Introduction:

it would be useful to establish a model for predicting child and adolescent obesity through modifiable factors from the first thousand days of life.

Objective:

to develop an early predictive model of child and adolescent obesity.

Methods:

a longitudinal, retrospective and observational case-control study was conducted at the "Manuel González Díaz" Polyclinic, period March/2021-May/2023. Universe, consisting of 3091 children under 19 years of age (Basic Working Group 2), forming a Sample A [105 children under two years of age (35 cases and 70 controls)] and a Sample B [(144 adolescents (72 cases and 72 controls)]. The factors associated with obesity were identified, with them an early predictive model of obesity was developed and validated using methods such as: logistic regression analysis, goodness of fit test of Hosmer and Lemeshow and estimation of the Area Under Curve-ROC.

Results:

in Sample A two factors were associated with childhood and adolescent obesity: history of maternal obesity [OR 3.32 (CI: 1.11-9.88)] and maternal smoking [OR 6.15 (CI: 1.07-35.15)], calibration (p=0.1729), AUC-ROC 0.68. The results were similar in Sample B.

Conclusions:

the history of maternal obesity and maternal smoking were predictive factors of child and adolescent obesity with an adequate adjustment of reality in sample A and similar results in sample B, a fact consistent with the fact that both factors achieve a significant reduction in distance, so its use is proposed in clinical practice to predict from the first thousand days of life the future development of child and adolescent obesity.

        · resumen en Español     · texto en Español     · Español ( pdf )