| BackgroundAs the development of society and the economy, national population structure is undergoing great challenges. In China, the percentage of individuals over 60 years had risen to 15.6% in 2014. Since the dramatic ageing of the population, and the family structure has changed from extended to nuclear families, problems such as financial and social burdens of providing for the aged, and empty-nest syndrome have risen. These could potentially lead to mental health problems among elderly individuals.Anxiety and depression are highly prevalent and challenging psychological problems experienced by elderly individuals, which are linked to occurrence and progression of various illnesses. Anxiety is related to increased occurrence and progression of cardiovascular diseases, and it is considered as an independent risky factor. Depression is also increased occurrence and progression of cardiovascular diseases, and is considered as high risk factor of acute cardiovascular events. Besides, depression is related to occurrence of multiple sclerosis, diabetes mellitus, and cognitive impairment. There have also been lots of studies about the relation of depression and suicide. So far, treatment measures of anxiety or depression contain medication treatment, cognitive-behavioral therapy and the others. Studies showed that outpatient rate of anxiety and depression is low, and the outcome is controversial. Accordingly, early detection of high-risk groups is especially important to prevent the occurrence of anxiety and depression.Studies on prevention for mental diseases showed that people with certain factors carried higher risk than others, and confirm these factors can create smaller, manageable groups which brings more substantial health gains. How to discover risk group becomes essential. Current studies focus on common multivariate methods such as Logistic regression analysis and use these methods to filter out relative risk factors to discover risk groups. However, epidemiologists consider combined effects of multiple factors to be the determinants of disease occurrence. The same factor’s impact is not consistent across various contexts due to the interplay of various factors. For example, because of the interplay of chronic disease, income, gender, age, and related factors, different combinations affect anxiety differently. For instance, although studies in China show that chronic diseases are associated with anxiety, data from the European Union and United States do not. So it is not critical to filter out risk group based on factors respectively.Individuals are exposed to several different factors that result in interaction effects, affecting disease/illness. Risk group is the category of these people who are exposed to the combination of the riskiest factors. To date, studies employing typical multiple factor analysis methods and linear logistic regression analysis have not conclusively identified nor established the extent of impact of such combinations on anxiety or depression among elderly individuals. A prospective study showed that incidence of anxiety can be significantly decreased when prevention measures were taken on individuals with certain combination of same risk factors.Regression trees can classify the population as continuous to yield various combinations comprising several factors. In the present study, this algorithmic process is represented as a tree model. Regression trees can be used to explore the potentially riskiest combination, which can help identify high-risk groups and engage in focused prevention. Additionally, the impact of a single factor on anxiety can be explained through its interaction effect and interdependencies. This model has been used in other diseases, but report about the model’s apply on anxiety and depression are rarely seen.Subjects comprised aged 60 or older, from four urban communities in Jinan, China. Self-rating Anxiety Scale (SAS) scores and Self-rating Depression Scale (SDS) were used to measure the independent variable-anxiety/depression-upon which the regression tree was built. This study aimed at analyzing combinations of various factors and the extent of each combination’s impact on anxiety and depression, and identifying the potentially riskiest combination in order to inform psychological interventions for anxiety experienced by elderly individuals.Objectives1. Analyze combinations of various factors and the extent of each combination’s impact on anxiety and depression, and identifying the potentially riskiest combination.2. Analyze the effect of the same factor in various combination.MethodsSubjects comprised 1241 community-dwelling individuals, aged 60 or older, from four urban communities with good relationship of cooperation in Jinan, China who satisfied the criteria. Inclusion criteria comprised:(1) being 60 years or older, (2) participating voluntarily and providing informed consent, (3) completing psychological scales, and (4) having permanent residence.1188 subjects have completed all the psychological scales, and 53 subjects only have completed Self-rating Anxiety Scale. Finally, there were 1241 subjects meet the criteria of the study on anxiety, and 1188 subjects meet the criteria of the study on depression. A questionnaire about general characteristics obtained self-reported information about gender, age, marital status, level of education, medical insurance, residence status, source of income, monthly income, number of chronic diseases, and physical functioning (PF). PF is measured using one dimension from the MOS36-Item Short-from Health Survey (SF-36) termed "Physical functioning" (PF). It was used to assess whether subjects’health status interfered with normal physical activity. The SAS scale was adopted for anxiety assessment among elderly individuals, and the SDS scale was adopted for depression. Statistics description was adopted to analyze subjects’ anxiety symptom and depression symptom. The regression tree model was adopted to analyze combinations of various factors and the extent of each combination’s impact on anxiety and depression, and identifying the potentially riskiest combination and analyze the effect of related factors in various combination.Results1. Subjects’ anxiety symptom and depression symptomThe 1241 community-dwelling individuals met the criteria of the study on anxiety who had a mean SAS score of 36.71±6.20.There were 52 participants experienced anxiety symptom, and the prevalence of anxiety was 4.19%. The 1188 community-dwelling individuals met the criteria of the study on depression who had a mean SDS score of 41.23±8.69. There were 203 participants experienced depression symptom, and the prevalence of depression was 17.09%.2. The regression tree model of anxietyAll variables were inserted into the regression tree and six variables (physical functioning, age, gender, disease, income, medical insurance) were selected to constitute 18 combinations. The combination of "PF≤52.5,65≤age<80, and single/widowed" had the highest SAS score (43.0±6.0); the combination of "male, no chronic disease, PF>92.5, and 70≤age<80" had the lowest SAS score (31.6±4.4). The difference between the means was 11.4.3. The regression tree model of anxietyAll variables were inserted into the regression tree and six variables (physical functioning, gender, income, medical insurance, education, age, disease) were selected to constitute 12 combinations. The combination of "PF≤57.5, income<1000 or income>2000, and female" had the highest SDS score (48.5±9.5); the combination of "PF>77.5, education level is primary school or lower, without medical insurance, no disease, income<1000 or income>2000" had the lowest SDS score (37.8±6.7). The difference between the means was 11.4.Conclusions1. Considering that the occurrence of a disease is the result of multiple factors, high-risk groups need to be identified by not only specific isolated risk factors, but by combinations of various factors. Degrees of anxiety or depression are different in various combination.2. One factor’s impact on anxiety is not consistent across combinations. |
PDF Full Text Request