Childhood nutrition and cardiovascular disease risk: People in training for a plant-centered diet

Evidence is strong for a holistic approach for diet to prevent cardiovascular disease, diabetes, and other conditions. This includes dietary strategies, such as following a Mediterranean diet or otherwise eating a dietary pattern with low meat intake and a high intake of fruit, vegetables, legumes, whole grains, nuts, berries, seeds generally, unrefined unsaturated oils, fish, coffee, tea, chocolate, and alcohol (among drinkers, not in excess; intoxication should be avoided).1 Water is very important as a beverage. Some forms of dairy foods may also be recommended.2-4 Although an understanding of nutrients is important for controlling deficiency diseases, nutrients are not a good model for choosing what foods to eat. A nutrient approach to nutrition has been confusing to consumers. Paramount in this concept are the terms “low fat” and “low carbohydrate”; neither is particularly useful for choosing what foods to eat. “Low fat” as a food choice rule excludes apparently healthy foods, such as olives, walnuts, and salmon. “Low carbohydrate” would include these high-fat foods, but seems to discourage the consumption of whole plant foods. The idea that there is a concerted action of food constituents is called “food synergy”; this idea also supports that certain dietary patterns may be healthy. The food synergy and holistic idea is further supported by the notion that variety in the consumption of nutrient- or bioactive-rich plant food helps maintain health because there are many compounds with many functions. A simple example of the highly nuanced nature of biochemicals on health is the difference between oleic acid and its trans counterpart elaidic acid. The two molecules differ only in the placement of one double bond, yet the cis version, oleic acid, is a major part of olive oil, which is considered to be healthy,5 whereas the trans version, elaidic acid, is rare in nature, produced by hydrogenation of vegetable oils, and is associated with excess risk of coronary heart disease.6 The consumer does not know in advance of eating which compounds are needed on any given day.

There is also an epidemiologic advantage to studying diet patterns. Although many individual foods and nutrients have been seen to have very high within-person variability, diet patterns act like many other cardiovascular or diabetes risk factors, showing correlations of ~0.5 over periods up to 20 years. Jacobs and Orlich¹ reviewed reports of 16 diet pattern scores and in all but one case found tracking correlations over several to many years of follow-up to range from 0.39 to 0.65.

Jacobs and colleagues have formulated the A Priori Diet Quality Score (APDQS).1, 7-10 The theoretical basis for the APDQS is discussed by Jacobs and Orlich.1 Because the APDQS is based on foods, it differs somewhat from study to study, depending on which food groups were asked about. In Project EAT (Hu et al.12, table 4), positively rated food groups were formed, namely beans, chocolate, coffee, fish, fruit, green vegetables, low-fat dairy, nuts, poultry, tea, tomato, whole grain, and yellow vegetables. The negatively rated food groups were butter, fried foods, fried potato, grain dessert, organ meat, processed meat, red meat, salty snacks, sauces, soft drinks, sweet bread, and whole fat dairy. Each food group was placed in quintiles (or a non-consumer group and quartiles among consumers), and categories were scored 0–4 for lowest to highest reported consumption. The APDQS was calculated by summing category scores (0–4) of positively rated food groups with the reverse scores (4–0) of the negatively rated food groups. Additional foods were rated as neutral, namely diet soft drinks, eggs, fruit juice, margarine, other vegetables, potatoes, refined grains, seafood, and soup; these provide the full dietary context, but do not enter the score except through energy balance (replacing either positively or negatively rated foods). The APDQS has been shown to be inversely associated with total death that occurred over 26 years of follow-up from 1986 in Iowa women who were then aged 55–69 years.11 The APDQS is also related to health parameters in people followed from young adulthood.8, 10 However, APDQS has not been the subject of much study in children. Using data from Project EAT,12 we hypothesized that: (i) APDQS tracks from adolescence into young adulthood; and (ii) higher and increasing APDQS is associated with less weight gain in these participants. The following summarizes published data.12

In Project Eat, 4746 adolescents were recruited from middle or high school in Minneapolis–Saint Paul (MN, USA) in 1998–1999, with a mean age of 15 years (range 11–18 years). The study sample for the present report was the 2656 participants with follow-up visits in 2004 (mean age 20 years) and/or 2009 (mean age 25 years), adequate diet data, and self-reported weight and height at each examination. There were 1226 males and 1633 subjects were White in a racially diverse sample.

Between mean age 15 and 25 years, mean (±SD) self-reported weight increased from 61 ± 15 to 76 ± 19 kg, whereas the corresponding mean APDQS was 45 ± 12 and 53 ± 12 points. There was no significant change in mean energy intake over these ages.

As hypothesized, within-person tracking correlation of APDQS was 0.50 from mean age 20 to 25 years, but was a little lower (0.31) during the teenage years, from mean age 15 to 20 years.

Per 15-point APDQS baseline difference (mean age 15 years), there was no cross-sectional weight difference: (mean 0.1 kg; 95% confidence interval [CI] –0.8, 0.8 kg). However, as hypothesized, there was less weight gain over the period 1999–2009 (mean − 1.8 kg; 95% CI –2.6, −1.0 kg) per 15-point higher APDQS at mean age 15 years. There was also less concurrent weight gain per 15-point APDQS increase between mean age 15 and 25 years (mean − 1.8 kg; 95% CI –2.2, −1.4 kg). There was a 5.6-kg mean difference in weight at mean age 25 years between participants who were below the median APDQS at mean age 15 years and had a decrease in APDQS between mean age 15 and 25 years and those participants who were above median APDQS at mean age 15 years and had an increase in APDQS between mean age 15 and 25 years (Table 1).

In conclusion, a plant-centered dietary pattern with some resemblance to a Mediterranean diet is known to protect against chronic disease when eaten in adulthood.5, 7, 8, 10, 11 Children experiment with eating and learn dietary patterns,13 and have high energy intake during adolescence. Family meals, in particular, are a strategy that may promote better eating in adolescence.14, 15 Therefore, non-obese children may not suffer the same adverse metabolic consequences from eating high-energy foods that are not nutrient dense as adults. Children have a remarkable amount of latitude in food choice. We observed that eating a plant-centered diet by adolescence led to lower body mass index in young adulthood and appeared to set these individuals up for chronic disease avoidance throughout adulthood.12

Disclosure

DRJ is a consultant to the California Walnut Commission. TH has nothing to declare.

既往已有强有力的证据表明,  全面控制饮食的方法可以预防心血管疾病、糖尿病以及其他疾病。这包括饮食策略,  例如遵循地中海饮食或者其他饮食模式,  其饮食特征为少量摄入肉类与大量摄入水果、蔬菜、豆类、全谷物、坚果,  浆果、一般的种子、未精炼的不饱和油、鱼、咖啡、茶、巧克力与酒精(饮酒者不应过量;应避免醉酒)¹。水是一种非常重要的饮料。也推荐可以饮用某些类型的乳制品^2–4。虽然对于控制营养缺乏性疾病来说了解营养成分非常重要,  但是对于选择吃什么食物来说根据营养素选择并不是一个很好的模式。补充营养素的营养疗法一直困扰着消费者。在这个概念中最重要的是“低脂肪”与“低碳水化合物”这两个术语;它们对于要选择吃什么食物来说都不是特别有用。“低脂肪”作为一种食物选择规则显然要将一些健康的食品如橄榄、核桃与三文鱼排除在外了。“低碳水化合物”将包括那些高脂肪的食物,  但是似乎还阻碍了全植物食品的消费。各种食物成分协同作用的理念被称为“食物协同作用”;这个理念也支持某些可能是健康的饮食模式。食物协同作用与整体理念进一步得到了以下观念的支持,  亦即消费多样化的富含营养成分或者生物活性的植物性食品将有助于保持健康,  因为其中包含了许多具有多种功能的化合物。生物化学品对健康的影响具有高度细微的差别,  有一个简单的例子那就是油酸与反式油酸之间的区别。这两个分子仅在一个双键的位置上有所不同,  但是油酸是顺式版本,  它是橄榄油的主要部分,  目前认为它是健康的⁵,然而反油酸是反式版本,  它在自然界中很罕见,  由植物油氢化后生成,  它是冠心病的极度危险因素⁶。消费者在任何一天的进食前都不知道需要的是哪种化合物。

另外,  使用流行病学调查的方法有利于我们研究饮食模式。虽然目前已经发现许多个体消费的食物以及营养物质在个体内的变异性非常高,  但是饮食模式就像其他许多心血管疾病或糖尿病的危险因素一样,  在长达20年的随访期间发现它的相关系数大约为0.5。Jacobs与Orlich¹对16种饮食模式的评分报告进行了综述,  在这些报告中为了跟踪相关性随访了数年至许多年,  结果发现除了一例以外在所有人中的相关系数范围为0.39至0.65。

Jacobs与同事们为我们确切地阐述了先验性饮食质量评分法(A Priori Diet Quality Score,APDQS)^1,7-10。Jacobs与Orlich针对APDQS的理论基础进行了讨论¹。因为APDQS是以食物为基础的研究方法,  研究与研究之间都有所不同,  取决于被问及的是哪类食物。在EAT项目(Hu等¹²,表4)中,  规定了正面评价的食物类别,  即豆类、巧克力、咖啡、鱼、水果、绿色蔬菜、低脂乳制品、坚果、家禽、茶、番茄、全谷物以及黄色蔬菜。负面评价的食物类别是黄油、油炸食品、炸土豆、谷物甜点、内脏肉、加工肉类、红肉、咸味点心、酱料、软饮料、甜面包以及全脂乳制品。每种食物类别都按照五分位分组(或是一个非消费者分组,  而在消费者中按照四分位数分组),并且按照最少至最多报告的消费量进行分类(评分为0至4)。通过将正面评价的食物组类别得分(0-4)与负面评价的食物组类别得分(4-0)进行求和计算后得出APDQS评分。其他食物被评价为中性食品,  即软饮料、鸡蛋、果汁、人造奶油、其他蔬菜、土豆、精制谷物、海产品以及汤类;加上这些食物后饮食环境才够完整,  但它们不进入评分,  除非能够(替代正面或者负面评价的食物)影响能量平衡。有一项研究从1986年开始对爱荷华州的妇女进行了长达26年的随访,  刚开始的时候她们的年龄在55岁至69岁之间,  研究结果发现APDQS与总死亡率之间呈负相关¹¹。对年轻成人进行随访后也发现APDQS与健康参数相关^8,10。然而,APDQS一直还都没有成为儿童研究的对象。利用来自EAT项目的数据¹²,我们假设如下:(i)APDQS的轨迹是从青春期到成年早期;并且(ii)在这些参与者中,APDQS更高以及增加都与体重增加较少有关。以下我们总结了已公布的数据¹²。

在EAT项目中,1998至1999年间从明尼阿波利斯—圣保罗(明尼苏达州,  美国)中学或者高中招募了4746名青少年,  平均年龄为15岁(年龄范围为11–18岁)。这份报告的研究样本有2656名参与者,  分别在2004年(平均年龄20岁)和/或2009年(平均年龄25岁)进行了随访,  在每次调查的时候都要记录足够多的饮食数据以及自我报告的体重与身高。在这个具有多种多样种族的样本中,  有1226名男性以及1633名白种受试者。

在平均年龄为15至25岁之间的受试者中,  自我报告的平均体重(±SD)从61±15 kg增加到了76±19 kg,而相应的平均APDQS评分从45±12分增加到了53±12分。在这些年龄段受试者的平均能量摄入没有显著的变化。

就像假设的一样,  从平均年龄20岁至25岁,  追踪到的受试者个体内的与APDQS的相关系数为0.50,但是在十几岁(平均年龄为15至20岁)的受试者中相关系数要小一点(0.31)。

基线时每15分以内的APDQS差异(平均年龄为15岁),体重都没有横截面差异:(平均0.1 kg;95%置信区间[CI]为-0.8,0.8 kg)。然而,  就像假设的一样,  从1999年至2009年,  在平均年龄为15岁的受试者中,APDQS评分每升高15分,  体重增加就会显著减少(平均-1.8 kg;95%CI为-2.6,-1.0 kg)。在平均年龄为15岁至25岁之间的受试者中,APDQS评分每升高15分,  体重增加也会同时显著减少(平均-1.8 kg;95% CI为-2.2,-1.4kg)。平均年龄15岁时APDQS评分低于中位数并且平均年龄为15至25岁之间时APDQS评分有所减少的参与者与平均年龄15岁时APDQS评分高于中位数并且平均年龄为15至25岁之间时APDQS评分有所增加的参与者相比,  当他们的平均年龄到了25岁时,  两组之间的平均体重差异达到了5.6 kg(表1)。

总之,  以植物为主的饮食模式与地中海饮食有某些相似之处,  目前已经明确成年以后如果采用这种饮食模式可以预防慢性疾病^5,7,8,10,11。儿童在不断地尝试不同的饮食并且在学习采用何种饮食模式¹³,到了青春期还存在高能量摄入。我们特别推荐家庭餐,  这是一种可以促进青少年采用更好饮食模式的策略^14,15。因此,  非肥胖儿童摄入并非富含营养素的高能量食品后可能不会像成人那样出现同样的不良代谢结果。儿童在食物的选择范围上明显更广。我们观察到,  青少年进食以植物为主的饮食会导致成年早期时较低的体重指数,  并且似乎这样就会导致这些个体在整个成年期间都避免发生慢性疾病¹²。