Study Design and Setting
This was a cross-sectional study with retrospective recall using data from a Japanese study on stratification, health, income, and neighborhood (J-SHINE) survey (30). The J-SHINE survey is a prospective cohort study aimed at clarifying the complex associations between social factors and various health policy evaluations. The J-SHINE survey randomly sampled adult community residents from a registry in each of four municipalities (two in Tokyo and two in neighboring prefectures). The Wave 1 survey was from July 2010 to February 2011, the Wave 2 survey from August 2011 to December 2011, and the Wave 3 survey in 2017. We used data regarding demographic variables and outcome variables (i.e., meaning in life, life satisfaction, and subjective health status) in the Wave 1 survey and data of personal values at the age of 15 in the Wave 3 survey. The study protocol and informed consent procedure were approved by the ethics committee of the Graduate School of Medicine of The University of Tokyo [No.630–7,3361].
Participants
Study participants were adult community residents recruited in the J-SHINE survey. Eligibility criteria of the study were adult community inhabitants aged 20 to 50 years old who were selected from the registry in four municipalities (two in the Tokyo metropolitan area and two in neighboring prefectures). Systematic sampling was conducted for each unit. Except for age, there were no inclusion and exclusion criteria. Trained surveyors visited the home of the inhabitants after sending an invitation letter. If they agreed to participate in the study, the participants were asked to provide informed consent and answered the self-report questionnaire by using a computer-aided personal instrument (CAPI). For those who were unfamiliar with computer use, a personal interview with the CAPI was provided.
Measurements
All variables in this study (personal values at the age of 15, meaning in life, life satisfaction, subjective health status, and covariates) were measured by self-report questionnaires.
Personal values at the age of 15 were measured by two components: value priorities and degree of commitment to the values. For measurement of the value priorities, we used a list of 11 values (31, 32), which were developed based on the 57-item Protrait Values Questionnaire (PVQ–57) (33): not bothering others, being evaluated by others, having and keeping a belief, economically succeeding, improving society, exploring what you were interested in, having influence on society, actively challenging, cherishing familiar people, graduating from a famous school, and maintain a stable life. Respondents were asked to recall their value priorities at the age of 15 (“When you were 15–16 years old, how important did you think the following values were in your life?) and rate for each on a seven-point Likert scale (1 = Not at all, 7 = Very important). Commitment to the values was measured by the Personal Values Questionnaire Ⅱ (PVQ-Ⅱ) (34). The original version consists of nine items, but for the Japanese version of PVQ-Ⅱ, one item was dropped. Each item was rated on a five-point Likert scale. The internal consistency, concurrent, and structural validity were reported both for the English and Japanese versions of the PVQ-Ⅱ (34). In this study, we modified the PVQ-Ⅱ questions to the past tense and asked participants to rate their commitment to a value that was most important for them at the age of 15. Internal consistency and factor-based and construct validity were already found to be acceptable (Watanabe et al., under review). The total score was calculated, and higher ratings indicate more commitment to the important values.
Meaning in life was measured by the Meaning in Life Questionnaire (MLQ) (6). The MLQ consists of two five-item subscales; presence of meaning in life and search for meaning in life. All items were rated on a seven-point Likert scale ranging from 1 (Strongly disagree) to 7 (Strongly agree). We calculated the total score of each subscale and used this for analyses. Higher scores indicate a higher meaning in life. The original English scale was translated into Japanese and preliminarily tested, and it proved to have an acceptable level of internal consistency reliability and content validity (6).
Life satisfaction was also measured by one item: “How satisfied are you with your current life in general?” The item was rated on a five-point Likert scale (1 = Satisfied, 5 = Dissatisfied). This question was ab item from the national survey of Social Stratification and Social mobility (SSM) (35). The score of life satisfaction was reverse coded, with a higher score meaning a more satisfying life.
Subjective health status was measured by one item, “How is your current health condition?” The item was developed in a national survey, the Comprehensive Survey of Living Conditions. This survey was conducted to investigate basic matters such as health, medical care, welfare, pension, income, etc., by the Household Statistics Office, the Ministry of Health, Labour and Welfare, Japan (Ministry of Health, Labour and Welfare online (36). The item was rated on a five-point Likert scale (1 = Good, 5 = Not good). The score for subjective health status was reverse coded, with a higher score meaning a better health condition.
As the covariates, demographic variables, and one’s circumstances at the age of 15 were measured. The demographic variables included gender (male [reference], female), age, educational status (junior high school [reference], senior high school, professional school, university), marital status (not married [reference], married), work status (not working [reference], working, absence), household income (Japanese yen) (< 3 million [reference], 3 ~ 5 million, 5 ~ 7.5 million, > 7.5 million, not known). One’s circumstances at the age of 15 consisted of one item, rated on a 5-point Likert scale ranging from 1 (very poor) to 5 (very rich).
Statistical analysis
Participants who completed all the variables used in the study were included in the analyses. First, descriptive statistics among the completers were calculated. For the main analysis, multiple linear regression analysis was done to investigate the association between personal values at the age of 15 and meaning in life, life satisfaction, and subjective health status. The analysis estimated both crude associations and adjusted associations for the covariates. For the categorical covariates (gender, educational status, marital status, work status, and household income), dummy variables were created and entered into the models. Statistical analysis was done with SPSS (IBM SPSS Statistics for Windows, version 22.0 Armonk, NY, USA).