Hyperuricemia is associated with cardiovascular diseases clustering among very elderly women - a community based study in Chengdu, China

Cardiovascular epidemiological features among very elderly Chinese are still uncertain. This study aimed to describe the distribution of cardiovascular diseases and sex difference, and investigate potential risk factors for diseases clustering among very elderly Chinese. From May 2013 to May 2015, a total of 1056 very elderly were sampled in this cross-sectional study. Demographic characteristics collection, physical examination and biochemical analysis were performed. Totally, 1038 participants (men: 49.8%) with a median age of 83.0 years (age range: 80.0–100.0 years) were included. In this very elderly group, the prevalences of hypertension, diabetes mellitus, renal dysfunction, hyperuricemia, obesity, visceral obesity, and hypercholesterolaemia were 76.9%, 20.0%, 23.3%, 33.0%, 9.4%, 54.5% and 35.8%, respectively. About 17.5% of very elderly (men vs. women: 15.1% vs 19.8%, p = 0.007) have ≥3 cardiovascular diseases clustering. Logistic analysis found that hyperuricemia (odds ratio 3.850, 95%CI 2.189–6.770) was associated with of cardiovascular diseases clustering among very elderly women. Prevalences of prehypertension, hyperuricemia, visceral obesity and dyslipidaemia are apparent in very elderly women, while diabetes mellitus, renal dysfunction are common in very elderly men. Women are more likely to have ≥3 cardiovascular diseases. Hyperuricemia is associated with cardiovascular diseases clustering among very elderly women.


Results
Baseline characteristics. Totally, 1056 participants were enrolled in this survey. The overall response rate was 92.6%. Finally, 18 participants with incomplete data were excluded and therefore 1038 participants (men vs women: 49.8% vs 50.2%) were included in final analysis. The median age was 83.0 years (lower quartile: 81.0, upper quartile: 85.0, age range: 80.0-100.0 years), in which men were slightly older than women (p = 0.342). Among all participants, more than 10% were current smokers and 8% of them were current alcohol drinker. Not surprisingly, both proportions in men were much significantly higher than in women (both p < 0.001). Nearly 70% of these men were educated in a primary school or a middle/high school. While, more than 30% of women were illiterate compared with 11% in men (p < 0.001). Moreover, the proportion of women educated in middle/ high school and college/university were significantly lower than which in men (both p < 0.001) ( Table 1).
Estimated prevalence of cardiovascular abnormalities. Estimated prevalences of pre-hypertension and hypertension among very elderly participants were 16.6% and 76.9%, respectively. While, both prevalences were not significantly different between sexes, although both of them were a bit higher in very elderly women (both p > 0.05). The estimated prevalence of DM was about 20.0% among overall very elderly participants. The prevalence of DM among very elderly men did not differ significantly from women (21.4% vs 18.8%, p = 0.285). The estimated prevalence of renal dysfunction among the very elderly participants was 23.3%, which was doubled in men than in women (32.6% vs 14.1%, p < 0.001). As well, we could not ignore that the prevalence of hyperuricemia among very elderly was as high as 33.0%, which was even higher in women than that in men (35.6% vs 30.2%, p > 0.05). We also presented the estimated prevalence of renal dysfunction (eGFR < 60 ml/(min · 1.73 m 2 )) according to Cockcroft-Gault equation in Table 2.
Compared with obesity, overweight and visceral obesity were more common in very elderly with a prevalence of 30.0% and 54.5%, respectively. Especially in very elderly women, the prevalence of visceral obesity was significantly higher than in men (62.7% vs 45.1%, p < 0.001).
As among the very elderly, estimated prevalences of persistent atrial fibrillation and three degrees atrial ventricular block were also higher in men than in women, while not significantly (all p > 0.05).  The estimated prevalence of hypercholesterolaemia in very elderly was one and half times as the prevalence of hypertrigleicemiea (35.8% vs 21.6%). And both prevalences were significantly higher in women than in men (both p < 0.05) ( Table 2).

Cluster of cardiovascular diseases. Among participants with DM, the prevalence of comorbid visceral
Obesity was 63.2%, while among participants with hypertension, the prevalence of comorbid visceral Obesity was 57.3%, which was even lower (p = 0.129). Among participants with both hypertension and DM, 54.1% of them had visceral Obesity and 27.3% of them had renal dysfunction. Prevalences of hypertrigleicemiea and hyperuricemia among participants with DM were higher than which among participants with hypertension, while the prevalence of hypercholesterolaemia was lower. Moreover, in participants with both hypertension and DM, prevalences of all other cardiovascular abnormalities were lower than in participants separately with hypertension or DM, except for renal function (Table 3).
Receiver-operating characteristic (ROC) analysis found that the optimal cutoff point for UA predicting cardiovascular diseases clustering was 352.5 μmol/L among very elderly women (p < 0.001) (Fig. 1). The sensitivity and specificity of UA ≥ 352.5 μmol/L for predicting AF were 66.7% and 67.3% among very elderly women.

Discussion
To our knowledge, this survey is the first community population-based study providing general epidemiological information across wide spectrum of very elderly Chinese in Chengdu. This survey demonstrates as follows: (1) Cardiovascular risk factors are common among very elderly. (2) Sex differences of risk factors are apparent among very elderly, and women are more likely to cluster ≥3 diseases. (3) Hyperuricemia is associated with cardiovascular diseases clustering among very elderly women.
According to the newest report by World Health Organization (WHO), the life expectancy at birth for Chinese in 2015 is 76.1 years 9 . The average age of participants in this study is more than 83 years. Although much of the evidence refers to young and middle-aged adults, older people currently account for most cardiovascular disease morbidity and mortality. Our study shows that burdens of cardiovascular diseases in very elderly are even worse than in other age group or in general population 7 . Hypertension, DM, renal dysfunction, obesity, and lipid abnormalities are currently not rare in very elderly residents. Especially prevalences of some of them are relatively high. It is already reported that among very elderly American the incidence of dementia is higher than that of coronary heart disease, although both diseases are complications of cardiovascular and metabolic diseases 10 . A recent study in Shanghai community elderly and very elderly residents, which included about 600 very elderly, reported similar cardiovascular and metabolic parameter levels (SBP, DBP, FG, TG, and TC) and prevalences of hypertension and DM compared with results in our study 11 .
Based on results of this survey, hypertension is one of the most predominant health problems in very elderly Chinese in Chengdu. The proportion of very elderly with normal blood pressure (BP) is only about 6.5%. In another word, most population would suffer from hypertension when they are older than 80 years. The mean level     aged 20-39 years to 21.6% in population aged 65 years and older 12 . It is already known that total DM prevalence in China rises from 3.2% in population aged 20-39 years to more than 21% in population aged 70 or older 13 . More than one quarter Chinese adults have two cardiovascular risk factors (i.e. smoking, overweight, hypertension, dyslipidaemia, or hyperglycaemia), and nearly one fifth adults have three risk factors 7 . In this survey, it is showed that more than 17.5% very elderly residents have ≥3 cardiovascular risk factors. Therefore, there is no doubt that the burden of cardiovascular diseases increases gradually with aging. Compared with our previous survey in 2008, current study finds that total smoking rate in very elderly is only half of which in middle aged and elderly population (aged 40-79 years, 10.7% vs 23.7%, p < 0.001). Moreover, that difference is mainly from men (18.0% vs 54.3%, p < 0.001) while not from women (4.4% vs 4.1%, p = 0.737) 14 . Furthermore, prevalences of overweight (30.0% vs 35.1%, p = 0.001), obesity (9.4% vs 10.6%, p = 0.262) in very elderly are lower than which in middle aged and elderly population, while the prevalence of visceral obesity (54.5% vs 29.3%, p < 0.001) is much higher, which is more obviously in women. Surprisingly, the prevalence of hypertension (76.9% vs 29.4%, p < 0.001) was obviously higher in very elderly than in middle aged and elderly population. Both in very elderly men (75.6% vs 31.0%, p < 0.001) and women (78% vs 27.8%, p < 0.001), the prevalence of hypertension is two times than in middle aged and elderly population. Moreover, in very elderly, there is no obvious decrease in prevalence of DM (20.0% vs 11.3%, p < 0.001). In contrast, the prevalence of hyperuricemia increases apparently (33.0% vs 11.9%, p < 0.001). Besides, prevalences of hypercholesterolemia (35.8% vs 24.7%, p < 0.001) and hyper-low density lipoproteinemia (14.9% vs 12.0%, p = 0.009) are even higher in the elderly Chinese, while the prevalence of hypertrigleicemiea (21.6% vs 31.4%, p < 0.001) is lower. Therefore, cardiovascular risk factors distribute differently among very elderly and middle aged and elderly Chinese.
Sex differences in cardiovascular diseases are common in very elderly. Our survey demonstrates that not only prevalences of hypertension, visceral Obesity and obesity, but also prevalences of hypercholesterolaemia, hypertrigleicemiea, and hyper-low density lipoproteinemia in very elderly women are higher than in men. It seems that these abnormalities are more likely to cluster in very elderly women while not in men. Nearly one quarter of the total global disease burden is attributable to diseases in people aged 60 years and older, and half of them is in high income countries and a fifth is in low-income and middle-income countries 15 . Although China remains a developing country with the fastest acceleration in urbanization, industrialization, the burden of disease increases dramatically fast. A meta-analysis included 15 trials found that in people aged 60 years or older, 5 year event rates per 1000 people were reduced by 18 for all deaths, by 19 for cardiovascular deaths, and by 51 for combined cardiovascular morbidity and mortality by diuretics or β blockers 16 . Another meta-analysis with 14 randomized trials found that benefits from statins based cholesterol-lowering treatment were apparent for patients aged 75 years and older, with risk reduction of major vascular and coronary events 17 . Since hypertension, DM, dyslipidemia, and obesity are major modifiable risk factors for cardiovascular diseases, benefits of risk factor modification in older people are clear for control of hypertension and hypercholesterolemia.
It has been reported that UA could predict mortality in patients with heart failure 18 and the increase of incidence of hypertension 19 . Moreover, hyperuricemia could increase the risk of cardiovascular diseases and DM in patients with hypertension 20 and the mortality in patients with coronary heart diseases 21 . In accordance with results of previous studies above, our study finds that hyperuricemia is associated with cardiovascular diseases clustering among very elderly women.
In conclusion, prevalences of cardiovascular diseases are high among very elderly Chinese in Chengdu. Cluster of cardiovascular diseases are more apparently among vey elderly women. Hyperuricemia is associated with cardiovascular diseases clustering among very elderly women.
There are several limitations in this study. First, this is a cross-sectional study with relatively small sample, which could not describe any causality. Therefore, results of this survey should be applied with caution and longitudinal prospective cohort studies investigating the impact of hyperuricemia on cardiovascular diseases in very elderly are needed. Second, oral glucose tolerance test, hemoglobin A1c, echocardiogram, and pulse wave velocity could not been obtained in this survey because of economic reasons.

Methods
Participants. This cross-sectional survey was conducted from May 2013 to May 2015. Participants were recruited by using a stratified three-stage cluster sampling design. The first stage of sampling involved the random selection of five districts in Chengdu. And the second stage involved the random selection of four neighborhoods from each of the selected districts. From each of the neighborhoods, one residential committees was randomly selected.
A total number of 1056 participants from 20 residential committees were sampled according to registration data from Chengdu government. The study participants were defined as permanent residents of the households with a record in the household registration and living in local committees at least 3 years. Ethics approval was obtained from the Ethics Committee of the Second People's Hospital of Chengdu, China. The methods in the study were in accordance with relevant guidelines and the Declaration of Helsinki. All participants gave informed consent.
Study Design. This study was designed as a cross-sectional survey in general very elderly Chengdu residents.
All participants were assessed at a survey site in their committees by well-trained researchers. In oder to avoid uncertain influences of abominable weather, the survey was not conducted when the temperature outside was below 18 °C or over 25 °C. All participants completed a questionnaire-based interview, physical examination, and electrocardiogram and fast venous blood sample taken for subsequent biochemical analysis.
The survey questionnaire included questions about basic demographic characteristics, disease history, and family history, and so on. The physical examination involved general estimation, auscultation of lung and heart, measurements of height, weight, WC, and BP (HEM-7300, Omron, Kyoto, Japan). Electrocardiogram was taken for every participant after resting for more than 15 minutes (ECG-1350P, Nihon Kohden, Japan). The biochemical analysis included FG, TC, TG, LDL-C, HDL-C, UA and creatine, which were measured in the central laboratory of the Second People's Hospital of Chengdu.
Abnormalities definition. Pre-hypertension was defined as 120 ≤ SBP < 140 mmHg and/or 80 ≤ DBP < 90 mmHg. Hypertension was defined as SBP ≥ 140 mmHg and/or DBP ≥ 90 mmHg and/or self-reported treatment of hypertension with antihypertensive medication in last 2 weeks according to the Chinese Guidelines on Prevention and Control of Hypertension and the Seventh Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure guidelines 22,23 . Impaired fasting glucose was diagnosed if 6.1 mmol/L ≤ FG < 7.0 mmol/L. DM was diagnosed if FG ≥ 7.0 mmol/L, or FG < 7.0 mmol/L with a past history of DM 24 .
Renal dysfunction was defined as creatine ≥115 μmol/L according to the reference ranges recommend by central laboratory. Hyperuremia was defined as serum UA ≥ 416 μmol/L (7 mg/dL) in men and 357 μmol/L (6 mg/dL) in women 25,26 . Hypercholesterolaemia was defined as TC ≥ 5.2 mmol/L according to the guidelines of Chinese adult dyslipidemia prevention and control 27 . Hypertrigleicemiea was defined as TG ≥ 1.7 mmol/L. Hyper-low density lipoproteinemia defined as LDL-C ≥ 3.4 mmol/L. Hypo-high density lipoproteinemia was defined as HDL-C level of <0.9 mmol/L in men, or HDL-C level <1.0 mmol/L in women. Dyslipidaemia was defined as any or more than one of disorders as follows: hypercholesterolaemia, hypertrigleicemiea, hyper-low density lipoproteinemia and hypo-high density lipoproteinemia. According to WHO guidelines for the Asian Pacific population, overweight was defined as 24.0 ≤ BMI <28.0 kg/m 2 , and obesity as BMI at least 28.0 kg/m 2 28 . Visceral obesity was defined as WC at least 85 cm in women and at least 90 cm in men 24 .
Statistical analysis. Data were inputted by two researchers independently and checked by a third researcher.
Continuous variables were expressed as mean ± standard deviation. Frequencies were presented as percentages. Statistical comparison of continuous variables between men and women was conducted by using Student's t -test (Mann-Whitney U test for skewed data). And x 2 test was applied to compare frequencies between men and women. Prevalences among total elderly participants were standardized for age and sex. Stepwise multiple logistic regression models were used to evaluate the association between potential factors and disease clustering. ROC curves were used to demonstrate the sensitivity and specificity of related parameters. Calculations were performed by SPSS software (Version 17.0, SPSS Inc, Chicago, IL). A p value < 0.05 was considered statistically significant.