Effect of serum uric acid to creatinine ratio on cognitive function decline in middle-aged adults: Longitudinal evidence from CHARLS

Abstract

Background

Serum uric acid (SUA) was a predictor of cognitive function. The association of SUA/serum creatinine ratio (Scr), which represents renal function-normalized SUA and cognitive function is unknown.

Objective

This study investigated the association of the SUA/Scr with cognitive function and the potential mediation effect of inflammation in the above relationship.

Methods

This study used 1–5 waves of data from the China Health and Retirement Longitudinal Study. 3302 participants aged 45–60 years at baseline were included. Among them, 1129 who attended subsequent 2–3 waves were further included for cumulative exposure calculation to the SUA/Scr ratio. The Cox models were used to evaluate the impact of baseline SUA/Scr ratio and its cumulative exposure on cognitive function decline.

Results

During a median follow-up of 8.6 years, there were 1512 (45.8%) cognitive function declined. After adjustment, the highest quartile of the SUA/SCr ratio was associated with the highest risk of cognitive function decline (Hazard ratio, 1.175; 95% confidence interval, 1.015–1.360). Restricted cubic spline showed a linear association between the SUA/Scr ratio and the risk of cognitive function decline (p_non−linear= 0.514). There were a stronger association of cumulative SUA/Scr ratio and its exposure burden with cognitive function decline [the highest versus lowest quartile: 1.635 (1.006–2.656), the high versus low group: 1.729 (1.212–2.466), respectively]. No significant mediating effect through white blood cell count or C-reactive protein in SUA/Scr ratio-cognitive function decline was found.

Conclusions

The SUA/Scr ratio was associated with a higher risk of cognitive decline, whereas the mechanism mediated by inflammation indicators was not found.

Keywords

Alzheimer's disease cognitive function cohort study inflammation indicators middle-aged serum uric acid to serum creatinine ratio

Introduction

One of the significant challenges facing public health care in China is the trend of an aging society. The acceleration of aging inevitably promotes the occurrence and development of chronic diseases.¹ Alzheimer's disease is a leading cause of disability in old adults, including China.² Cognitive function decline, a neurodegenerative process due to increasing age, develops into Alzheimer's disease under the interaction of genetic susceptibility, environmental damage and unhealthy lifestyle.³ Notably, middle age, a period defined by unique central and peripheral processes, is related to changes in the volume of parts of the brain, atrophy of the hippocampus, and decreased connectivity between different parts of the brain. Cognitive and brain health decline are most pronounced in this stage.^4,5 Neuropathological lesions could be observed 20 years before the onset of the clinical symptoms of Alzheimer's disease.⁶ Individual interventions in midlife could slow down the process of aging and its related diseases before organ frailty occurs.

Serum uric acid (SUA) is considered to be a risk marker for major chronic diseases, especially cardiovascular disease.⁷ The link between SUA levels and cognitive function and Alzheimer's disease is still controversial.^8–10

High SUA levels may play a negative role in the course of vascular dementia, whereas the antioxidant properties of SUA support the hypothesis of a neuroprotective effect in the brain. These may be explained partly by the fact that levels of endogenous uric acid depend on renal clearance function, for 90% of SUA is filtered and reabsorbed by the kidneys¹¹ or by a single measurement of SUA at baseline rather than the dynamic changes. To our knowledge, few studies have investigated the effect of renal function-normalized SUA (SUA to serum creatinine ratio, SUA/SCr ratio), a superior indicator of endogenous uric acid, and its cumulative exposure on cognitive function. In addition, given the characteristic of SUA activation of inflammatory factors⁹ and the neuroinflammation-associated cognitive decline in midlife,⁵ the mediating role of inflammatory indicators in the association between SUA/Scr ratio and cognition also needs to be evaluated.

Therefore, this study aimed to explore the associations of the SUA/Scr ratio and its cumulative exposure with cognitive decline among middle-aged Chinese adults. The mediating role of inflammation indicators in the investigated associations was also evaluated.

Methods

Study population

This prospective cohort study used data from waves 1–5 (2011–2020) of the China Health Retirement Longitudinal Study (CHARLS), an ongoing nationally representative longitudinal survey of Chinese adults aged ≥45 years and older from both urban and rural areas in 28 provinces. Detailed information about the CHARLS can be found in a previous paper.¹² The protocol was approved by the Ethical Review Committee of Peking University (IRB00001052-11015). The participants were informed of the research purpose and signed written informed consent forms.

Among the 17,708 participants interviewed at baseline (wave 1), 9641 middle-aged adults were included. Participants meeting the following criteria were excluded: (1) 3490 participants with missing data on cognitive test scores at baseline and in all follow-up (wave 2 to wave 5); (2) 2848 participants with missing data on confounders; (3) missing data on SUA or Scr (n = 1). This left 3302 participants for analysis of the baseline effect of the SUA/Scr ratio on cognitive function. (4) then missing data on SUA or Scr at wave 3 (n = 1298); (5) 875 participants were excluded due to cognitive function decline between wave 2 and wave 3. Finally, a total of 1129 participants were included for analysis of the effect of cumulative exposure of SUA/Scr ratio on cognitive function. The detailed selection of participants is illustrated in Figure 1.

Figure 1.

Flowchart with exclusion criteria and definition of the study sample.

Data collection and definitions

Detailed information on individual demographics, lifestyle factors, and histories of disease and medications was collected using face-to-face computer-assisted personal interviews. Demographics included age, gender, marital status, residence and education levels. Lifestyle factors included smoking and drinking, defined as having smoked and an alcoholic drink in the past, respectively. History of six self-reported chronic diseases included hypertension, diabetes, dyslipidemia, heart problems, stroke and psychological problems. Histories of medication included lipid-lowering drugs and drugs for stroke and psychological problems.

Height and weight were measured with participants wearing light dressing and without shoes. Body mass index (BMI) was calculated as weight/ height² (kg/ m²). Waist circumference (WC) was measured at the navel when participants held their breath at the end of the exhalation. Abdominal obesity was defined as WC ≥ 85 cm in males and ≥80 cm in females.¹³ Blood pressure was measured using a standardized automatic electronic sphygmomanometer (HEM-7112; Omron, Dalian, China), and the average of the three values was used. Fasting blood samples were collected after an overnight fast at waves 1 and 3. Venous blood samples were frozen at −20°C, shipped to the Chinese Center for Disease Control and Prevention in Beijing, and stored at −80°C until assayed at the Youanmen Center for Clinical Laboratory of Capital Medical University.¹⁴ Glucose and lipids were measured using the enzymatic colormetric test. SUA was measured using the UA Plus method. Scr was measured using the Rate-blanked and compensated Jaffe creatinine method. C-reactive protein (CRP) was measured using the immunoturbidimetric assay. White blood cell (WBC) count analysis was done on automated analyzers available in the local Centers for Disease Control or town/ village health centers laboratories. To determine cumulative SUA/Scr ratio burden on cognitive function declined, the duration of cumSUA/Scr ratio was calculated as (SUA/Scr ratio1 + SUA/Scr ratio3)/ 2 × time1–3, where SUA/ Scr ratio1 and SUA/ Scr ratio3 indicated SUA/Scr ratio at wave 1 and wave 3. Time1–3 indicated the examination time between wave1 and wave 3. cumSUA/Scr ratio burden was calculated as (SUA/Scr ratio1 + SUA/Scr ratio3)/2 − cutoff × time1–3, where cutoff value was used 5.6, the median of SUA/Scr ratio at wave 1.¹⁵

Cognitive function assessment

Cognitive function was measured in waves 1 to 5. Two dimensions of cognitive function were captured: episodic memory and executive function, As described in previous studies.^16,17 Episodic memory was measured by immediate recall (score range, 0 to 10 points) and delayed recall (score range, 0 to 10 points) for 10 unrelated Chinese nouns and was calculated as the sum of two recall tests (ranging from 0 to 20 points). Executive function was measured by time orientation (score range, 0–5 points), numerical ability (score range, 0–5 points), and visuospatial ability (score range, 0–1 points). It was also calculated as the sum of these three tests (ranging from 0 to 11 points). Global cognition was defined as the sum score of these two dimensions (ranging from 0 to 31 points), with a higher cognitive score indicating better cognitive performance. Detailed measurements of each cognitive dimension were provided in the Supplemental Methods. In the present study, Cognitive function decline was defined as a reduction of 3 or more points at any time during follow-up.¹⁷

Statistical analysis

Continuous variables were presented as the means and standard deviation (SD), and categorical variables were expressed as numbers (%). Values between groups were compared using variance (ANOVA) or independent t-test and Chi-square test or Fisher's exact test, as appropriate. To evaluate the association of baseline SUA/Scr ratio, cumSUA/Scr ratio and cumSUA/Scr ratio burden with cognitive decline, we respectively used Cox regression to calculate the Hazard ratios (HRs) with corresponding 95% confidence intervals (CIs). Three models were established: Model 1 was unadjusted. Model 2 was adjusted for age, sex and education levels. Model 3 was adjusted additionally adjusted for a propensity score. The propensity score was calculated with a multivariate linear regression model, entering the SUA/Scr ratio as the dependent variable. The independent variables included marital status, smoking, drinking, hypertension, diabetes, stroke, heart problems, dyslipidemia, psychological problems, antihypertensive medications, hypoglycemic medications, lipid-lowering medications, BMI, WC, systolic blood pressure (SBP); diastolic blood pressure (DBP), blood glucose, total cholesterol (TC), triglyceride (TG), high-density lipoprotein cholesterol (HDLC), and low-density lipoprotein cholesterol (LDLC). P values for trend were computed using SUA/Scr ratio quartiles as ordinal variables. Restricted cubic spline (RCS) with three knots at the 5th, 33.3th, 66.6th, and 95th centiles based on Cox regression was performed to flexibly model the association of SUA/Scr ratio with future cognitive function decline risk. Subgroup analyses were stratified by traditional risk factors and summarized with forest plots. An interaction term within each subgroup was calculated using the likelihood ratio test. Four sensitivity analysis were done. Considering possible reverse causality, we first analyzed the association between the SUA/Scr ratio and cognitive function by excluding high-risk individuals who had a history of stroke or psychological problems.¹⁸ Second, we analyzed whether the association would change if individuals who had a low cognitive score (score range, <15 points) at the baseline were excluded. Then, we analyzed the association only in individuals with hyperuricemia (SUA concentrations >6 and 7 mg/dL for females and males, respectively).¹⁹ Finally, we employed the E-value as a measure to evaluate the stability of Cox regression models. This metric aids in assessing whether an unmeasured underlying factor may influence an observed association. Mediation analyses were carried out to investigate the potential mediating effect of inflammatory factors (mediator, WBC count and CRP) on the association between SUA/Scr ratio, cumSUA/Scr ratio and cumSUA/Scr ratio burden (exposure) and changes in cognitive function scores (outcome). The mediated (indirect) effect, direct effect and total effect were obtained. Data analysis was performed using the SAS 9.4 software (SAS Institute Inc., Cary, NC, USA) and IBM SPSS 26.0 (SPSS Inc., Chicago, IL, USA). A two-tailed p value less than 0.05 was accepted as indicating statistical significance.

Results

Baseline characteristics

Baseline characteristics grouped by residence was shown in Table 1. Among the 3302 participants, the mean age was 52.19 ± 4.44 years, and 56.35% of the participants were female. Participants in urban were significantly older and more likely to have a history of hypertension, smoking and drinking. They had lower blood glucose, TC, TG, WC, SDP, and DBP than those in rural areas. Figure 2 showed the incidence rates of cognitive function decline according to baseline SUA/Scr ratio by residence and the distribution of cognitive test scores at baseline and the last follow-up. There was a significant difference in incidence rates of cognitive function decline among baseline SUA/Scr ratio quartiles in rural. Cognitive test scores were concentrated in 10 to 20 points, and more significantly in the last follow-up. The distribution of high marked a slight decline in the last follow-up.

Figure 2.

Baseline characteristics of participants. (a) Incidence rates of cognitive function decline according to the SUA/Scr ratio by residence. (b) Frequency distribution of total cognitive scores at baseline and the last follow-up. SUA: serum uric acid; Scr: serum creatinine.

Table 1.

Characteristics of participants at baseline.

Characteristics	Total	Rural	Urban	p
Demographic
Age, y	52.19 ± 4.44	52.23 ± 4.80	52.16 ± 4.41	0.667
Female, n (%)	1736 (56.35)	732 (42.17)	1004 (50.12)	<0.001
Education levels, n (%)				<0.001
Less than lower secondary education	2763 (83.68)	995 (76.60)	1768 (88.27)
Upper secondary and vocational training	502 (15.20)	274 (21.09)	228 (11.38)
Tertiary education	37 (1.12)	30 (2.31)	7 (0.35)
Never Married, n (%)	14 (0.42)	3 (0.23)	11 (0.55)	0.169
Lifestyle
Smoking, n (%)	1284 (38.89)	447 (34.41)	837 (41.79)	<0.001
Drinking, n (%)	1319 (39.95)	489 (37.64)	830 (41.44)	0.030
History of disease and medications
Hypertension, n (%)	699 (21.17)	304 (23.40)	395 (19.72)	0.011
Diabetes, n (%)	193 (5.84)	99 (7.62)	94 (4.69)	<0.001
Stroke, n (%)	53 (1.61)	19 (1.46)	34 (1.70)	0.600
Heart problems, n (%)	316 (9.57)	140 (10.78)	176 (8.79)	0.058
Dyslipidemia, n (%)	363 (10.99)	188 (14.47)	175 (8.74)	<0.001
Psychological problems, n (%)	24 (0.73)	9 (0.69)	15 (0.75)	0.853
Antihypertensive medications, n (%)	505 (15.29)	239 (18.40)	266 (13.28)	<0.001
Hypoglycemic medications, n (%)	131 (3.97)	73 (5.62)	58 (2.90)	<0.001
Lipid-lowering medications, n (%)	181 (5.48)	84 (6.47)	97 (4.84)	0.045
Laboratory values
Blood glucose, mg/dL	108.60 ± 32.87	111.70 ± 37.18	106.58 ± 29.58	<0.001
TC, mg/dL	193.51 ± 37.58	195.18 ± 38.21	192.43 ± 37.12	0.040
TG, mg/dL	135.24 ± 101.29	144.71 ± 105.05	129.10 ± 98.30	<0.001
HDLC, mg/dL	50.36 ± 14.93	48.42 ± 14.49	51.62 ± 15.07	<0.001
LDLC, mg/dL	116.56 ± 34.70	117.58 ± 35.24	115.90 ± 34.33	0.174
Physical values
BMI, kg/m²	25.03 ± 7.97	26.70 ± 6.74	23.94 ± 9.74	0.071
Waist circumference, cm	85.01 ± 11.91	85.99 ± 13.40	84.38 ± 10.79	<0.001
SBP, mm Hg	125.55 ± 19.38	126.71 ± 19.55	124.81 ± 19.23	0.006
DBP, mm Hg	76.12 ± 12.39	77.21 ± 12.09	75.41 ± 12.53	<0.001

Data were shown as mean ± SD for continuous variables and n (%) for categorical variables. TC: total cholesterol; TG: triglycerides; HDLC: high-density lipoprotein cholesterol; LDLC: low-density lipoprotein cholesterol; BMI: body mass index; WC: waist circumference; SBP: systolic blood pressure; DBP: diastolic blood pressure.

Association of baseline and changes in SUA/Scr ratio with risk factors

SUA/Scr ratio was associated with several baseline characteristics in linear regression models (Supplemental Table 1). Smoking, drinking, hypoglycemic medications, HDLC and WC were independently associated with baseline SUA/Scr ratio, with a standardized β of −0.204 in smoking, 0.141 in drinking, −0.397 in hypoglycemic medications, −0.010 in HDLC and 0.009 in WC. The associations with change in SUA/Scr ratio were similar to those of baseline, but not in blood glucose, TC, LDLC, and SBP.

Associations of baseline and changes in SUA/Scr ratio with cognitive function decline

During 21007.9 person-years of follow-up, there were 1512 (45.8%) incident cases of cognitive function decline. As Table 2 showed, the age-, sex-, and education-adjusted hazard ratio (HR) for cognitive function decline was increased for the Q4 (1.162, 95% CI 1.007–1.342) baseline SUA/Scr ratio compared to the Q1 (p for trend = 0.022). In a multivariate-adjusted model for age, sex, education and a propensity score. Cognitive function decline was also associated with baseline SUA/Scr ratio: HR 1.175 (95% CI 1.015–1.360) for the Q4, compared with the Q1 (p for trend = 0.017). Baseline SUA/Scr ratio as per unit increase was positively associated with cognitive function decline (HR 1.049, 95% CI 1.013–1.087). Among 1129 participants attending waves 2, 3 and 4 or 5, a total of 115 (13.7%) participants developed cognitive function decline (after wave 3). During 9888.6 person-years of follow-up, in fully adjusted model, cumSUA/Scr ratio and its burden were both still associated with incident cognitive function decline [for cum SUA/Scr ratio per unit increase: HR (95% CI),1.036 (1.006–1.066); Q3 versus T1: 1.642 (1.031–2.615), Q4 versus T1:1.635 (1.006–2.656), p for trend = 0.025 and for cum SUA/Scr ratio burden per unit increase: HR (95% CI), 1.035 (1.005–1.065); High versus Low: 1.729 (1.212–2.466)]. In Figure 3, the restricted cubic splines showed that higher levels of baseline SUA/Scr ratio were associated with a progressively increased risk of cognitive function decline (p nonlinear = 0.514). Similar trends were also found in cum SUA/Scr ratio and its burden (Supplemental Figures 1 and 2).

Figure 3.

Adjusted cubic spline model of the association between the baseline SUA/Scr ratio and risk of cognitive function decline. SUA: serum uric acid; Scr: serum creatinine; HR: hazard ratio; CI: confidential interval.

Table 2.

Baseline SUA/Scr exposure indicators and prediction of cognitive function decline.

Exposure	Group	Model 1HR (95% CI)	p_{for trend}	Model 2HR (95% CI)	p_{for trend}	Model 3HR (95% CI)	p_{for trend}
Baseline SUA/Scr	Per unit increase	1.043 (1.008–1.079)		1.044 (1.010–1.080)		1.049 (1.013–1.087)
	Q1	Ref.	0.029	Ref.	0.022	Ref.	0.017
	Q2	1.038 (0.895–1.203)		1.037 (0.894–1.202)		1.040 (0.897–1.205)
	Q3	1.121 (0.973–1.292)		1.122 (0.973–1.293)		1.127 (0.977–1.299)
	Q4	1.153 (0.999–1.330)		1.162 (1.007–1.342)		1.175 (1.015–1.360)
Cumulative SUA/Scr	Per unit increase	1.026 (0.998–1.053)		1.026 (0.999–1.054)		1.036 (1.006–1.066)
	Q1	Ref.	0.079	Ref.	0.074	Ref.	0.025
	Q2	1.216 (0.749–1.974)		1.226 (0.755–1.992)		1.251 (0.770–2.003)
	Q3	1.537 (0.970–2.434)		1.547 (0.976–2.453)		1.642 (1.031–2.615)
	Q4	1.439 (0.902–2.293)		1.452 (0.911–2.316)		1.635 (1.006–2.656)
	Per unit increase	1.024 (0.097–1.052)		1.052 (0.998–1.053)		1.035 (1.005–1.065)
Cumulative SUA/Scr burden	Low	Ref.		Ref.		Ref.
Cumulative SUA/Scr burden	High	1.589 (1.130–2.261)		1.609 (1.137–2.276)		1.729 (1.212–2.466)

Model 1 was unadjusted. Model 2 was adjusted for age, sex and education levels. Model 3 was adjusted additionally adjusted for a propensity score. SUA: serum uric acid; Scr: serum creatinine; HR: hazard ratio; CI: confidence interval.

Sex, educational levels, smoking, drinking, hypertension, diabetes, dyslipidemia, BMI, and abdominal obesity were stratified for further analysis of the association between SUA/Scr ratio and cognitive function decline (Figure 4). The associations were not statistically significant in most subgroups, and no significant interactions were found in any subgroups.

Figure 4.

Associations between the SUA /Scr ratio and the risk of cognitive function decline in participants stratified by the sex, educational levels, smoking, drinking, hypertension, diabetes, dyslipidemia, BMI and abdominal obesity. Data were fully adjusted other than variables for stratification. Abdominal obesity was defined as WC ≥ 85 cm in males and ≥80 cm in females. HRs for per unit increased in SUA /Scr ratio. p-value for interaction was calculated. SUA: serum uric acid; Scr: serum creatinine; HR: hazard ratio; CI: confidence interval. Other abbreviations as in Table 1. * A small sample without an effective model convergence.

When excluding participants who had a history of stroke or psychological problems at baseline, the results did not change substantially. In model 3, Compared with Q1, participants with the highest SUA/Scr ratio group (Q4) were still significantly associated with an increased risk of cognitive function decline [HR (95% CI), 1.034 (1.005–1.064) for baseline SUA/Scr ratio, 1.033 (1.004–1.064) for cum SUA/Scr ratio and 1.719 (1.204–2.453) for cumSUA/Scr ratio burden]. SUA/Scr ratio (per unit increase) was related to a higher risk of cognitive function decline (Supplemental Table 2). Similar results were also found after excluding participants with a low cognitive score (<15 points) at baseline (Supplemental Table 3). There were no statistically significant associations between SUA/Scr ratio and cognitive function decline in hyperuricemia (Supplemental Table 4). An E-value was calculated to assess the potential impact of unaccounted confounders. An E-value of 1.22, with a lower bound of the 95% CI at 1.10. These confounding factors would need to have an HR of at least 1.22 to account for the association between baseline SUA/Scr ratio and cognitive function decline (Supplemental Figure 3).

Mediation analysis

We did not find any statistically significant associations between WBC count, CRP and the changes in cognitive test scores, and non-significant relationships were also found between WBC count, CRP and SUA/Scr ratio (all p ≥ 0.05, data not shown). Table 3 demonstrated the direct, indirect, and total effect of the SUA/Scr ratio on changes in cognitive test scores through WBC count and CRP. The direct effect of baseline SUA/Scr ratio though WBC count and CRP on MACE was [−0.165 (−0.276 – −0.053) and −0.157 (−0.268– −0.045)], respectively. No statistical significance was found in the direct effect of the cmuSUA/Scr ratio and its burden on changes in cognitive test scores through WBC count or CRP. A non-significant indirect effect through WBC count or CRP was found.

Table 3.

Mediation analysis of the effects of inflammatory indicators on the association of SUA/Scr with cognitive function score changes.

	Baseline SUA/Scr		Cumulative SUA/Scr		Cumulative SUA/Scr burden
Effect	WBC count	CRP	WBC count	CRP	WBC count	CRP
	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)
Total Effect	−0.164 (−0.276–0.052)	−0.158 (−0.269– −0.046)	−0.034 (−0.080–0.013)	−0.034 (−0.080–0.013)	−0.030 (−0.076–0.017)	−0.030 (−0.077–0.017)
Direct Effect	−0.165 (−0.276– −0.053)	−0.157 (−0.268– −0.045)	−0.035 (−0.081–0.012)	−0.031 (−0.078–0.015)	−0.031 (−0.077–0.016)	−0.028 (−0.074–0.019)
Mediated Effect	0.001 (−0.002–0.005)	−0.001 (−0.006–0.003)	0.001 (−0.002–0.004)	−0.003 (−0.010–0.001)	0.001 (−0.002–0.004)	−0.003 (−0.009–0.001)

Mediation analysis was adjusted for sex, age, education levels, marital status, smoking, drinking, hypertension, diabetes, stroke, heart problems, dyslipidemia, psychological problems, antihypertensive medications, hypoglycemic medications, lipid-lowering medications, BMI, waist, SBP, DBP, blood glucose, TC, TG, HDLC, and LDLC. WBC: white blood cell; CRP: C-reactive protein; CI: confidence interval. Other abbreviations are the same as in Table 1.

Discussion

This study investigated the association between SUA/Scr ratio and cognitive function decline in middle-aged participants (aged 45–60 years old) using a longitudinal national representative cohort dataset. SUA/Scr ratio, an indicator of endogenous UA, had a linear association with the risk of cognitive function decline. That is, both a single evaluation and cumulative exposure of the SUA/Scr ratio were associated with raised cognitive function decline risk, and this finding persisted even after adjusting for possible confounders. However, the above associations were weakened in most subgroups, with CIs included 1. We did not find any significant mediating effects of WBC count, CRP in SUA/Scr ratio-cognitive test score links

Previous findings of the effect of SUA on cognition were consistent. Some observational studies suggest that higher SUA was associated with poorer cognitive performance.^20–22 However, Chen et al. reported that high SUA was associated with reduced risks of mild cognitive impairment (MCI) among 3013 Chinese older adults. Global cognition in the above study was measured by using a Mini-Mental State Examination (MMSE) and then identifying MCI.²³ Consistently, two cross-sectional studies in China found that an appropriate increase in SUA could slow down the occurrence and development of MCI.^24,25 The National Health and Nutrition Examination Survey (2011–2014) data in America found a slight increase in uric acid within the normal range, which was beneficial for cognitive performance.²⁶ Interestingly, the results from CHALS also support the protective effect of a moderate increase in SUA on cognitive function in non-hyperuricemia or -normotensive adults,^27–29 which were inconsistent with our findings. RCS analysis showed a positive linear relationship between the SUA/Scr ratio and the risk of incident cognitive function decline without any significant inflexion points. The present study outcome definition of cognitive function (dichotomous variable) and the age of the study population (aged 45–60 years old) were different from those described above. Huang et al. found an interaction between baseline age and SUA levels, and SUA levels were not associated with cognitive performance in non-hyperuricemia adults aged <60 years.²⁸ More importantly, the above studies did not consider the confounding effect of renal function without including Scr or estimated glomerular filtration rate in fully adjusted multivariate models. In fact, baseline renal function-normalized SUA, which may reflect the net production of UA, will be better than SUA as the predictor of incident chronic kidney disease.³⁰ Although SUA testing has been normalized in community clinics, considering the impact of renal function when assessing the relationship between SUA and cognitive impairment is necessary and meaningful in daily healthcare.

Numerous previous studies focused on the relationship between a single measurement of SUA and cognitive performance.^31,32 In the present study, the cumulative exposure method was used to consider the change level of the SUA/Scr ratio over a certain period. Our findings suggest that the cumulative exposure of the SUA/Scr ratio significantly affected future cognitive function. Compared with the lowest quartile, the highest cumSUA/Scr group had a 63.5% increased risk of cognitive decline after adjusting for confounders, and a high burden of cmuSUA/Scr persisting for 4 years (wave1 to wave3) was associated with up to a 72.9% increased risk compared with a low burden. In addition to focusing on the SUA/Scr at a single time point, the cumulative exposure of SUA/Scr also deserved attention in the primary prevention of cognitive health. Unfortunately, only 1129 middle-aged adults consistently participated in waves 1 to 3 and 4 or 5 follow-ups (Figure 1). A 4-year interval between two SUA/Scr measurements could not adequately show the long-term SUA/Scr exposure trends, which should be considered when interpreting our results.

Previous studies showed that the SUA/Scr ratio was a predictor of chronic disease.^30,33,34 As we found, the high baseline SUA/Scr ratio and its cumulative exposure were associated with incident cognitive function decline. This relationship remained with participants without histories of stroke or psychological problems. Inconsistently, a cross-sectional study with a sample size of 3874 found that the baseline SUA/Scr ratio was negatively associated with the risk of cognitive impairment in Americans over 60 years of age.¹⁸ The type of cognitive test used or other factors such as sample size, study design, and characteristics of participants were different from ours. In line with present funding, SUA levels were not significantly associated with cognitive function in patients with hyperuricemia.²⁴ Abnormalities in UA metabolism and the use of UA-lowering medications may confound the association between SUA/Scr and cognitive function. In addition, the results may be unreliable due to insufficient statistical power due to a small sample size (n = 145). Noteworthily, the potential mechanisms that link UA to cognitive function effects are not fully understood. Inflammation defense plays a vital role in the above relationship.⁹ SUA was correlated with neutrophil count, CRP, IL-6, IL-1ra, IL-18 and TNF-α.³⁵ A positive relationship between higher CRP, IL-6 levels and SUA levels and between inflammatory markers and white matter hyper-intensity lower gray matter and hippocampal volumes were found,⁹ which motivated us to examine the association between SUA levels, inflammation, and cognitive function in a nationally prospective cohort. Regrettably, we did not find any significant connection between WBC count or CRP to SUA/Scr ratio or cognitive test score. The precipitation of urate crystals cloud release inflammatory factors such as IL-β1 and IL-6, leading to damage to the vascular endothelium. Lesions in the brain microvasculature led to cognitive impairment.³⁶ The 3157 (95.6%) participants with non-hyperuricemia in the present study may attenuate the pro-inflammatory effects of SUA. Therefore, it is meaningful to examine the complex interactions among SUA, cognition, and inflammation in patients with hyperuricemia in a large sample cohort while focusing on the levels of oxidative stress in vivo.⁹

The main strength of this study was that we used the data of CHARLS, a nationally representative, prospective cohort design and long-term follow-up survey dataset, to explore the association between SUA/Scr ratio and incident cognitive function decline. While effectively controlling for selection bias, a relatively large baseline sample of middle-aged participants (median age 53.0 years, interquartile range: 48.0, 56.0) was collected. This is important for gaining an insight into the preclinical stage of dementia.³⁷ In addition, unlike a single measurement in previous studies,³¹ the assessment of renal function-normalized SUA cumulative exposure could help to identify dynamic patterns during cognitive decline and reveal inherent potential value in predicting disease occurrence.²⁹ There are several limitations to our study. Firstly, the interpretation of the findings on the Chinese middle-aged population to other countries and age groups may be limited. Secondly, causality could not be inferred because of the potential residual confounders in observational study designs. We did not analysis some potential confounders such as the urate-lowering medications, APOE genotype and the contents and frequency of food intake due to the lack of related data in the CHARLS. Thirdly, CHARLS provided two biochemical data (wave1 and wave3) to the public, which could not adequately assess the long-term trend of SUA levels. In other words, we could not evaluate the longitudinal trajectory of SUA to identify potential categories of variation trends that would improve the accuracy of dynamic prediction. Finally, although the assessment tool for cognitive function test used was relatively simple without well-established cut-off values to identify cognitive impairment, the moderate-to-high validity of this cognitive test was reported compared with the MMSE and the Clinical Dementia Rating Scale.³⁸

Conclusion

This prospective cohort study demonstrated that the SUA/Scr ratio was positively associated with the risk of incident cognitive function decline in middle-aged participants, independent of potential confounders. This indicated that community doctors should focus on the renal function-normalized SUA indexes, such as the SUA/Scr ratio, in routine health management. We did not find a significant effect of the SUA/Scr ratio in cognition by regulating the levels of WBC count or CRP. Further studies with various ethnic were needed to verify this potential mechanism.

Supplemental Material

sj-docx-1-alz-10.1177_13872877241303789 - Supplemental material for Effect of serum uric acid to creatinine ratio on cognitive function decline in middle-aged adults: Longitudinal evidence from CHARLS

Supplemental material, sj-docx-1-alz-10.1177_13872877241303789 for Effect of serum uric acid to creatinine ratio on cognitive function decline in middle-aged adults: Longitudinal evidence from CHARLS by Xiangjun Zhou, Zhe Wang, Yi Zhu, Shuang Feng, Haijian Wu, Dongliang Zhu, Zheyuan Wu and Qingjun Kao in Journal of Alzheimer's Disease

Footnotes

Acknowledgments

The authors would like to express genuine gratitude to the participants and staff of the CHARLS, who contributed greatly to the academic community and made this study possible.

ORCID iDs

Xiangjun Zhou

Zhe Wang

Haijian Wu

Author contributions

Xiangjun Zhou (Data curation; Formal analysis; Methodology; Software; Writing – original draft; Writing – review & editing); Qingjun Kao (Methodology; Software; Supervision); Zhe Wang (Methodology; Software; Supervision); Yi Zhu (Formal analysis); Shuang Feng (Software); Haijian Wu (Data curation; Formal analysis); Dongliang Zhu (Data curation; Methodology); Zheyuan Wu (Data curation; Software).

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Hangzhou Health Science and Technology Projects [Grant # A20230326].

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data availability

Researchers can gain access to the CHARLS data ().

Supplemental material

Supplemental material for this article is available online.

References

Fang

Xie

Schenkel

, et al. A research agenda for ageing in China in the 21st century (2nd edition): focusing on basic and translational research, long-term care, policy and social networks. Ageing Res Rev 2020; 64: 101174.

Jia

Quan

, et al. Dementia in China: epidemiology, clinical management, and research advances. Lancet Neurol 2020; 19: 81–92.

Elliott

. MRI-based biomarkers of accelerated aging and dementia risk in midlife: how close are we? Ageing Res Rev 2020; 61: 101075.

Strittmatter

Sunde

Zegners

. Life cycle patterns of cognitive performance over the long run. Proc Natl Acad Sci U S A 2020; 117: 27255–27261.

Dohm-Hansen

English

Lavelle

, et al. The ‘middle-aging’ brain. Trends Neurosci 2024; 47: 259–272.

Therriault

Schindler

Salvadó

, et al. Biomarker-based staging of Alzheimer disease: rationale and clinical applications. Nat Rev Neurol 2024; 20: 232–244.

Ndrepepa

. Uric acid and cardiovascular disease. Clin Chim Acta 2018; 484: 150–163.

Latourte

Bardin

Richette

. Uric acid and cognitive decline: a double-edge sword? Curr Opin Rheumatol 2018; 30: 183–187.

Tana

Ticinesi

Prati

, et al. Uric acid and cognitive function in older individuals. Nutrients 2018; 10: 975.

10.

Yuan

Yin

. Uric acid and the risk of Alzheimer’s disease: is there an association? Am J Alzheimers Dis Other Demen 2016; 31: 294–295.

11.

Bobulescu

Moe

. Renal transport of uric acid: evolving concepts and uncertainties. Adv Chronic Kidney Dis 2012; 19: 358–371.

12.

Zhao

Smith

, et al. Cohort profile: the China health and retirement longitudinal study (CHARLS). Int J Epidemiol 2014; 43: 61–68.

13.

Luo

Zhang

, et al. Obesity and the onset of depressive symptoms among middle-aged and older adults in China: evidence from the CHARLS. BMC Public Health 2018; 18: 909.

14.

CHARLS, https://charls.charlsdata.com/ (2011, accessed 10 July 2024).

15.

Guan

Shi

, et al. Cumulative non-high-density lipoprotein cholesterol burden and risk of atherosclerotic cardiovascular disease: a prospective community-based study. Front Cardiovasc Med 2023; 10: 1105342.

16.

McArdle

Fisher

Kadlec

. Latent variable analyses of age trends of cognition in the health and retirement study, 1992–2004. Psychol Aging 2007; 22: 525–545.

17.

, et al. Association between solid cooking fuel and cognitive decline: three nationwide cohort studies in middle-aged and older population. Environ Int 2023; 173: 107803.

18.

Chen

Tong

. Association between the serum uric acid/serum creatinine ratio and cognitive function in older adults: NHANES in the United States. Sci Rep 2024; 14: 16312.

19.

Veronese

Stubbs

Trevisan

, et al. Hyperuricemia is a significant predictor of poor physical performance in the elderly: THE Pro.V.A. Study. Arthritis Care Res (Hoboken) 2017; 69: 1238–1244.

20.

Khaled

Abdelhamid

Al-Mazroey

, et al. Higher serum uric acid is associated with poorer cognitive performance in healthy middle-aged people: a cross-sectional study. Intern Emerg Med 2023; 18: 1701–1709.

21.

Ouyang

Xiong

, et al. Elevated serum uric acid is associated with cognitive impairment in acute minor ischemic stroke patients. Heliyon 2023; 9: e21072.

22.

Tong

Zhang

, et al. Uric acid index is a risk for mild cognitive impairment in type 2 diabetes. Hormones (Athens) 2023; 22: 425–439.

23.

Chen

, et al. High blood uric acid is associated with reduced risks of mild cognitive impairment among older adults in China: a 9-year prospective cohort study. Front Aging Neurosci 2021; 13: 747686.

24.

Liu

Wang

Zeng

, et al. Relationship between serum uric acid level and mild cognitive impairment in Chinese community elderly. BMC Neurol 2017; 17: 146.

25.

Xue

Liu

Xue

, et al. Low uric acid is a risk factor in mild cognitive impairment. Neuropsychiatr Dis Treat 2017; 13: 2363–2367.

26.

Geng

Zhang

Liu

, et al. Elevated serum uric acid is associated with cognitive improvement in older American adults: a large, population-based-analysis of the NHANES database. Front Aging Neurosci 2022; 14: 1024415.

27.

Zhang

Jia

, et al. The association between serum uric acid and cognitive function among middle-aged and older adults without hyperuricemia: the mediating role of depressive symptoms. J Gerontol A Biol Sci Med Sci 2024; 79: 203.

28.

Huang

Zhang

Shen

, et al. Association of plasma uric acid levels with cognitive function among non-hyperuricemia adults: a prospective study. Clin Nutr 2022; 41: 645–652.

29.

Wang

Jin

, et al. Moderate increase of serum uric acid within a normal range is associated with improved cognitive function in a non-normotensive population: a nationally representative cohort study. Front Aging Neurosci 2022; 14: 944341.

30.

Huang

, et al. Serum uric acid to creatinine ratio: a predictor of incident chronic kidney disease in type 2 diabetes mellitus patients with preserved kidney function. Diab Vasc Dis Res 2017; 14: 221–225.

31.

Khan

Quinn

Hewitt

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

32.

Liu

Peng

Yang

, et al. Uric acid levels and risk of cognitive impairment: dose-response meta-analysis of prospective cohort studies. PLoS One 2023; 18: e0293832.

33.

Wang

Tian

, et al. Metabolic factors mediate the association between serum uric acid to serum creatinine ratio and cardiovascular disease. J Am Heart Assoc 2021; 10: e023054.

34.

Al-Daghri

Al-Attas

Wani

, et al. Serum uric acid to creatinine ratio and risk of metabolic syndrome in Saudi type 2 diabetic patients. Sci Rep 2017; 7: 12104.

35.

Ruggiero

Cherubini

Ble

, et al. Uric acid and inflammatory markers. Eur Heart J 2006; 27: 1174–1181.

36.

Perez-Ruiz

Becker

MA.

Inflammation: a possible mechanism for a causative role of hyperuricemia/gout in cardiovascular disease. Curr Med Res Opin 2015; 31(Suppl 2): 9–14.

37.

Sperling

Mormino

Johnson

. The evolution of preclinical Alzheimer’s disease: implications for prevention trials. Neuron 2014; 84: 608–622.

38.

Meng

Wang

Strauss

, et al. Validation of neuropsychological tests for the China health and retirement longitudinal study harmonized cognitive assessment protocol. Int Psychogeriatr 2019; 31: 1709–1719.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.17 MB