of our analyses may not be generalizable to groups of women who do not change their smoking status between pregnancies. Appendix Tables 6a-h present the proportions and means in the outcomes for each possible combination of SDP in the offspring (where we used the first- and second-born of each mother having her first child in 1983 or later, and having at least two children in exposure periods defined as in the main article). To investigate the generalizability of pairs discordant for SDP to pairs concordant, we calculate the concordant proportions/means and an expected value for the discordant pairs (Appendix Table 7). We realize that “expected value” is scale-dependent, so we consider additive scale and multiplicative scale in our analyses. The expected value was calculated under the assumption that SDP had a linear effect (either additive scale [unity link] or a multiplicative scale [logit link], where applicable) for both offspring in the nuclear family, regardless if the mother smoked in the offspring’s pregnancy or his/her sibling’s. As can be seen in the table, for all associations the observed prevalences and means in the outcomes for SDP-discordant pairs were roughly halfway between that of SDP-concordant pairs. All values were closer to the predicted values than to the observed values in any of the concordant pairs. This pattern is what is expected if SDP is an expression of maternal characteristics (e.g., genes) and/or the association is causal. Thus, the data suggest that the SDP discordant mothers carry a liability that is in between the liabilities of the concordant mothers or that the association is causal, or a combination them between. Further, if discordant siblings were very different from the general population we would not expect within sibling results to be in agreement with results comparing cousins (Table 4 in the main paper). Taken together, our analyses don’t provide any evidence for the hypothesis that SDP results from discordant mothers should not be generalizable to other mothers. 25 Carry-over and contagion effects. Does smoking behavior in first pregnancy influence the next pregnancy? Another issue is whether there are carry-over effects from child 1 to child 2, that is, when the exposure status of sibling 1 (SDP1) affects the outcome of sibling 2 (OUT2) directly; for example, later born offspring to mothers who had their first child born during adolescence were affected by the effect of the adolescent child-bearing regardless of the mothers age at their own birth, maybe because environmental factors specifically associated with early maternal age of first child (e.g., diminished financial and social resources in families in which the mother had her first child at an early age (Coyne et al. 2013)). The carry-over effects should be distinguished from the contagion effects, which occurs when the outcome of sibling 1 (OUT1) affects the outcome of sibling 2 (OUT2). An example would be if criminal behavior of an older sibling would influence criminality in the younger siblings. We note that both of these potential effects would introduce an association between SDP1 and OUT2 through a pathway which is assumed to be non-existing in the sibling analyses. Thus, if no carry-over or contagion effect exists we would expect the association between the SDP1 and OUT2 to be of the same magnitude as the association between SDP2 on OUT1 .On the other hand, if carry-over or contagion effects are important, the effect of SDP1 on OUT2 should be stronger than the effect of SDP2 on OUT1. To investigate carry-over and contagion effects we tested whether the SDP1-OUT2 association differed from the SDP2-OUT1 association by including an interaction term between being first born and SDP in one pregnancy on the outcome in the other pregnancy (after controlling for the effects of being firstborn; Appendix Table 8). As evident in Appendix Table 8, in none of the associations where our analyses suggested that familial confounding was responsible for the entire association (i.e., all the cognitive/behavioral outcomes) there were no evidence for carry-over or contagion effects (note that for Low academic achievement, the indication is that child 2 influence child 1 rather than the other way around). We only observe significant interaction effects of the SDP1-OUT2 association being stronger than the SDP2-OUT1 association in pre-term birth and born small for gestational age. Thus, our main conclusion that smoking during pregnancy seem to be associated to adolescent cognitive capacity and late adolescent behavioral/conduct problems only due to genetic transmission of common susceptibility genes do not seem to be affected by violation of carry-over or contagion effects. References Appendix B Coyne, C. A., Langstrom, N., Rickert, M. E., Lichtenstein, P., D'Onofrio, B. M. (2013). Maternal age at first birth and offspring criminality: using the children of twins design to test causal hypotheses. Development and Psychopathology 25(1): 17-35. D'Onofrio, B. M., Lahey, B. B., Turkheimer, E. & Lichtenstein, P. (2013). Critical need for family-based, quasi-experimental designs in integrating genetic and social science research. American Journal of Public Health 103 Suppl 1, S46-55 26