A Prospective Study on the Association between Uric Acid and Cognitive Function among Middle-Aged and Older Chinese

Abstract

Background: Uric acid (UA) is a powerful antioxidant that may have neuroprotective properties, yet it is also a risk factor of vascular disease that predisposes individuals to cognitive impairment. Results from longitudinal studies on UA and cognitive decline remain controversial.

Objective: We examined the associations of baseline plasma UA level with follow-up cognitive function as well as cognitive decline over time among a large sample of middle-aged and older Chinese.

Methods: Data from China Health and Retirement Longitudinal Study (CHARLS) were used. Cognitive function, including episodic memory, mental intactness, and global cognition, were tested twice with 2-year interval. Plasma UA was measured at baseline. Basic demographics, life habits, and health status were considered as potential confounders. Multiple linear regression models and mixed-effects regression models were fitted.

Results: A total of 12,798 individuals aged above 45 years were eligible with the follow-up time ranging from 1.33 to 2.42 years. Both global cognitive function and mental intactness declined, while episodic memory remained stable over time. In multiple linear regression models, compared with the lowest baseline UA level, 3rd baseline UA quartile was associated with better follow-up global cognitive function (b = 0.425, p = 0.041) and episodic memory (b = 0.413, p = 0.004), and highest baseline UA quartile was associated with better follow-up mental intactness (b = 0.253, p = 0.041) in males; highest baseline UA level was associated with better follow-up cognition for each measure (b = 0.281∼0.768, p≤0.046) in females. Mixed-effects regression models suggested no significant baseline UA-by-time interactions on any cognitive measure.

Conclusion: Higher baseline UA level was associated with better cognition in later life but not with rates of cognitive decline among middle-aged and older Chinese.

Keywords

Aged cognition episodic memory mental intactness uric acid

INTRODUCTION

Cognitive function is ‘the mental action or process of acquiring knowledge and understanding through thought, experience, and the senses’ [1]. All aging humans will develop some degree of decline in cognitive function that predisposes some individuals to neurological and psychiatric disorders leading to poor quality of life [2]. It is estimated that 20% of the elderly aged >70 years had mild cognitive impairment in the United States [3] and 5% ∼7% of people aged ≥60 years had dementia worldwide [4]. With the aging of the global population, the burden of cognitive impairment and dementia is expected to increase. A better understanding of the influence factors of cognitive function would inform strategies for the prevention of cognitive disorder.

Uric acid, as a major natural antioxidant, accounts for a substantial part of the antioxidative capacity of the plasma [5] that may have neuroprotective properties [6]. However, UA was proved to be associated with an increased risk of vascular disease [7], which may predispose individuals to cognitive impairment [8]. During the past years, epidemiological studies on UA and various cognitive outcomes have drawn conflicting conclusions. Some of them suggested that higher baseline UA level was associated with better subsequent cognitive function or was potentially protective against cognitive decline [8, 9], while others found the opposite association [10], and a few had mixed findings [6, 11]. Additionally, there is growing appreciation that UA may have different relationships with disease and outcomes in women and men[6 , 11–14].

This study thus aimed to examine the sex-specific associations of baseline plasma UA level with follow-up cognitive function as well as cognitive decline over time in a large, population-based sample derived from “China Health and Retirement Longitudinal Study (CHARLS)”.

SUBJECTS AND METHODS

Study sample

CHARLS is a nationally representative longitudinal survey on the middle-aged and older population (≥45 year) conducted by the National School of Development at Peking University. Procedures involved in the CHARLS and details concerning its multistage stratified sampling were described elsewhere [15]. Briefly, the survey included two waves covering 150 county-level units distributed in 28 provinces of China except Tibet. The baseline (wave 1, W1) survey was conducted in 2011-2012, and data were collected on 17,708 participants with the respondent rate above 80%. The second wave (wave 2, W2) survey successfully re-interviewed 15,770 of these individuals in 2013-2014. Both surveys including questionnaire, blood sample collection, as well as anthropometric and laboratory measurements were carried out by well-trained clinicians in a face-to-face, computer-aided personal interview (CAPI) manner. In this study, the 15,770 individuals who performed two wave surveys were included.

Cognitive function

Cognitive performance was calculated by two measures which named episodic memory and mental intactness in CHARLS. Each respondent was asked to repeat as many as possible of the 10 Chinese nouns just read to him/her (immediate word recall) and to recall the same list of these words 5 minutes later (delayed recall) [16]. Then episodic memory measure was defined as the summation of immediate and delayed recall scores. Mental intactness including numerical ability, time orientation, and picture drawing [17] were obtained from the following questions: serial 7 subtractions from 100 (up to five times), naming today’s date (month, day, year, and season), the day of the week, self-rated memory, and to redraw a picture shown to him/her. Answers to these questions were aggregated into a single mental intactness score that ranges from 0 to 15. The global cognitive function was the summation of episodic memory and mental intactness scores.

Uric acid

Blood samples of participants were collected after an overnight fast by medically-trained staff in the Chinese Center for Disease Control and Prevention (Chinese CDC). Venous blood was separated into plasma and buffy coat and then they were immediately stored frozen at –20°C and transported to the Chinese CDC in Beijing within 2 weeks where they were placed in a deep freezer and stored at –80°C until assay at Capital Medical University (CMU) laboratory. UA levels (mg/dL) were analyzed by UA Plus method [18].

Based on the quartiles of UA level, males were grouped as ‘≤4.0790 mg/dL’, ‘>4.0790 and ≤4.8048 mg/dL’, ‘>4.8048 and ≤5.6784 mg/dL’, and ‘>5.6784 mg/dL’, and females were grouped as ‘≤3.2793 mg/dL’, ‘>3.2793 and ≤3.8791 mg/dL’, ‘>3.8791 and ≤4.6158 mg/dL’, and ‘>4.6158 mg/dL’. The lowest UA group was chosen as the reference.

Other variables

Other covariates included age, educational level, cigarette smoking, alcohol drinking, memory-related disease (Alzheimer’s disease, Parkinson’s disease, and brain atrophy), obesity (body mass index ≥28 kg/m² [19]), hypertension, type 2 diabetes,cardiovascular disease, stroke, anxiety, and depressive symptoms. Educational level was categorized as ‘primary school or lower’ and ‘higher than primary school’. Cigarette smoking and alcohol drinking were classed as ‘never’, ‘current’, or ‘former’. Memory-related disease, obesity, hypertension, type 2 diabetes, cardiovascular disease, anxiety, and stroke were dichotomized as ‘no’ or ‘yes’. Center for Epidemiologic Studies Short Depression Scale (CES-D 10) was used to assess depressive symptoms.

Statistical analysis

The associations of baseline UA level with participants’ characteristics were first tested using univariate linear or logistic regression models, as appropriate. Paired sample t-test examined changes in cognitive measures between two visits. Age-adjusted linear regression models were used to test the cognitive differences across baseline UA quartiles. Then multiple linear regression models adjusting for potential confounders (age, educational level, cigarette smoking, alcohol drinking, memory-related disease, obesity, hypertension, type 2 diabetes, cardiovascular disease, stroke, anxiety, and depressive symptoms) with and without baseline cognitive measures were fitted to estimate the relationship between baseline UA and 3 follow-up cognitive test score(s) as alternative outcomes. Finally, multiple mixed-effects regression [11] models including interaction of time variable with baseline UA were conducted to determine whether rates of cognitive decline varied by levels of baseline UA. All data were analyzed by using STATA version 14 (StataCorp LP, College Station, Texas, USA).

RESULTS

A total of 12,798 participants (5,949 males and 6,849 females) were eligible for the currentstudy after excluding the individuals who did not complete cognitive measurements at W1 or W2. The follow-up time ranged from 1.33 to 2.42 years. Baseline UA ranged from 1.67 to 12.02 mg/dL for males (Mean = 4.96 mg/dL, SD = 1.27 mg/dL), and from 1.02 to 10.68 mg/dL for females (Mean = 4.01 mg/dL, SD = 1.07 mg/dL). Descriptive statistics for each covariate and the associations of participants’ characteristics with baseline UA level were shown in Table 1.

Table 1

Characteristics of male and female participants in CHARLS according to baseline UA quartiles

Characteristics	Total	UA for Males					UA for Females
		≤4.0790	>4.0790	>4.8048	>5.6784	p-value	≤3.2793	>3.2793	>3.879	>4.6158	p-value
			≤4.8048	≤5.6784				≤3.8791	≤4.6158
Age, year ( $\bar{X} \pm s$ )	59.3±9.7	59.6±8.8	59.8±9.3	59.6±9.1	61.0±9.8	0.001	57.4±9.1	58.3±9.0	58.9±9.6	60.8±10.0	<0.001
Educational level (%)						0.994					0.552
≤primary school	67.6	60.1	59.9	58.7	60.5		79.0	77.7	76.9	78.2
>primary school	32.4	39.9	40.1	41.3	39.5		21.0	22.3	23.1	21.8
Cigarette smoking (%)						<0.001					0.110
Never smoked	62.4	25.3	23.6	27.2	27.9		92.1	91.8	93.7	92.0
Former smoked	8.3	14.5	15.4	18.1	19.8		1.4	1.8	1.2	2.4
Current smoked	29.2	60.1	61.0	54.7	52.4		6.6	6.3	5.1	5.6
Alcohol drinking (%)						<0.001					0.119
Never drank	68.7	51.0	46.3	42.7	37.1		91.2	91.2	90.3	90.4
Former drank	6.0	8.2	10.4	11.2	11.5		1.9	1.6	2.8	3.1
Current drank	25.4	40.9	43.3	46.0	51.4		6.9	7.3	6.9	6.5
Memory-related disease (%)	1.4	1.5	1.9	1.4	1.3	0.534	1.1	1.1	1.0	0.8	0.303
Stroke (%)	2.2	1.6	2.3	2.6	3.7	0.001	2.1	2.1	2.0	2.6	0.421
Hypertension (%)	24.2	17.4	20.9	25.0	32.4	<0.001	18.5	23.6	27.1	37.5	<0.001
Type 2 Diabetes (%)	5.6	5.6	5.0	5.5	5.3	0.863	6.3	5.9	4.7	7.7	0.308
Cardiovascular disease (%)	11.9	9.5	9.8	12.4	11.2	0.057	13.4	13.6	13.1	15.9	0.097
Obesity (%)	11.3	4.9	7.0	8.3	12.5	<0.001	9.4	12.4	14.4	22.3	<0.001
Anxiety (%)	1.4	1.5	0.7	1.0	0.8	0.133	1.9	1.9	2.2	1.4	0.535
Depressive symptom ( $\bar{X} \pm s$ )	8.4±6.3	8.0±6.2	7.6±5.8	7.5±6.0	7.0±5.4	<0.001	9.8±6.8	9.4±6.5	9.3±6.5	9.2±6.6	0.015

Bold for p≤0.05 in multi-sample comparisons. Univariate logistic regression models were fitted for educational level, cigarette smoking, alcohol drinking, memory-related disease, stroke, hypertension, diabetes, cardiovascular disease, obesity, and anxiety. Univariate linear regression models were fitted for age and depression scores.

Both global cognitive function and mental intactness declined between two visits in males and females. Performance on episodic memory remained stable over time (Table 2). As shown in Table 3, after adjusting for age, each cognitive measure was significantly different (p < 0.001) across baseline UA quartile groups. Both males and females with higher UA quartiles had better cognitive function for each cognitive domain.

Table 2

Cognitive changes between two visits

Cognitive measure	2011	2013	d	p
	Mean±SD	Mean±SD	Mean±SD
Male
Global cognition	16.60±5.07	16.40±5.30	–0.19±4.86	0.004
Episodic memory	7.34±3.21	7.37±3.34	0.03±3.57	0.581
Mental intactness	9.03±3.08	8.84±3.16	–0.19±2.99	<0.001
Female
Global cognition	14.99±5.67	14.80±5.92	–0.18±4.93	0.005
Episodic memory	7.24±3.42	7.28±3.49	0.04±3.66	0.360
Mental intactness	7.40±3.42	7.24±3.48	–0.17±2.96	<0.001

Bold for p ≤ 0.05. d is the difference of cognitive measure between 2013 and 2011.

Table 3

Age-adjusted comparison of cognitive measures across UA quartiles by sex

Sex	UA Quartiles	2011						2013
		Global Cognition		Episodic memory		Mental intactness		Global Cognition		Episodic memory		Mental intactness
		$\bar{X} \pm s$	post-hoc p	$\bar{X} \pm s$	post-hoc p	$\bar{X} \pm s$	post-hoc p	$\bar{X} \pm s$	post-hoc p	$\bar{X} \pm s$	post-hoc p	$\bar{X} \pm s$	post-hoc p
Male	UA 1st quartile (ref.)	16.0±5.1	–	7.1± 3.2	–	8.7±3.2	–	15.8±5.3	–	7.1±3.3	–	8.7±3.1
	UA 2st quartile	16.4±5.1	0.035	7.2±3.2	0.186	8.9±3.1	0.067	15.9±5.6	0.603	7.2±3.4	0.244	8.6±3.3	0.582
	UA 3st quartile	16.3±5.2	0.042	7.2±3.3	0.113	9.0±3.1	0.013	16.5±5.3	0.002	7.5±3.4	0.001	8.9±3.1	0.093
	UA 4st quartile	16.6±5.2	<0.001	7.3±3.3	0.003	9.1±3.2	<0.001	16.4±5.3	<0.001	7.2±3.4	0.070	9.1±3.1	<0.001
Female	UA 1st quartile (ref.)	14.3±6.7	–	6.9±3.3	–	7.1±3.5	–	14.0±6.0	–	6.9±3.5	–	6.9±3.5	–
	UA 2st quartile	14.6±5.8	0.002	7.2±3.4	0.002	7.2±3.4	0.037	14.4±6.1	0.015	7.1±3.6	0.072	7.02±3.4	0.021
	UA 3st quartile	14.7±5.7	<0.001	7.1±3.5	0.003	7.3±3.4	0.001	14.3±5.9	0.006	6.9±3.6	0.225	7.2±3.4	<0.001
	UA 4st quartile	14.8±5.9	<0.001	7.0±3.6	<0.001	7.4±3.5	<0.001	14.6±6.1	<0.001	7.0±3.6	<0.001	7.3±3.5	<0.001

Bold for p≤0.05. Age-adjusted linear regression models were fitted for cognitive measures.

Table 4 showed the relationship between the baseline UA level and the follow-up cognitive measures in multiple linear regression models. For males, after adjusting for potential confounders, referenced to the lowest UA level, the 3rd quartile level of UA was associated with better global cognitive function (b = 0.455, p = 0.044) and episodic memory (b = 0.427, p = 0.003), and the highest level of UA was related to higher global cognitive function (b = 0.582, p = 0.011) and mental intactness (b = 0.346, p = 0.013). In model 2, baseline analogous cognitive measure was also included as an independent variable. The associations of 3rd quartile UA level with global cognitive function (b = 0.425, p = 0.041) and episodic memory (b = 0.413, p = 0.004), as well as of highest UA level with mental intactness (b = 0.253, p = 0.041) were still statistically significant. For females, positive relations between highest UA level and better cognitive measures were found (b = 0.281∼0.768, p≤0.046) in both models.

Table 4

Multiple linear regression model testing the association between baseline UA level and follow-up cognitive measures

Sex	Independent Variable	Global Cognition		Episodic Memory		Mental intactness
		β±SE (β)	p	β±SE (β)	p	β±SE (β)	p
Male	Model 1
	UA 1st quartile (≤4.0790 mg/dL) (ref.)	–	–	–	–	–	–
	UA 2nd quartile (4.0791–4.8048 mg/dL)	0.033±0.225	0.882	0.143±0.145	0.323	–0.122±0.136	0.372
	UA 3rd quartile (4.8049–5.6784 mg/dL)	0.455 ± 0.226	0.044	0.427 ± 0.145	0.003	0.023±0.137	0.864
	UA 4th quartile (>5.6784 mg/dL)	0.582 ± 0.229	0.011	0.252 ± 0.147	0.086	0.346 ± 0.139	0.013
	Model 2
	UA 1st quartile (≤4.0790 mg/dL) (ref.)	–	–	–	–	–	–
	UA 2nd quartile (4.0791–4.8048 mg/dL)	–0.084±0.208	0.688	0.071±0.143	0.621	–0.173±0.123	0.158
	UA 3rd quartile (4.8049–5.6784 mg/dL)	0.425 ± 0.208	0.041	0.413 ± 0.143	0.004	0.001±0.123	0.996
	UA 4th quartile (>5.6784 mg/dL)	0.345±0.211	0.102	0.182±0.145	0.208	0.253 ± 0.124	0.041
Female	Model 1
	UA 1st quartile (≤3.2793 mg/dL) (ref.)
	UA 2nd quartile (3.2794–3.8790 mg/dL)	0.335±0.230	0.145	0.141±0.142	0.321	0.167±0.136	0.219
	UA 3rd quartile (3.8791–4.6158 mg/dL)	0.272±0.230	0.237	–0.023±0.142	0.870	0.281±0.135	0.038
	UA 4th quartile (>4.6158 mg/dL)	0.768 ± 0.234	0.001	0.290 ± 0.145	0.045	0.480 ± 0.137	< 0.001
	Model 2
	UA 1st quartile (≤3.279 3 mg/dL) (ref.)
	UA 2nd quartile (3.2794–3.8790 mg/dL)	0.373±0.206	0.071	0.144±0.140	0.305	0.199±0.117	0.090
	UA 3rd quartile (3.8791–4.6158 mg/dL)	0.168±0.207	0.417	–0.007±0.141	0.962	0.225±0.117	0.054
	UA 4th quartile (>4.6158 mg/dL)	0.557 ± 0.212	0.009	0.288 ± 0.144	0.046	0.337 ± 0.119	0.005

Bold for p≤0.05. Model 1: adjusted for age, educational level, depression scores, cigarette smoking, alcohol drinking, memory-related disease, stroke, hypertension, type 2 diabetes, heart problems, anxiety, and obesity. Model 2: adjusted for model 1+ baseline cognition.

Table 5 displayed associations between baseline UA and longitudinal cognitive changes, based on mixed-effects regression analyses. No significant UA-by-time interactions were found, indicating that baseline UA level did not predict rates of cognitive decline. Figure 1 depicted predictive margins from the mixed-effects regression models. Both in males and females, for each cognitive measure, the decline rates across different baseline UA levels weresimilar.

Table 5

Longitudinal cognitive change by baseline UA: Mixed-effected linear regression models

Cognitive measure	Intercept		Time		Baseline UA×Time
	γ±SEE	p	γ±SEE	p	γ±SEE	p
Male
Global cognition	20.26±0.59	<0.001	–0.15±0.09	0.113	0.02±0.03	0.580
Episodic memory	9.97±0.39	<0.001	–0.01±0.07	0.843	–0.00±0.03	0.988
Mental intactness	9.26±0.49	<0.001	–0.12±0.06	0.031	0.01±0.02	0.597
Female
Global cognition	18.13±0.70	<0.001	–0.13±0.10	0.178	0.01±0.04	0.751
Episodic memory	9.75±0.35	<0.001	0.02±0.07	0.727	0.00±0.03	0.986
Mental intactness	9.01±0.38	<0.001	–0.13±0.05	0.019	0.01±0.02	0.442

Fig.1

Longitudinal association of baseline UA with cognitive changes over time: Mixed-effects regression model. A) for Global cognition, B) for Episodic memory and C) for Mental intactness. “ua = 1” is baseline UA 1st quartile, “ua = 2” is baseline UA 2nd quartile, “ua = 3” is baseline UA 3rd quartile, “ua = 4” is baseline UA 4th quartile.

DISCUSSION

This study examined the longitudinal relationship between plasma UA level and cognitive function in middle-aged and older Chinese. The analysis including 12,798 participants identified that higher baseline UA level was associated with better cognition in later life but not with rates of cognitive decline both in males and females.

Our findings about benefits of high baseline level of UA on follow-up cognitive function are consistent with some previous literatures [8 , 21]. A prospective population-based cohort study among 4,618 participants aged 55 years and over showed that after correcting for several cardiovascular risk factors, higher serum UA levels were associated with a decreased risk of dementia [8]. And their further analyses in 1,724 individuals who remained free of dementia during follow-up suggested higher serum UA levels at baseline were associated with better cognitive function later in life for all cognitive domains [8]. A latest meta-analysis including 46 papers that assessed the association between serum UA and any measure of cognitive function or a clinical diagnosis of dementia concluded that compared to controls, serum UA was lower in dementia [22]. It is interesting that this meta-analysis found different associations by dementia type with an apparent association for Alzheimer’s and Parkinson’s disease-related dementia but not in cases of mixed dementia or vascular dementias [22]. However other researchers reported a potentially adverse effect of higher level UA on cognitive outcomes. For instance, a community-dwelling study in America suggested that even mild elevated level of UA might increase the risk of cognitive decline among older adults [23]. A cohort study of 423 cognitively healthy older women observed that a higher baseline UA was linked to poorer working memory, a trend toward slower manual speed and dexterity [6].

We did not found the significant baseline UA-by-time interactions on any cognitive changes, indicating that baseline UA level did not predict rates of cognitive decline over time, which is similar to a previous study with 9 years of follow-up time [6]. However, another study with a mean of 4.64-year follow-up conducted by Beydoun et al. [11] concluded that a higher baseline serum UA was associated with a faster cognitive decline in a visual memory/visuo-construction ability test. It is noted that they also found that an increasing change of UA over-time had particular benefit on the domain of attention/processing speed among older men [11]. The different conclusions could possibly arise from different study design, studied samples or follow-up time.

UA was higher in men compared to women within each age group [24], and it may play different roles in cognitive outcomes/decline in men and women. For example, a study by Heo et al. uncovered a dose-response relationship between UA and brain infarction only among women [13]; Beydoun et al. found the potentially beneficial effect in the case of attention only among older men. But our data did not show the sex difference pattern in the longitudinal association between baseline UA and cognitive function.

The mechanisms underlying the relationship between UA level and cognitive function are still not fully understood. Two potential mechanisms exist by which high UA level may decrease the risk of cognitive decline. First, based on biochemical evidence, UA is considered to be a reactive oxygen species scavenger, and has strong anti-oxidant properties. It is particularly effective on quenching hydroxyl, super-oxide, and peroxynitrite radicals, and may play a protective physiological role by preventing lipid peroxidation [25]. In various organs and vascular beds, local UA concentrations increase in time of acute oxidative stress and ischemia, and the elevated concentrations might be a compensatory mechanism that provides protection against increased free radical activity [26]. Second, a genetic study found that four single nucleotide polymorphisms in the uric acid transporter gene (SLC2A9) were significantly related to memory performance [27], suggesting the association between UA and cognitive function is perhaps attributed to a common set of genetic factors.

A major strength of this study was large number of participants from a prospective study, allowing a much greater possibility of reasonable conclusions. However, there were also some limitations. First, measuring UA at middle age or older may not capture long-term exposure of UA over time because greater variability may have been present earlier in the period of older adulthood at which time vascular changes may have developed [6]. Second, factors such as diet and nutrients which might influenced UA level were not included in our models as potential confounders. Third, the cognitive domains wererelatively limited. Fourth, measures interval for cognitive outcome was too short to explore the interaction of time and baseline UA on cognition.

Conclusion

In sum, the present findings corroborated the notion that higher baseline UA level was associated with better cognition in later life among middle-aged and older Chinese. However, baseline UA level did not predict rates of cognitive decline over time.

Footnotes

ACKNOWLEDGMENTS

The authors appreciate the Chinese CDC for making the survey and making it available online freely, and all the participants for providing these data. This research project was supported by grants from the National Natural Science Foundation of China (31371024).

Authors’ disclosures available online ().

References

“Cognition – definition of cognition in English from the Oxford dictionary”, Oxford University Press, http://www.oxforddictionaries.com/definition/english/cognition, Accessed February 04, 2016.

Mikkelsen

, Stojanovska

, Tangalakis

, Bosevski

, Apostolopoulos

(2016) Cognitive decline: A vitamin B perspective. Maturitas 93, 108–113.

Plassman

, Langa

, Fisher

, Heeringa

, Weir

, Ofstedal

, Burke

, Hurd

, Potter

, Rodgers

, Steffens

, McArdle

, Willis

, Wallace

(2008) Prevalence of cognitive impairment without dementia in the United States. Ann Intern Med 148, 427–434.

Prince

, Bryce

, Albanese

, Wimo

, Ribeiro

, Ferri

(2013) The global prevalence of dementia: A systematic review and metaanalysis. Alzheimers Dement 9, 63–75 e62.

Davies

, Sevanian

, Muakkassah-Kelly

, Hochstein

(1986) Uric acid-iron ion complexes. A new aspect of the antioxidant functions of uric acid. Biochem J 235, 747–754.

Vannorsdall

, Kueider

, Carlson

, Schretlen

(2014) Higher baseline serum uric acid is associated with poorer cognition but not rates of cognitive decline in women. Exp Gerontol 60, 136–139.

Martinez-Quintana

, Tugores

, Rodriguez-Gonzalez

(2016) Serum uric acid levels and cardiovascular disease: The Gordian knot. J Thorac Dis 8, E1462–E1466.

Euser

, Hofman

, Westendorp

, Breteler

(2009) Serum uric acid and cognitive function and dementia. Brain 132, 377–382.

Molshatzki

, Weinstein

, Streifler

, Goldbourt

, Tanne

(2015) Serum uric acid and subsequent cognitive performance in patients with pre-existing cardiovascular disease. PLoS One 10, e0120862.

10.

Cicero

, Desideri

, Grossi

, Urso

, Rosticci

, D’Addato

, Borghi

, Brisighella

Heart Study G

(2015) Serum uric acid and impaired cognitive function in a cohort of healthy young elderly: Data from the Brisighella Study. Intern Emerg Med 10, 25–31.

11.

Beydoun

, Canas

, Dore

, Beydoun

, Rostant

, Fanelli-Kuczmarski

, Evans

, Zonderman

(2016) Serum uric acid and its association with longitudinal cognitive change among urban adults. J Alzheimers Dis 52, 1415–1430.

12.

Alderman

, Cohen

, Madhavan

, Kivlighn

(1999) Serum uric acid and cardiovascular events in successfully treated hypertensive patients. Hypertension 34, 144–150.

13.

Heo

, Lee

(2010) High levels of serum uric acid are associated with silent brain infarction. J Neurol Sci 297, 6–10.

14.

O’Reilly

, Gao

, Weisskopf

, Chen

, Schwarzschild

, Spiegelman

, Ascherio

(2010) Plasma urate and Parkinson’s disease in women. Am J Epidemiol 172, 666–670.

15.

Zhao

, Hu

, Smith

, Strauss

, Yang

(2014) Cohort profile: The China Health and Retirement Longitudinal Study (CHARLS). Int J Epidemiol 43, 61–68.

16.

McArdle

, Ferrer-Caja

, Hamagami

, Woodcock

(2002) Comparative longitudinal structural analyses of the growth and decline of multiple intellectual abilities over the life span. Dev Psychol 38, 115–142.

17.

, Lei

, Smith

, Zhao

(2012) Effects of social activities on cognitive functions: Evidence from CHARLS National Academies Press (US), National Research Council (US) Panel on Policy Research and Data Needs to Meet the Challenge of Aging in Asia.

18.

Zhao

, Crimmins

, Hu

, Ge

, Kim

, Strauss

, Yang

, Yin

, Wang

(2014) China Health and Retirement Longitudinal Study, 2011-2012 National Baseline Blood Data Users’ Guide. China Center for Economic Research, Peking University, Beijing.

19.

Zhou

(2002) Effect of body mass index on all-cause mortality and incidence of cardiovascular diseases–report for meta-analysis of prospective studies open optimal cut-off points of body mass index in Chinese adults. Biomed Environ Sci 15, 245–252.

20.

Mendez-Hernandez

, Salas-Pacheco

, Ruano-Calderon

, Tellez-Valencia

, Cisneros-Martinez

, Barraza-Salas

, Arias-Carrion

(2015) Lower uric acid linked with cognitive dysfunction in the elderly. CNS Neurol Disord Drug Targets 14, 564–566.

21.

Al-khateeb

, Althaher

, Al-khateeb

, Al-Musawi

, Azzouqah

, Al-Shweiki

, Shafagoj

(2015) Relation between uric acid and Alzheimer’s disease in elderly Jordanians. J Alzheimers Dis 44, 859–865.

22.

Khan

, Quinn

, Hewitt

, Fan

, Dawson

(2016) Serum uric acid level and association with cognitive impairment and dementia: Systematic review and meta-analysis. Age (Dordr) 38, 16.

23.

Schretlen

, Inscore

, Jinnah

, Rao

, Gordon

, Pearlson

(2007) Serum uric acid and cognitive function in community-dwelling older adults. Neuropsychology 21, 136–140.

24.

Gephardt

, Hanlon

, Matson

(1964) Blood uric acid values as related to sex and age. JAMA 189, 1028–1029.

25.

Squadrito

, Cueto

, Splenser

, Valavanidis

, Zhang

, Uppu

, Pryor

(2000) Reaction of uric acid with peroxynitrite and implications for the mechanism of neuroprotection by uric acid. Arch Biochem Biophys 376, 333–337.

26.

Nieto

, Iribarren

, Gross

, Comstock

, Cutler

(2000) Uric acid and serum antioxidant capacity: A reaction to atherosclerosis? Atherosclerosis 148, 131–139.

27.

Houlihan

, Wyatt

, Harris

, Hayward

, Gow

, Marioni

, Strachan

, Price

, Starr

, Wright

, Deary

(2010) Variation in the uric acid transporter gene (SLC2A9) and memory performance. Hum Mol Genet 19, 2321–2330.