The National Integrated Project for Prospective Observation of Non-communicable Disease and its Trends in the Aged 2010 (NIPPON DATA2010): Objectives, Design, and Population Characteristics

Background The structure and risk factors for cardiovascular diseases (CVD) in Japan may change because lifestyle, particularly nutrition, socioeconomic status, and medical care, which affect CVD, may markedly change over time. Therefore, a new prospective cohort study on a representative general Japanese population based on national surveys is required. Methods In November 2010, the baseline survey of the National Integrated Project for Prospective Observation of Non-communicable Disease and its Trends in the Aged 2010 (NIPPON DATA2010) was performed with the National Health and Nutrition Survey of Japan (NHNS2010) in 300 randomly selected districts throughout Japan. The survey included a questionnaire, electrocardiogram, urinalysis, and blood biomarkers added to the NHNS2010 examinations. Physical measurements, blood biomarkers, and dietary data were also obtained in NHNS2010. Socioeconomic factors were obtained by merging with the Comprehensive Survey of Living Conditions 2010 (CSLC2010) dataset. Participants are followed annually for the incidence of diabetes mellitus, CVD events (acute coronary events, heart failure, atrial fibrillation, and stroke), and cause-specific mortality. The activities of daily living are followed every 5 years. Results A total of 2,898 individuals aged 20 years or older agreed to participate in the baseline survey of NIPPON DATA2010. The participation rate was 74.6%. Of these, data from NHNS2010 was merged for 2,891 participants (1,236 men and 1,655 women). The data of 2,807 participants were also merged with CSLC2010 data. Conclusions We established NIPPON DATA2010 as a cohort study on a representative general Japanese population that covers all of Japan.

conditions, the expansion of health inequality has been reported. 13 Although the social determinants of health inequality need to be identified in order to prompt political action, the effects of SES on health have not yet been fully investigated in Japan. 14,15 Therefore, a new cohort study on a representative general Japanese population based on national surveys that cover all of Japan with standardized methods and enable assessments of health-related factors from multiple aspects, including SES, is required. We herein describe the objectives, design, and characteristics of a new cohort study, NIPPON DATA2010, which started in 2010 with the purpose of monitoring and revealing factors related to CVD in a recent Japanese population.

Study participants
NIPPON DATA2010 was established as a prospective cohort study of participants of the National Health and Nutrition Survey of Japan in 2010 (NHNS2010) and Comprehensive Survey of Living Conditions in 2010 (CSLC2010), which were conducted by the Ministry of Health, Labour and Welfare of Japan ( Figure 1).

CSLC2010
In June 2010, CSLC was conducted in order to survey national living conditions, such as health, medical care, welfare, pension, income, and related factors. 16 Sampling was based on the national population census enumeration area (CEA), which covered all of Japan. Each CEA consisted of approximately 50 households. In the 2010 survey, approximately 290,000 households from 5,510 randomly selected CEAs were sampled for the Household Survey and Health Status Survey. The response rate was 79.4%. Among these CEAs, the Saving Survey and Income Survey covered 2,500 randomly selected unit blocks; each unit block consisted of 20-30 households, approximately half of which were in the CEA. The Nursing Care Survey also covered 2,500 unit blocks that partially overlapped. Detailed information on CSLC was described elsewhere. 16

NHNS2010
Among the 5,510 CEAs (11,000 unit blocks) for CSLC2010, 300 unit blocks from throughout Japan were randomly selected for NHNS2010. All household members who resided in the blocks and participated in CSLC2010 were announced to participate in NHNS2010 in November 2010. A total of 8,815 residents aged 1 year and older from 3,684 households participated in the survey ( Figure 1). The participation rate by households was 68.8%. Adult participants aged 20 years or older were also asked to take lifestyle questionnaires and blood examinations for NHNS2010, in addition to the dietary survey and physical examination, with 7,881 (2,672 men and 4,209 women) taking the lifestyle questionnaire and 3,873 (1,598 men and 2,275 women) taking the blood examination for NHNS2010. Details of NHNS2010 were described elsewhere. 17,18

NIPPON DATA2010
The baseline survey for NIPPON DATA2010 was performed simultaneously with NHNS2010 in November 2010. 19,20 Trained staff explained the aim and methods of NIPPON DATA2010 to the 3,873 participants aged 20 years or older who underwent the blood examination for NHNS2010 at the NHNS2010 places such as public health centers. A total of 2,898 participants (1,239 men and 1,659 women; participant rate, 74.6%) agreed to participate in the baseline survey for NIPPON DATA2010. Staff obtained written informed consent from all participants before enrollment. The Institutional Review Board of Shiga University of Medical Science (No. [22][23][24][25][26][27][28][29]2010) approved this study.
Of the 2,898 participants, 7 were excluded because it was not possible to merge data from NHNS2010. Thus, the remaining 2,891 participants (1,236 men and 1,655 women) provided baseline data for NIPPON DATA2010. In the analysis of socioeconomic factors, 2,807 participants who were also merged to CSLC2010 were listed. Regarding the follow-up survey, 2,732 participants (1,170 men and 1,562 women) agreed to be included in the study.

Measures
The measures of NIPPON DATA2010 were composed of three parts: NHNS2010, CSLC2010, and specific measures in NIPPON DATA2010. The outline of the integrated dataset, variables, and their original survey are listed in Table 1. Health professionals for NHNS2010 and trained staff for NIPPON DATA2010 collected information on smoking, alcohol consumption, and medical history. Lifestyle-related factors, including knowledge of CVD risk factors, were asked using self-administered questionnaires. Regarding physical activity, the time (hours) spent at each activity level was also asked: heavy activity, moderate activity, slight activity, watching television, other sedentary time, and no activity (sleeping). The interviewer then ensured that the total time added up to 24 hours, and the physical activity index was calculated by multiplying the time and corresponding weighting factor estimated in the Framingham study. 21 The activities of daily living (ADL) were also asked for five aspects: eating, using the toilet, dressing, bathing or taking a shower, and walking, with answers of "independent" or "need assistance". 22 Information on instrumental activities of daily living (IADL), intellectual activities, and social roles was also obtained based on the Tokyo Metropolitan Institute of Gerontology Index of Competence. 23 Socioeconomic factors, such as household composition and monthly household expenditure, were obtained from CSLC2010 with the permission of the Ministry of Health, Labour and Welfare. Equivalent household expenditure (EHE) was estimated using the following formula: EHE = monthly household ex-penditure=square root of the number of household members. Information about medical insurance and pension was also obtained from CSLC2010. The dietary intake data of NHNS2010, which was assessed by 1-day semi-weighted household dietary records, were also obtained with the allowance of the Ministry of Health, Labour and Welfare. The detailed procedure for the dietary survey was described elsewhere. 17,18 Physical measurements were obtained by trained health professionals. They measured blood pressure in duplicate using a standard mercury sphygmomanometer on the right arm of seated participants after 5 minutes of rest. A 12-lead resting electrocardiogram (ECG) was also recorded. Each ECG record was manually read according to the Minnesota codes by two trained researchers independently. 24,25 When the coding results mismatched, the central committee of ECG reading adjudicated the codes.
In the baseline survey, casual blood samples were obtained. Serum was separated and centrifuged soon after blood coagulation. Plasma samples were collected into siliconized tubes containing sodium fluoride and shipped to a central laboratory (SRL, Tokyo, Japan) for blood measurements. Serum triglycerides (TG), total cholesterol (TC), low-density lipoprotein (LDL) cholesterol, and high-density lipoprotein (HDL) cholesterol were measured using enzymatic methods, which have been standardized by the Lipid Standardization Program of the US Centers for Disease Control and Prevention=Cholesterol Reference Method Laboratory Network (CDC=CRMLN). 26 Blood glucose was measured using hexokinase UV methods, and hemoglobin A1c (HbA1c) was measured using latex agglutination inhibition assays with the standardized method of the Japan Diabetes Society (JDS). HbA1c values were converted into the National Glycohemoglobin Standardization Program (NGSP) value using the following formula: HbA1c (NGSP) (%) = 1.02 × HbA1c (JDS) (%) + 0.25. 27 Serum creatinine was measured enzymatically. High-sensitivity C-reactive protein (CRP) was measured using nephelometry methods, and brain natriuretic peptide (BNP) was measured via CLEIA using MI02 Shionogi BNP (Shionogi Co. Ltd., Osaka, Japan). Information on blood chemistry data measurements and their performance was described elsewhere. 26 Spot urine samples were also collected and shipped to the same laboratory at which blood measurements were examined. Urine creatinine was measured enzymatically. Urine sodium and potassium were measured using selective ion electrode methods. Urine albumin and protein were measured using immunonephelometry and pyrogallol red methods.

Follow-up
As the first step, the incidence of stroke, heart disease, and diabetes mellitus is surveyed annually from participants using the self-administered questionnaire via mail or telephone interviews. In the second step, when the incidence of these diseases is reported by participants or their family members, detailed medical records will be referred to the hospitals from the central office of the NIPPON DATA Research Group. The incidence of these diseases will then be assessed at the event adjudication committee of the study group. Information on medication for hypertension, dyslipidemia, and diabetes mellitus is also collected annually. The ADL and IADL survey is also performed every 5 years using the self-administered questionnaire.
Participants who die during the follow-up are confirmed by computer matching with the National Vital Statistics database, using area, sex, date of birth, and date of death as key codes, with the permission of the Management and Coordination Agency of the Government of Japan. The underlying causes of death in the National Vital Statistics database are coded according to the 10 th International Classification of Disease (ICD-10). Details of these classifications have been described elsewhere. 28

Main outcome measures
The study main outcome measures are listed in Table 2. All diagnoses will be based on medical records obtained from the hospitals. Each case is independently diagnosed by two trained medical doctors. When their diagnoses are mismatched, the committee adjudicates. The diagnostic criteria of main outcome events are described as follows.

Heart disease
Myocardial infarction is diagnosed using the modified MONICA criteria or third universal definition of myocardial infarction by ESC=ACCF=AHA=WHF. 29,30 Invasive procedures for coronary arteries, such as coronary artery bypass grafting (CABG) and percutaneous coronary intervention (PCI), are also considered as outcome events. Heart failure is diagnosed according to the Framingham Heart Study criteria. 31 Researchers diagnose the case, taking into account symptoms and laboratory tests when the components of criteria are not fully collected. The acute exacerbation of chronic heart failure is also considered to be an outcome event when the case meets the Framingham criteria and requires hospitalization. Regarding arrhythmia, atrial fibrillation is diagnosed based on ECG findings. Sick sinus syndrome, atrioventricular block, and other arrhythmias are considered to be outcome events when the participant requires the insertion of a cardiac pacemaker.

Stroke
Stroke is diagnosed with neurological symptoms that continue for 24 hours or longer. 32 Secondary stroke due to injuries, leukemia, and tumors is excluded from outcome events. The stroke subtype is diagnosed based on imaging findings.

Diabetes mellitus
Diabetes mellitus is diagnosed according to the modified criteria of the JDS as follows: 1) fasting blood glucose 126 mg=dL or higher, 2) casual blood glucose 200 mg=dL or higher, 3) HbA1c 6.5% or higher, and=or 4) medication for diabetes mellitus. 33 A case that meets any one of these criterion items is considered to be an incident case of diabetes mellitus.

Baseline descriptive statistics
The baseline characteristics of participants are shown in Table 3.

Strengths and limitations
We established NIPPON DATA2010 as a cohort study on a representative general Japanese population. Because the extraction method of study participants and ECG coding method were the same as in the NSCD, which were performed recently in 2000, we could consider NIPPON DATA2010 as the successor survey of NSCD. The study obtained not only physical measures, but also lifestyle factors, diet (NHNS2010), and socioeconomic factors (CSLC2010), and data collection and measurement methods were highly standardized. 26 Thus, we will provide important information from multiple aspects for future strategies for CVD prevention and management in Japan. The integration method we adopted to use data from CSLC2010 may be applicable to NIPPON DATA80 and NIPPON DATA90, which will enable us to investigate the effects of changes in SES on health inequality from 1980.
Several limitations need to be noted for this study. The reporting bias for the measures obtained using the self-  34,35 They also showed that sociodemographic factors and lifestyle, such as being active, were related to cooperation for blood testing in NHNS. However, due to the lack of other cohort studies based on recent national surveys using a cluster random sampling design in Japan, NIPPON DATA2010 is considered to be the best available cohort that represents a recent Japanese population from all over Japan.

ACKNOWLEDGMENTS
We deeply appreciate the Japanese Association of Public Health Center Directors and the Japan Medical Association for their support with NIPPON DATA2010's baseline and follow-up survey. We also appreciate Shionogi Co. Ltd. for their support measuring brain natriuretic peptide. The authors thank Japanese public health centers and medical examination institutions listed in the Appendix for their support with NIPPON DATA2010's baseline survey.
Conflicts of interest: This study was provided reagents for measuring BNP from Shionogi Co. Ltd.  Co-morbidities were defined as follows: Hypertension (SBP 140 mm Hg and higher and=or DBP 90 mm Hg and higher and=or on medication), diabetes mellitus (fasting blood glucose 126 mg=dL and higher and=or non-fasting blood glucose 200 mg=dL and higher and=or HbA1c 6.5% and higher and=or on medication. Samples obtained more than 8 hours after the last meal were considered to be fasting blood samples), hypercholesterolemia (LDL 140 mg=dL and higher and=or medication), hypertriglyceridemia (TG 150 mg=dL and higher and=or on medication), low HDL cholesterolemia (HDL-cholesterol less than 40 mg=dL), dyslipidemia (hypercholesterolemia and=or hypertriglyceridemia and=or low HDL cholesterolemia). A history of stroke and myocardial infarction was based on self-reports. JPY, Japanese Yen. Data were expressed as mean (standard deviation [SD]) for age, as a median (25%-75%) for equivalent household expenditure (EHE), and in % for dichotomous variables. P values for differences between men and women were estimated by t-tests for continuous variables, Mann Whitney tests for EHE, and chi-squared tests for dichotomous variables. Working status was categorized as follows: unemployed (unemployed, full-time homekeeper or retiree), self-employed (self-employed or family employee), indefinite-term employee (exective, indefinite-term employee or limited-term employee with 1 year and longer), limited-term employee (limited-term employee shorter than 1year). EHE was estimated using the following formula: EHE = monthly household expenditure=square root of the number of household members.