Page 27 - IJPS-1-1

Yasuhiko Saito, Shieva Davarian, Atsuhiko Takahashi, Edward Schneider and Eileen M. Crimmins

2. Data and Methods

2.1 Data Sources
This study used data from the 2006 (fourth wave) of the Nihon University Japanese Lon-
gitudinal Study of Aging (NUJLSOA), and data from the 2006 Health and Retirement
Study (HRS). The NUJLSOA is a nationally representative sample of the Japanese popula-
tion aged 65 and older at baseline collected by researchers at Nihon University. The fourth
wave contains data from participants aged 68 years and older. Complete information on
the design and methods of the study can be found on the website of the survey (Nihon
University Japanese Longitudinal Study of Aging).
The fourth wave of the survey includes 3,403 participants, of whom approximately 68%
participated in the collection of blood pressure measurements (n = 2,330). Our sample
consists of the 2,309 cases for whom both measured and self-reports of hypertension are
available. The average age of the sample was 76.1 years, of whom 54.9% were female.
The participants who were missing on measured blood pressure differed from those in-
cluded in the analysis in that they were older (1.01 years, p<0.01) and had lower rates of
antihypertensive medication usage (9%, p<0.01).
The Health and Retirement Study (HRS) surveys more than 22,000 Americans over the
age of 50 every two years. Data for the eighth wave of HRS used in this study were col-
lected in 2006, the same time as the fourth wave of NUJSLOA. A detailed description of
the study is provided at a website hosted by the University of Michigan. Half of the HRS
2006 sample was randomly chosen to participate in an enhanced face-to-face interview
where blood pressure was measured. Respondents who used a proxy or who resided in a
nursing home did not have measured blood pressure. The total number of the respondents
who participated in the enhanced face-to-face interviews and who were 68 years of age
and over was 4,730; of these 1,137 did not have blood pressure measurements and were
not included in this study. The sample of 3,517 with information on both measured and
self-reported hypertension used for this study was quite similar in composition to the Jap-
anese sample and included 58.8% females and had an average age of 76.7. Relative to
those included in the analysis, the participants who were missing on blood pressure mea-
surements were older (3.21 years, p<0.01). There was no significant difference in antihy-
pertensive medication usage (p = 0.97) between those included in the study and those who
were missing on blood pressure measurements.

2.2 Measurements
Protocols for blood pressure recordings were almost identical between HRS and NUJL-
SOA except that the measurements were taken during the midst of an interview for HRS
and at the end of an interview for NUJLSOA. The blood pressure monitor used in NUJL-
SOA was OMRON HEM-762 (OMRON, Kyoto, Japan) and for HRS it was OMRON
HEM-780N (OMRON, Kyoto, Japan); the device has been validated in independent stu-
dies (Shirasaki, Asou and Takahashi, 2007). Blood pressure was measured three times with
approximately one minute between each measurement and the average of the available
readings was used for analysis. If fewer than three readings were recorded, then the aver-
age of those readings was included. The participant was generally in a seated position and
the recording was generally performed on the left arm for both surveys.
Measured hypertension is a binary variable that indicates a person had an average sys-
tolic measurement of 140 mmHg or greater and/or a diastolic measurement of 90 mmHg
or greater. Use of antihypertensive medication was reported in the same way in both sur-
veys. Respondents who responded affirmatively to ever having been diagnosed or having

International Journal of Population Studies | 2015, Volume 1, Issue 1 21