Dietary patterns and associated factors among the elderly

Abstract The aim of the present study was to identify dietary patterns among the elderly, and associate the same with sociodemographic characteristics, health and lifestyle. A cross-sectional study was carried out with a non-probabilistic sample of 295 elderly residents of both genders of the city of São Caetano do Sul, São Paulo, Brazil, who were users of public health units. Dietary patterns were identified considering 44 food items identified through 24-hour dietary recall. Factor analysis by the principal components method was used, adopting loads greater than 0.25 as the cut-off point, followed by Varimax orthogonal rotation. The associations were analyzed by the linear regression method, with variables with p<0.20 in univariate regression selected, and those with p<0.05 maintained in the multivariate model. There was a greater percentage of elderly females (85.1%) and those aged 60-69 years (46.4%). Three dietary patterns were identified: the "traditional" pattern was associated with the male gender (β=0.59, p<0.001) and physically active individuals (β=0.39, p=0.020); the "Pastas, pork and sweets" pattern was associated with the male gender (β=0.38, p=0.025) and retired individuals (β=0.55, p=0.017); and the "Coffee with milk and bread and butter" pattern was associated with an age of 80 years or older (β=0.55, p=0.004) and elderly persons who had difficulty chewing (β=0.38, p=0.013). The dietary patterns identified show that the elderly tend to maintain similar eating habits to the Brazilian population in general, and that diet changes to a similar degree as health and life style conditions.

摘要 本研究旨在明确老年人群的膳食模式，并探究其与社会人口学特征、健康状况及生活方式的关联。本研究为横断面研究（cross-sectional study），采用非概率抽样样本（non-probabilistic sample），招募对象为巴西圣保罗州南圣卡埃塔诺市的295名老年居民，涵盖两性，均为公共卫生服务机构的使用者。膳食模式的确定基于通过24小时膳食回顾法（24-hour dietary recall）收集的44项食物条目。研究采用主成分分析法（principal components method）进行因子分析（factor analysis），以载荷值大于0.25作为截断阈值，随后进行方差最大正交旋转法（Varimax orthogonal rotation）。关联分析采用线性回归（linear regression）方法：先筛选单变量回归（univariate regression）中p<0.20的变量，最终将p<0.05的变量纳入多变量模型（multivariate model）。本次研究对象中女性老年人占比更高（85.1%），60~69岁年龄段人群占比为46.4%。研究共识别出3种膳食模式：‘传统型’模式与男性性别（β=0.59，p<0.001）及体力活动活跃人群（β=0.39，p=0.020）呈显著正相关；‘意面、猪肉与甜食’模式与男性性别（β=0.38，p=0.025）及退休人群（β=0.55，p=0.017）呈显著正相关；‘咖啡配牛奶、面包与黄油’模式与80岁及以上高龄（β=0.55，p=0.004）及存在咀嚼困难的老年人（β=0.38，p=0.013）呈显著正相关。本研究识别的膳食模式表明，老年人群的饮食习惯整体与巴西普通人群相近，且膳食结构的变化程度与健康及生活方式状况的变化程度呈对应关系。