Bern University Hospital for Mental Health, Centre for Psychiatric Rehabilitation, Bern, Switzerland.
Department of Health Professions, Bern University of Applied Sciences, Bern, Switzerland.
Acta Psychiatr Scand. 2019 Nov;140(5):393-407. doi: 10.1111/acps.13083. Epub 2019 Sep 3.
The question whether mental illness prevalence rates are increasing is a controversially debated topic. Epidemiological articles and review publications that look into this research issue are often compromised by methodological problems. The present study aimed at using a meta-analysis technique that is usually applied for the analysis of intervention studies to achieve more transparency and statistical precision.
We searched PubMed, PsycINFO, CINAHL, Google Scholar and reference lists for repeated cross-sectional population studies on prevalence rates of adult mental illness based on ICD- or DSM-based diagnoses, symptom scales and distress scales that used the same methodological approach at least twice in the same geographical region. The study is registered with PROSPERO (CRD42018090959).
We included 44 samples from 42 publications, representing 1 035 697 primary observations for the first time point and 783 897 primary observations for the second and last time point. Studies were conducted between 1978 and 2015. Controlling for a hierarchical data structure, we found an overall global prevalence increase in odds ratio of 1.179 (95%-CI: 1.065-1.305). A multivariate meta-regression suggested relevant associations with methodological characteristics of included studies.
We conclude that the prevalence increase in adult mental illness is small, and we assume that this increase is mainly related to demographic changes.
精神疾病发病率是否增加是一个备受争议的话题。研究这一研究问题的流行病学文章和综述出版物往往存在方法学问题。本研究旨在使用元分析技术，通常用于干预研究的分析，以实现更高的透明度和统计精度。
我们在 PubMed、PsycINFO、CINAHL、Google Scholar 和参考文献中搜索了基于 ICD 或 DSM 诊断、症状量表和困扰量表的成人精神疾病患病率的重复横断面人群研究，这些研究至少在同一地理区域内使用相同的方法学方法进行了两次以上。该研究已在 PROSPERO（CRD42018090959）中注册。
我们首次纳入了 42 篇文献中的 44 个样本，代表了 1035697 个第一次观察和 783897 个第二次也是最后一次观察的主要观察结果。这些研究是在 1978 年至 2015 年期间进行的。在控制了分层数据结构后，我们发现总体优势比的全球患病率增加了 1.179（95%-CI：1.065-1.305）。多元元回归表明与纳入研究的方法学特征存在相关关联。
我们得出结论，成人精神疾病的患病率增加很小，我们假设这种增加主要与人口统计学变化有关。
