The study population comprised 19,231 subjects aged 40–74 years in seven cities in Yamagata Prefecture (Yamagata, Sakata, Kaminoyama, Tendo, Sagae, Higashine, and Yonezawa) who consented to the Yamagata Study during the baseline survey between 2009 and 2015. Details of the Yamagata study have been described in a previous study . The Yamagata Study was a community-based prospective cohort study, supported by the regional characteristics of the twenty-first Century program and Global Center of Excellence programs in Japan. The target population of this study was the local population covered by national health insurance. All subjects in this study provided written informed consent to participate. Of the 19,231 subjects who consented to the study, 14,264 subjects were included in the analysis after excluding those who withdrew their consent and those with missing responses to the questions on yogurt/milk intake and baseline characteristics (n = 3387) and the subjects with history of cancer or cardiovascular diseases (stroke, myocardial infarction and angina pectoris) (n = 1580). This study was approved by the Ethics Review Committee of Yamagata University Faculty of Medicine (Approval No. 2019–391), and it was conducted in accordance with the Declaration of Helsinki.
Information on yogurt and milk intake and baseline characteristics
We collected information on yogurt and milk intake at the baseline survey using the short food frequency questionnaire (FFQ) with high reproducibility and validity reported . In the questionnaire frequency of yogurt and milk intake was assessed via a single-item question: “How often do you have yogurt?” and “How often do you drink milk?”. We provided six possible answers, namely, 1 or more times/day, 5–6 times/week, 3–4 times/week, 1–2 times/week, 1–3 times/month, and < 1 time/month. Then, we categorized the frequency of yogurt and milk into four groups as none (< 1/month), low (1–3/month), moderate (1–6/week), and high (> 1/day) intake.
Data on baseline characteristics, such as age, sex, current smoking status, alcohol consumption, education period and various laboratory parameters, were collected. The education period was categorized in to three groups (< 9 years: primary or junior high school, 10–12 years: high school, and > 13 years: college or higher). We ascertained the presence of hypertension, diabetes, and dyslipidemia using the information on laboratory data and medications according to the definitions used in previous studies [12, 14].
All subjects were followed up from 2009 to 2018 (9 years). Information on mortality was obtained from the certificate of residence. Information of the cause of death was collected from death certificate between 2009 and 2016 (7 years), and classified based on the International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD-10) codes. Cardiovascular mortality was defined as death due to circulatory system disorder (ICD-10 codes I00–I99), and cancer-related mortality was defined as death due to cancer (ICD-10 codes C00–D48).
Analysis of variance and chi-square tests were used for continuous and categorical variables, respectively, to compare the groups. Survival time analysis was performed using the Kaplan–Meier method for the association between yogurt/milk intake and mortality. To examine this association, Cox proportional hazards analysis was performed, adjusting for background factors including age, sex, smoking status, alcohol consumption, body mass index (BMI), hypertension, diabetes, and education period. The significance level for each test was set at P < 0.05. All statistical analyses were preformed using JMP Software 14.2 for Windows (SAS Institute Japan Ltd., Tokyo, Japan).