2.1 Sampling
Data were collected from March 1st to April 30th 2018 in Anshun city, Guizhou province, China, which was located in western China, and its economy is underdeveloped. Participants were selected using the methods of stratified sampling and cluster sampling. The sampling process involved a systematic approach and a three-step scheme: Firstly, according to geographical location, four districts of Anshun city were divided into two zones: east and west. Secondly, one district was selected from each zone: Xixiu and Kaifa, respectively. Thirdly, according to the method of cluster sampling, we selected some communities with the largest number of parents who lost their only child for the whole survey from each zone. The inclusion criteria were: The parents who lost their only child were equal to or older than 49, in view of the fact that the health bureau has the information of these parents aged at least 49 and mothers who lost their only child at this age would no longer be fertile. In addition, the participants must have normal cognitive functions, and be willing and able to cooperate throughout the investigation process. The exclusion criteria were: The parents who lost their only child had refused to accept the government's condolence and investigation. The parents moved to another place.
One teacher from Zhejiang Chinese Medical University and 15 family planning cadres in local communities were recruited and trained as interviewers. All respondents were clearly informed of the purposes of this study and asked to sign on the consent form. All respondents were assured of their rights to refuse to participate or to withdraw from this study at any time. Anonymity and confidentiality of the participants were assured. Each interview lasted for about 30 minutes. 130 participants were interviewed. There were 108 valid questionnaires, and the effective recovery rate was 83.1%.
2.2. Measurements
2.2.1. GDS
Geriatric Depression Scale (GDS) was developed by the U.S. Brink and Yesavage in 1982, which has 30 items in total. The simplified GDS-15 is a 15-item simplified scale designed by Sheikh and Yesavage in 1986 and is based on the standard version of 30 items according to the characteristics of the senior people[9]. It was used to evaluate their recent depression status, with particular focus on their feeling down, reduced activity, irritability, pain, withdrawal, and negative evaluation of the past, present and future. A higher score on the scale indicates more severe depression[9]. The GDS-15 has been widely used and its reliability and validity have been tested and supported[9]. The GDS-15 scores of the parents who lost their only child were regarded as the dependent variable in this study.
2.2.2. EQ-5D scale
European quality of life-5 dimensions (EQ-5D) scale mainly includes five dimensions: mobility, self-care, daily activities, pain or discomfort, anxiety or depression. Each dimension has three levels of "no difficulty", "some difficulty" and "total difficulty", which are used to evaluate the quality of life of the respondents. Due to the lack of calculation method of China's EQ-5D index score at present, we chose to use the integral conversion table of Japan for calculation, and the scale index score is -0.11~1.00 points. A higher EQ-5D score indicates better health[10]. The EQ-5D scale has high internal consistency and reliability[10].
2.2.3. SSRS
Social Support Rating Scale (SSRS) has 10 items in total, which is divided into three dimensions: subjective support, objective support, and utilization of support. The sum of the scores of the three dimensions is the total score of the scale. A higher score indicates a higher level of social support. A total score of less than 20 indicates less social support, a total score of 20-30 indicates general social support, and a total score of more than 30 indicates satisfactory social support[11]. The SSRS has adequate internal consistency and reliability[11].
2.2.4. Independent variables
The independent variables included: gender (1=male, 2=female, reference value), educational level (1=junior high school and below, 2=senior high school, 3= college and above, reference value), marital status (1=married, 2=not married, reference value), having grandchildren (1=yes, 2=no, reference value), and self-rated health status(1=good, 2=moderate, 3=bad). There were some continuous variables, including age, monthly income, number of chronic diseases, EQ-5D score, SSRS score.
The number of chronic diseases was measured by the multiple-choice question, "How many chronic diseases do you have?" Sixteen chronic diseases were listed for selection, including diabetes, hypertension, hyperlipidemia, the malignant tumor, Cerebral infarction (stroke), cerebrovascular disease, coronary heart disease, senile dementia, chronic liver disease, gout, asthma, gynaecology disease, arthritis, tuberculosis (TB), hematopathy, chronic low back pain, osteoporosis, cataract. A higher score meant that participants had more chronic diseases. However, self-rated health status was regarded as continuous variables in the generalized linear regression model.
2.3. Quality control
The respondents’ ages were confirmed using the household registration system. During the face-to-face field survey, the trained teachers and family planning cadres in local communities explained how to fill in the questionnaires and helped respondents complete them in their homes or neighborhood committee office. The database was established by EpiData3.1, and double input was conducted to ensure accuracy.
2.4. Data analysis
Data were analyzed using the SAS version 9.1 software. Participant sociodemographic variables were expressed as frequencies. In view of the non-normal distribution of GDS score data, the rank sum test was used to compare the GDS score of the parents who lost their only child with different basic characteristics. The generalized linear univariate and multivariate regression models were conducted to analyze factors influencing GDS score.