Haynes R B
Department of Clinical Epidemiology and Biostatistics, McMaster University Faculty of Health Sciences, Hamilton, Canada.
Am J Clin Nutr. 1997 Feb;65(2 Suppl):622S-625S. doi: 10.1093/ajcn/65.2.622S.
Does dietary sodium in excess of a specified daily consumption substantively elevate blood pressure in normotensive people? Does lowering the daily consumption of sodium reduce blood pressure in normotensive people? Sound observational and interventional studies can address these questions, but there are substantial differences in the ability of various research designs to provide clear, bias-free answers. I summarize established scientific principles for addressing issues of causation and the effects of interventions and compare observational and interventional designs. Observational studies are important in exploring the possible determinants of health problems but are subject to bias and cannot directly assess the effects of interventions. They are superseded by sound interventional studies, particularly randomized controlled trials, in answering the key questions concerning causes and benefits of intervention.
超过特定每日摄入量的膳食钠是否会实质性地升高血压正常者的血压？降低每日钠摄入量是否会降低血压正常者的血压？可靠的观察性研究和干预性研究可以回答这些问题，但各种研究设计在提供清晰、无偏差答案的能力上存在很大差异。我总结了解决因果关系问题和干预效果的既定科学原则，并比较了观察性设计和干预性设计。观察性研究在探索健康问题的可能决定因素方面很重要，但容易受到偏差影响，且无法直接评估干预效果。在回答有关干预的原因和益处的关键问题时，可靠的干预性研究，特别是随机对照试验，会取代观察性研究。
