Gonzalez-Perez G J, Vega-Lopez M G, Cabrera-Pivaral C
Center for Health, Population and Human Development Studies, University of Guadalajara, Mexico.
J Hum Lact. 1998 Dec;14(4):297-303. doi: 10.1177/089033449801400413.
This study seeks, through a logistic regression model, to describe the pattern of breastfeeding duration in Guadalajara, Mexico, during 1993. A multistage random sample of children under 1 year of age (n = 1036) was studied; observational data regarding breastfeeding duration, obtained through a "status quo" procedure, were compared with prevalence rates obtained from the logistic regression model. Modeling the duration of breastfeeding during the first year of life rather than only analyzing observational data helps researchers to understand this process in a dynamic and quantitative way. For example, uncommon indicators of breastfeeding were derived from the model. These indicators are impossible to obtain from observational data. The prevalence curve estimated through the logistic model was adequately fitted to observed data: there were no significant differences between the number or distribution of breastfed infants observed and those predicted by the model. Moreover, the model revealed that less than 40% of the children were breastfed in the fourth month of life; the median age for weaning was 39.3 days; 55% of the potential breastfeeding in the first 4 months did not occur; and the greatest abandonment of breastfeeding in the first 4 months was observed in the first 60 days. Thus, logistic regression seems a suitable option to construct a population-based model that describes breastfeeding duration during the first year of life. The indicators derived from the model offer health care providers valuable information for developing programs that promote breastfeeding.
本研究旨在通过逻辑回归模型描述1993年墨西哥瓜达拉哈拉市母乳喂养持续时间的模式。对1036名1岁以下儿童进行了多阶段随机抽样研究；通过“现状”程序获得的关于母乳喂养持续时间的观察数据，与从逻辑回归模型得出的患病率进行了比较。对生命第一年的母乳喂养持续时间进行建模，而不是仅分析观察数据，有助于研究人员以动态和定量的方式理解这一过程。例如，从模型中得出了不常见的母乳喂养指标。这些指标无法从观察数据中获得。通过逻辑模型估计的患病率曲线与观察数据拟合良好：观察到的母乳喂养婴儿数量或分布与模型预测的结果之间没有显著差异。此外，该模型显示，不到40%的儿童在出生后第四个月进行母乳喂养；断奶的中位年龄为39.3天；前四个月中55%的潜在母乳喂养没有发生；在前四个月中，母乳喂养的最大放弃率出现在前60天。因此，逻辑回归似乎是构建一个基于人群的模型来描述生命第一年母乳喂养持续时间的合适选择。从模型中得出的指标为医疗保健提供者制定促进母乳喂养的项目提供了有价值的信息。
