Predictive model of sperm DNA fragmentation using spermato-gram determined parameters

• Autores: Portella, Jimmy, López, Rosmary, Noriega-Hoces, Luis, Guzmán, Luis
• Formato: artículo
• Fecha de Publicación: 2014
• Institución: Sociedad Peruana de Obstetricia y Ginecología
• Repositorio: Revista SPOG - Revista Peruana de Ginecología y Obstetricia
• Lenguaje: español
• OAI Identifier: oai:ojs.spog:article/106
• Enlace del recurso: http://www.spog.org.pe/web/revista/index.php/RPGO/article/view/106
• Nivel de acceso: acceso abierto

Introduction: The spermatogram is used as a test of seminal quality. Recently, the sperm fragmentation test has demonstrated importance as sperm DNA integrity would affect clinical results in assisted reproduction treatments. Objectives: To determine spermatogram variables that would independently predict sperm DNA fragmentation index (SFI). Design: Retrospective, comparative study. Settings: Grupo PRANOR, Reprogenetics Latinoamerica, Clinica Concebir, Lima, Peru. Biologic material: Sperm. Methods: Individual variables and two models were compared: the first model considered percentage of sperm viability and patient’s age; the second model included percentage of motile sperms and age. Logistic regression analysis was done. Main outcome measures: Sperm viability, age. Results: Multivariate analysis showed that both models were significantly superior to individuals variables (p<0.01). The first model had non standardized coefficient values (95%CI) respectively of 0.200 (0.082 to 0.318) and -0.146 (-0.206 to -0.086). The second model had non standardized coefficient values (95%CI) respectively of -0.099 (-0.157 to -0.042) and 0.219 (0.99 to 0.339). Logistic regression analysis showed that the percentage of sperm viability and patient’s age predicted the probability of having an SFI over 30% with age non standardized coefficient values of 95%IC 0.034 (0.015 to 0.053) and viability percentage of-0.043 (0.034 to 0.052). Additionally the second model had 95%CI respectively of -0.04 (-0.031 to -0.049) and 0.035 (0.017 to 0.053). Finally, a ROC curve to determine the superiority of some model over individual variables showed that areas under the curve (ABC) of model 1 (age and sperm viability) was 0.727 (95%CI = 0.665 to 0.790) and model 2 (age and sperm total motility) 0.675 (95%CI = 0.606 to 0.744), compared with ABC of percentage of sperm viability = 0.295 (95%CI = 0.229 to 0.362), sperm total motility ABC = 0.333 (95%CI = 0.264 to 0.403) and patient’s age with ABC 0.584 (95%CI = 0.510 to 0.658). Conclusions: Age, and sperm motility and viability independently correlated with SFI and consequently these variables could be used as predictors of DNA fragmentation percentage.