Department of Applied Psychology, New York University Steinhardt School of Culture, Education, and Human Development, New York, New York, USA.
Department of Health Promotion and Development, University of Pittsburgh School of Nursing, Pittsburgh, Pennsylvania, USA.
Matern Child Nutr. 2019 Jan;15(1):e12655. doi: 10.1111/mcn.12655. Epub 2018 Sep 14.
Patterns of breastfeeding over time are not currently well understood. Limited qualitative and quantitative evidence suggests that there may be latent subgroups of mothers in the United States following very different trajectories of breast milk provision for their infants. This study used a quantitative modelling method (group-based trajectory modelling) to identify and describe these subgroups. Using data from the Infant Feeding Practices Study II (n = 3,023), the authors identified four distinct trajectories of breastfeeding intensity, each of which included a substantial subset of the total sample. A model with four groups fit the data well by objective and conceptual standards. Bivariate associations were analysed, and significant difference between trajectory group membership was found on an array of maternal and family determinants, including maternal demographics, hospital experience, and psychosocial resources, as well as on postweaning perceptions. These associations were used to create group profiles for each subgroup. Controlling for sociodemographic covariates, we also found that trajectory membership was significantly associated with several health outcomes for the target child 6 years later, including certain infections, picky eating, and health care utilization; trajectory group membership was also associated with maternal breastfeeding of subsequent children and maternal body mass index (BMI) at Year 6. Child BMI z-score and maternal breast cancer diagnosis were not significantly different across trajectory groups after accounting for confounding covariates, nor was time missed from school for the target child. Though exploratory, the initial identification and description of these four subgroups suggest future directions using breastfeeding trajectory methods, with potential implications for measurement, intervention development, and targeting.
目前，人们对母乳喂养时间模式的了解还不够充分。有限的定性和定量证据表明，在美国，可能存在一些潜在的母亲亚组，她们为婴儿提供母乳的轨迹非常不同。本研究使用定量建模方法（基于群组的轨迹建模）来识别和描述这些亚组。利用来自婴儿喂养实践研究 II（n=3023）的数据，作者确定了母乳喂养强度的四种不同轨迹，其中每一种轨迹都包含了总样本的一个很大子集。一个包含四个群组的模型在客观和概念标准方面都很好地拟合了数据。对双变量关联进行了分析，并发现轨迹组别的成员资格在一系列母婴和家庭决定因素上存在显著差异，包括母婴人口统计学特征、医院经历和心理社会资源，以及断奶后的认知。这些关联被用来为每个亚组创建群组档案。在控制社会人口学协变量的情况下，我们还发现，轨迹成员资格与目标儿童 6 年后的几个健康结果显著相关，包括某些感染、挑食和医疗保健利用；轨迹组别的成员资格也与母亲对后续儿童的母乳喂养以及母亲在第 6 年的体重指数（BMI）相关。在考虑了混杂协变量后，儿童 BMI z 分数和母亲乳腺癌诊断在轨迹组之间没有显著差异，目标儿童也没有因疾病而缺课。尽管这是初步的，但这四个亚组的初步识别和描述表明，未来可以使用母乳喂养轨迹方法进行研究，这可能对测量、干预措施的制定和目标人群的确定具有潜在意义。
