Blood pressure and hypertension prevalence among oldest-old in China for 16 year: based on CLHLS

Background There were little national data on hypertension based on the oldest-old, and lack of information on chronological changes. This study aimed to describe trends of blood pressure (BP) levels and hypertension prevalence for the past 16 years among the oldest-old in China. Methods All the oldest-old who had participated in the Chinese Longitudinal Healthy Longevity Survey (CLHLS) 1998 to 2014 with information about BP levels and hypertension were included in the analysis. Results There was fluctuation over the past 16 years for both SBP and DBP levels. The mean SBP level decreased from 148.4 ± 24.4 mmHg in 1998 to 130.8 ± 18.7 mmHg in 2005, and then increased to 139.7 ± 22.0 mmHg in 2014. The mean DBP level decreased from 84.3 ± 13.4 mmHg in 1998 to 78.9 ± 11.7 mmHg in 2008, and then increased to 79.7 ± 11.8 mmHg in 2014. The hypertension prevalence increased from 43.1 to 56.5% for the 16 years. The prevalence of isolated systolic hypertension was lowest in 2002–2005 (14.3%), and then increased to 30.7% in 2014. Multivariate logistic regression showed that older age, lower education and economic level, without health insurance were associated with higher hypertension prevalence. Conclusions There was a significant increase in hypertension prevalence among the Chinese oldest-old from 1998 to 2014. Greater efforts are needed for hypertension prevention among this specific population.


Background
Hypertension is one of the important risk factors for cardiovascular disease. The higher the blood pressure (BP), the greater the coronary heart disease and stroke risks [1][2][3]. Therefore, it is particularly important to understand the epidemic trend of hypertension. The prevalence of hypertension varies greatly among different age groups, especially among elderly. And isolated systolic hypertension (ISH) (systolic blood pressure (SBP) ≥140 mmHg while diastolic blood pressure (DBP) < 90 mmHg) was most existed in elderly [4,5].
On the other hand, there was a lack of basic data about hypertension prevalence among the oldest-old (aged 80 and over) in China. Previous studies were either among adults, or with small sample, or mainly with inpatients [6,7]. There were little national data on hypertension based on the oldest-old. Additionally, most studies were based on one time survey, the chronological changes had never been reported, which was a reflection of effects about national control measures of hypertension. Therefore, we reported the epidemiology characteristics of hypertension based on 63 thousand oldest-old from seven waves (1998,2002,2002,2005,2008,2011,2014) of Chinese Longitudinal Healthy Longevity Survey (CLHLS), the first and largest longitudinal survey focused on the oldest-old in China [8]. We evaluated the prevalence of hypertension by geography and subpopulations, and the chronological changes.

Study design
All the participants were from the seven waves of CLHLS, and those who aged more than 80 years old with complete records on BP and hypertension information were included. General characteristic of the seven survey waves was listed in Table 3 in Appendix. The details of the CLHLS and sample design have been described elsewhere [8]. The follow-up survey waves were conducted in 2000, 2005, 2008, 2011, and 2014. The use of CLHLS data was approved by the Biomedical Ethics Committee of Peking University, and written informed consent was obtained from each respondent.

Definitions
According to the BP levels, participants were divided into the following groups: normal BP, SBP ≤ 120 mmHg and DBP ≤80 mmHg among those who had never been diagnosed with hypertension; high-normal BP, 120 mmHg<SBP ≤ 139 mmHg or 80 mmHg<DBP ≤ 89 mmHg among those who had never been diagnosed with hypertension; Hypertension, SBP ≥ 140 mmHg or DBP ≥ 90 mmHg or self-reported being diagnosed as hypertension by II&III grade hospital before; ISH was defined as SBP ≥ 140 mmHg and DBP < 90 mmHg regardless of previous hypertension diagnosis history. Mean arterial pressure (MAP) was calculated as the following formula: (SBP+ (2 × DBP)) ÷3. Pulse pressure (PP) was calculated as SBP minus DBP.

Statistical analysis
Mean SBP, DBP, MAP, and PP levels were calculated and expressed as mean ± standard deviation (SD). Variance analysis was used to compare the differences among subgroups. Besides, we estimated the prevalence of hypertension among all the participants. The age and gender adjusted prevalence of hypertension for the first four waves (1998,2000,2002,2005) was calculated using the direct methods based on the fifth Chinese national census data, and the age and gender adjusted prevalence of hypertension for the last three waves (2008,2011,2014) was calculated using the direct methods based on the fifth Chinese national census data. Multivariate logistic regression was used to calculate Odds ratios (ORs) and their 95% confidence intervals (CIs).

Ethical consideration
The use of CLHLS data was approved by the Biomedical Ethics Committee of Peking University.   The prevalence of high-normal BP also showed an increasing trend. It went from 20.9% (95%: 20.0-21.7%) in  (Fig. 2). Sensitivity analysis was performed among those who were first included in the analysis. The BP levels, the prevalence of hypertension were a little higher than the results of the total participants in each survey wave ( Table 9 in Appendix).

Multivariable analysis of hypertension prevalence
Using the most current survey wave data (CLHLS 2014), Table 2 presented the adjusted ORs and 95%CI for the association between covariates in four multivariable logistic regressions each using the following two binary outcomes as dependent variables: (1) prevalence of hypertension; (2) prevalence of ISH. Gender, marriage status had no statistical association with hypertension prevalence. Compared with Han nationality, minorities had lower hypertension prevalence (OR = 0.76, 95%CI: 0.59-0.99). However, only

Discussion
Our study provided the mean BP levels and epidemic picture of hypertension among the oldest-old in China based on CLHLS 1998 to 2014 survey wave. The results indicated that BP levels were high especially DBP and PP levels. Besides, hypertension was of relatively high prevalence, and showed an increasing trend along with seven survey waves over the past 16 years.
The study of BP levels had filled in the blanks for mean BP levels based on community oldest-old in China. And there was fluctuation over the past 16 years for both SBP and DBP levels.. The mean BP levels were significantly higher than those of adults while they were about the same as those of the younger elderly who aged 65-74 years old from the interASIA study in 2000-2001 [9,10]. The above results and the comparison with the domestic and foreign studies indicated that the BP levels of the oldest-old in China was close to those of the younger elderly, which were lower than those of the elderly in the same age group in developing countries. However what cannot be ignored was that the increasing trend of BP levels for the past 16 years, which was worth paying attention to.
The prevalence of hypertension has gradually increased for the past 16 years was consistent with hypertension prevalence and increasing cardiovascular disease burden. And compared with previous studies based adults or younger elderly, the oldest-old had the highest prevalence [11]. Our study also provides data on ISH, which was an important subtype of serious harm among elderly. Data showed that about half of the hypertension oldest-old were classified as ISH subtype, which was higher than the situation in other age groups [12][13][14]. The result from our study showed that there was no less than 30% percent of total hypertension patients could be classified as stage II and above. This suggested that we need to pay attention to the serious situation of high ISH prevalence and high percentage of stage II & III hypertension in the oldest-old and prevent subsequent cardiovascular diseases.
In addition, it is worth noting that not only the hypertension prevalence was increasing, the prevalence of high-normal BP also showed an increasing trend. And evidence showed that high-normal BP was one of the important risk factors of hypertension and cardiovascular diseases [15]. Without enough early prevention measures, those who had high-normal BP will sooner or later develop into hypertension.
We also added information about related factors associated with hypertension (including ISH) prevalence. The gender difference was not significant, which was unlike adults and in line with expectations, since the protective effect of estrogen had disappeared for more than 30 years [16][17][18][19]. For unhealthy lifestyles, those who were ever smoking or alcohol drinking had higher prevalence. This was in line with the phenomenon of quitting smoking or stop drinking because of disease [20]. Obesity was a risk factor for hypertension, just like other studies about adults. Evidence showed that there were a series of endocrine and metabolic changes due to obesity, which might be associated with the hypertension [21].
There were several strengths. First, the CLHLS study was a large scale nationwide study covering 23 provinces of China for the past 16 years. The large sample was unique for representation of the oldest-old. Second, there was good design and strict quality control during the whole survey, which ensures the good quality data.
Our study had several limitations. First, the study sample was from 23 provinces, there was a lack of representativeness for other unselected provinces. Besides, because of the sampling method used in CLHLS, the participants were not representative samples. We calculated both the crude and weighted prevalence, and the trends were similar. Second, most of the participants of CLHLS were from communities, and there was only less than 5% who lived in living in nursing homes or other institutions. But this was in accordance with the situation in China, since more than 95% of the elderly were home-based care. Third, there was less than 3.5% frail oldest-old with severe diseases or disability who didn't have BP recorded,, and this may cause underestimation of prevalence. Fourth, BP levels were measured on the same day. Although the average of two times was use, long-time changes were not captured. Fifth, we didn't have information about treatment and medication information. Although this kind of definition had been adopted in quite a number of epidemiological investigations, the resulting bias cannot be ignored. Combined with the results of previous studies on treatment rates in China and the prevalence of hypertension in this study, this bias may lead to an overestimation. However, taking 2014 survey as an example, there were a total of 299 participants who were classified as hypertension according to previous history with SBP ≤ 140 mmHg and DBP ≤ 90 mmHg this time. And the misclassification participants were calculated as 299-299*30% (the treatment rate according to the China PEACE Million Persons Project) = 209, and the misclassification rate = 209/4587 = 4.6%. this number times. Sixth, only cross-sectional analysis was conducted, and there was lack of the lack of survival data analysis and sensitivity analysis in relation to survival of at least 1-2 years. Seventh, due to the cross-sectional nature, the results about risk factor of prevalence were of low evidence level.

Conclusions
In conclusion, this was the first large scale nationwide study about hypertension with long time comparisons for over 16 years among oldest-old in Chinese. The results also provided evidence about the trends of hypertension prevalence in China, which indicated that hypertension prevention was still a long and arduous task.