Supporting information.

Fig A in S1 Appendix: Key Word Strategy. Fig B in S1 Appendix: Quality assessment of cohort studies using Adapted Newcastle Ottawa Scale. Fig C in S1 Appendix: Quality assessment of case-controlled studies using Adapted Newcastle Ottawa Scale. Fig D in S1 Appendix: Quality assessment of cross-sectional studies using Adapted Newcastle Ottawa Scale. Fig E in S1 Appendix: Funnel plots of the association of miscarriage and (a) ischaemic heart disease and (b) breast cancer. Fig F in S1 Appendix: Forrest plots presenting pooled RRs for development of (a) ischaemic heart disease, (b) cerebrovascular disease and (c) circulatory diseases for the miscarriage arm using the Mantel–Haenszel random effects model with dichotomous data in meta-analysis. Fig G in S1 Appendix: Forrest plots presenting pooled RRs for development of (a) type 2 diabetes mellitus and (b) depression for the miscarriage arm using the Mantel–Haenszel random effects model with dichotomous data in meta-analysis. Fig H in S1 Appendix: Forrest plots presenting pooled RRs for development of (a) breast cancer, (b) ovarian cancer, and (c) uterine malignancies for the miscarriage arm using the Mantel–Haenszel random effects model with dichotomous data in meta-analysis. Fig I in S1 Appendix: Forrest plots presenting pooled RRs for development of (a) ischaemic heart disease, (b) cerebrovascular disease, (c) breast cancer, and (d) ovarian cancer for recurrent vs. single miscarriage using the Mantel–Haenszel random effects model with dichotomous data in meta-analysis. Fig J in S1 Appendix: Forrest plots presenting pooled RRs for development of (a) ischaemic heart disease, (b) cerebrovascular disease, and (c) breast cancer for recurrent vs. no miscarriage using the Mantel–Haenszel random effects model with dichotomous data in meta-analysis. Fig K in S1 Appendix: Forrest plots presenting pooled RRs for development of (a) ischaemic heart disease, (b) cerebrovascular disease, (c) breast cancer, and (d) ovarian cancer for recurrent vs. single miscarriage using the Generic inverse variance random effects model meta-analysis. Fig L in S1 Appendix: Forrest plots presenting pooled RRs for development of (a) ischaemic heart disease, (b) cerebrovascular disease, and (c) breast cancer for recurrent vs. no miscarriage using the Generic inverse variance random effects model with dichotomous data in meta-analysis. Fig M in S1 Appendix: Forrest plots presenting pooled RRs for development of (a) ischaemic heart disease, (b) cerebrovascular disease, and (c) circulatory disease for stillbirth arm using the Mantel–Haenszel random effects model with dichotomous data in meta-analysis. Fig N in S1 Appendix: Forrest plots presenting pooled RRs for development of (a) breast cancer, (b) female malignancies, and (c) all malignancies for stillbirth arm using the Mantel–Haenszel random effects model with dichotomous data in meta-analysis. Fig O in S1 Appendix: Forrest plots presenting pooled RRs for development of (a) type 2 diabetes, (b) renal disease, (c) depression for stillbirth arm using the Mantel–Haenszel random effects model with dichotomous data in meta-analysis. Table A in S1 Appendix: Outcome definitions of the miscarriage arm of the meta-analysis. Table B in S1 Appendix: Outcome definitions of the stillbirth arm of the meta-analysis. Table C in S1 Appendix: Covariate/Confounding variables the researchers adjusted for in the studies of miscarriage. Table D in S1 Appendix: Covariate/Confounding variables the researchers adjusted for in the studies of stillbirth.

(DOCX)