Leynen Françoise, De Backer Guy, Pelfrene Edwin, Clays Els, Kittel France, Moreau Michel, Kornitzer Marcel
Department of Epidemiology and Health Promotion, School of Public Health, Brussels Free University (ULB), Brussels, Belgium.
Eur J Cardiovasc Prev Rehabil. 2006 Apr;13(2):261-7. doi: 10.1097/01.hjr.0000194420.62379.de.
The 'labelling hypothesis' was introduced on the basis of the observation that labelling subjects with blood pressure elevation as hypertensive was associated with an increase in sickness absence. In the Belstress I study this hypothesis was analysed in the same way for the possible influence on sick leave of labelling persons with elevated cholesterol as hypercholesterolaemic.
The Belstress I cohort concerns a sample of more than 16,000 men and 5,000 women at work in 24 Belgian industries in various sectors. Baseline data were collected by questionnaire and clinical examination. Awareness was defined as answering positively to the question 'did a physician ever tell you that your blood pressure/serum cholesterol was too high?' Sick leave data were independently and objectively recorded during 1 year following the screening. Sick leave was treated in a dichotomous way whereby the event was defined as being in the highest quartile of the annual number of days of sick leave (10 days or more for men and 15 days or more for women) or as being in the highest quartile of the annual number of spells of sick leave (two spells or more for both sexes). Gender-specific logistic regression analyses were performed, with adjustment for a large set of covariates. A positive association was observed between both awareness of hypertension and awareness of hypercholesterolaemia and the various definitions of sick leave, in both sexes and after adjustment for different covariates. When dividing up aware subjects into treated versus untreated, we observed in men the highest sick leave incidence in aware and treated hypertensive patients as well as in aware and treated hypercholesterolaemic patients. In women findings were less consistent, probably due to the smaller sample size. When looking at cumulative effects by examining participants with both hypertension and hypercholesterolaemia and their level of awareness for one or both risk factors, a statistically significant gradient was noticed in men, with the highest sick leave incidence, whatever the definition, in men aware for both risk factors, followed by men aware for one. In women the same trends were observed, but no level of statistical significance was reached.
Without being able to test the effect of 'labelling' as such, our study provides support for the association between awareness of two different coronary risk factors and incidence of sick leave. Probably a common mechanism is at the base of these findings. Further research is needed, in order to reduce potential negative effects of screening on human wellbeing as well as on productivity.
“标签假说”是基于以下观察结果提出的，即给血压升高的受试者贴上高血压标签与病假天数增加有关。在“Belstress I”研究中，以同样的方式分析了给胆固醇升高的人贴上高胆固醇血症标签对病假可能产生的影响。
“Belstress I”队列研究涉及比利时24个不同行业的16000多名在职男性和5000多名在职女性。通过问卷调查和临床检查收集基线数据。知晓情况定义为对“医生是否曾告诉你血压/血清胆固醇过高？”这一问题回答“是”。在筛查后的1年里，独立且客观地记录病假数据。病假情况采用二分法处理，即事件定义为处于病假天数年度最高四分位数（男性为10天或更多，女性为15天或更多）或处于病假次数年度最高四分位数（男女均为两次或更多）。进行了性别特异性逻辑回归分析，并对大量协变量进行了调整。在男女两性中，经不同协变量调整后，高血压知晓情况和高胆固醇血症知晓情况与病假的各种定义之间均观察到正相关。当将知晓的受试者分为接受治疗和未接受治疗两组时，我们观察到，在男性中，知晓且接受治疗的高血压患者以及知晓且接受治疗的高胆固醇血症患者的病假发生率最高。在女性中，结果不太一致，可能是由于样本量较小。通过检查同时患有高血压和高胆固醇血症的参与者及其对一种或两种危险因素的知晓水平来观察累积效应时，在男性中发现了一个具有统计学意义的梯度，无论采用何种定义，知晓两种危险因素的男性病假发生率最高，其次是知晓一种危险因素的男性。在女性中也观察到了相同的趋势，但未达到统计学显著性水平。
尽管无法测试“贴标签”本身的效果，但我们的研究为两种不同冠心病危险因素的知晓情况与病假发生率之间的关联提供了支持。这些发现背后可能存在一个共同机制。需要进一步研究，以减少筛查对人类健康以及生产力的潜在负面影响。
