Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Feb 20.
Published in final edited form as: Sci Total Environ. 2020 Nov 26;756:144072. doi: 10.1016/j.scitotenv.2020.144072

Higher blood cadmium level is associated with greater cognitive decline in rural Chinese adults aged 65 or older

Hang Liu 1, Liqin Su 1,*, Xi Chen 1, Sisi Wang 2, Yibin Cheng 1, Shaobin Lin 1, Liang Ding 1, Jingyi Liu 1, Chen Chen 1, Frederick W Unverzagt 3, Ann M Hake 4, Yinlong Jin 1, Sujuan Gao 5
PMCID: PMC7775354  NIHMSID: NIHMS1652276  PMID: 33280862

Abstract

Cadmium (Cd) exposure has been reported to have neurotoxic effects in animal studies and associated with increased Alzheimer’s Disease mortality and lower cognitive function in cross-sectional and case-control studies. However, no results from longitudinal studies on Cd and cognitive decline are available. In this prospective cohort study, we recruited 1867 participants aged 65 years or older from rural areas in China, blood Cd and cognitive function were measured at baseline (2010–2012), and 1554 participants completed cognitive function tests during a 3-year follow-up (2013–2015). Cognitive function was evaluated using nine standardized cognitive tests: The Community Screening Instrument for Dementia, the CERAD Word List Learning, Word list recall, IU Story Recall, Animal Fluency Test, Boston Naming Test, Stick Design, Delayed Stick Design and the IU Token Test. Analysis of covariance models and logistic regression models were used to determine the association between Cd and standardized cognitive decline adjusting for covariates. The median blood Cd concentration of this study population was 2.12 μg/L, and the interquartile range was 1.42–4.64 μg/L. Significant association of higher Cd levels with lower cognitive scores were observed in five individual cognitive tests (Delayed Stick Design, Boston Naming Test, CERAD Word List Learning Test, Word List Recall Test and IU Story Recall Test) and the composite cognitive score adjusting for multi-covariates at baseline. Higher Cd levels were significantly associated with greater 3-year cognitive decline in Delayed Stick Design Test, Boston Naming Test, IU Token Test, Word List Recall Test and Composite cognitive score. For these cognitive tests, participants in the top two Cd quartile group had significantly greater decline than those in the lowest Cd quartile group, while the two lowest Cd quartile groups were not significantly different. Our findings suggest that higher Cd exposure is associated with greater cognitive decline in older Chinese adults.

Keywords: Cadmium, Cognitive function, Cognitive decline, Older adults, Prospective cohort study

1. Introduction

Cadmium (Cd) is an accumulative heavy metal that cannot be degraded in the environment and has been classified as a human carcinogen (Mezynska & Brzóska, 2018). In recent decades, greater amount of Cd has been discharged into the environment mainly by industrial and agricultural activities, resulting in an increasing trend of Cd pollution in the air, water, soil, and food (Garner and Levallois, 2016; Onakpa et al., 2018). Cd exposure has become a serious health risk factor for the Chinese population since approximately 19.4% of agricultural soil in China has reported to be contaminated with Cd (Zhao et al., 2015), which could lead to greater Cd exposure through food supply (Satarug et al., 2017) and tobacco use (Satarug & Moore, 2004). Recently, a higher level of Cd burden was found to associate with elevated Alzheimer’s Disease mortality in older US adults (Min & Min 2016; Peng et al. 2017). Cd-related cognitive impairment has attracted increasing concerns in the context of aging (Huat et al., 2019).

Toxicological studies have indicated that Cd can be transported directly through the blood-brain barrier and eventually accumulates in the brain activating various signaling pathways involved in inflammation, oxidative stress and neuronal apoptosis (Huat et al., 2019; Wang and Du, 2013). Cd can also induce β- and phosphorylated tau protein production and cell death through heat shock protein in SN56 cholinergic neurons (Moyano et al., 2018). Animal studies have shown that Cd exposure can significantly affect peripheral and central nervous system functions and present a wide range of clinical symptoms, including memory decline, mental retardation, learning disabilities and movement disorders (Branca et al., 2018).

Evidence on the association between Cd and cognitive function from epidemiological studies, however, is still limited and ambiguous so far. Some studies indicated that Cd exposure was related to adverse neurodevelopment of children (Gustin et al., 2018; Rodríguez-Barranco et al., 2014; Zhou et al., 2020), and that higher blood cadmium was associated with worse cognitive function in the elderly (Gao et al., 2008; Li et al., 2018; Souza-Talarico et al., 2017). There were also studies reporting non-significant association between Cd exposure and cognitive function in children (Cao et al., 2009) or in the elderly (Basun et al., 1994; Nordberg et al., 2000; Park et al., 2014). No longitudinal study on the association between Cd exposure and cognitive decline has been reported thus far.

In our previous small cross-sectional study, 188 participants with cognitive assessments had provided blood samples and trace elements including Cd were measured between December 2003 and May 2005. We observed that increasing blood Cd was significantly associated with lower composite cognitive score in adults aged 65 years or older (Gao et al., 2008). From 2010 to 2015, we had recruited a larger sample and measured blood Cd concentrations at baseline and cognitive scores at baseline and 3-year follow-up. Here we report on results from this longitudinal study on the relationship between Cd exposure and 3-year cognitive decline in this cohort.

2. Methods

2.1. Study Design and Participants

We conducted a prospective cohort study in four rural areas from southwestern and eastern China. Details of the study design were described previously (Gao et al., 2007; Gao et al., 2014). Residents aged 65 years or older from four rural counties were asked to join in the cohort if they met the following eligibility criteria: (a) had lived in the county for at least 30 years; (b) had no language communication problem; (c) agreed to complete a face-to-face questionnaire and provide blood samples. All participants signed written informed consents before each interview. The study was approved by the Indiana University Institutional Review Board and the Institute for Environmental Health and Related Safety, Chinese Center for Disease Control and Prevention.

From October 2010 to November 2012, a total of 1867 participants completed the cognitive evaluation and provided blood samples for cadmium measurements. We conducted a follow-up survey 3 years later (2013–2015), with 1554 (83.2%) completed cognitive function tests. 313 participants were not able to participate in the follow-up survey, most of whom were deceased, had moved out of the study area, had severe hearing loss, or other diseases which makes it impossible to conduct the cognitive assessment.

2.2. Cognitive Assessment

Nine cognitive tests were used at baseline and follow-up: the Community Screening Instrument for Dementia (CSID), the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) Word List Learning Test, Word list recall Test, IU Story Recall Test, Animal Fluency Test, Boston Naming Test, Stick Design Test, Delayed Stick Design Test, and IU Token Test. More details of the 9 instruments have been described in our previous cross-sectional study from the same cohort (Cheng et al., 2014). Cognitive evaluations were carried out in a quiet space to prevent interference. Interviewers were well trained before each wave of cognitive evaluation. Supervisors worked with interviewers in the entire interview process. In order to maintain consistent cognitive evaluation quality, almost the same group of interviewers remained for conducting cognitive evaluation at baseline and follow-up with less than 20% of new interviewers at follow-up.

CSID scores range from 0–30. The CERAD Word List Learning Test scores range from 0–30, and the CERAD Word List Recall Test scores range from 0–10. The IU Story Recall Test scores range from 0–14. Animal Fluency Test scores were the number of animals the participants could name in 60 seconds. The IU Token Test scores range from 0–24. The Boston Naming Test scores range from 0–20. Both Stick Design Test and Delayed Stick Design Test scores range from 0–12. A standardized score for each cognitive test was created by subtracting the baseline sample mean and dividing it by the baseline sample standard deviation. A composite cognitive Z score was derived using the average of standardized scores of the nine cognitive tests.

2.3. Blood Cadmium Measures

Fasting peripheral blood samples were collected in the morning using 5-mL purple top (EDTA) vacutainer tubes during the 2010–2012 baseline evaluation. All samples were stored in polypropylene tubes at −80 °C before laboratory analysis. Whole blood Cd concentrations were measured by inductively-coupled plasma mass spectrometry (ICP-MS) with an ELAN® DRC II (PerkinElmer, Norwalk, CT). Further methodological details of the laboratory analyses were described elsewhere (Wang et al., 2019). The limit of detection (LOD) for Cd was 0.1 μg/L. Each batch (20 samples) contained one positive sample (the Seronorm trace elements whole blood L-2 (LOT 1003192)) and one negative sample (deionized water) for quality control purposes. To monitor the testing procedure, blank samples were analyzed every 20 samples in all batches, and 12% duplicated samples were also included. Intra-assay and inter-assay coefficient of variances (CV) ranged from 0.23% to 2.53%. Samples were randomly assigned to batches.

2.4. Covariates

Covariates with potential association with cognitive function shown in previous literatures were included in our models. Information on sociodemographic characteristics (age, gender, education, marital status, and household composition), lifestyle factors (alcohol consumption and smoking) and medical history of illness was collected from questionnaires. Body Mass Index (BMI) was derived from height and weight measurements. Hypertension was defined as measured systolic blood pressure 140 mmHg or higher, diastolic blood pressure 90 mmHg or higher, or reported use of anti-hypertensive medication. APOE genotype (ε4 carriers or non-carriers) was determined by the multiplex tetra-primer amplification refractory mutation system (T-ARMS) PCR reactions (Kim et al., 2010). Serum Glucose, total cholesterol, triglycerides, high-density lipoprotein-cholesterol and low-density lipoprotein-cholesterol were measured using the Roche Diagnostic Kits by Hitachi Automatic Biochemistry Analyzer 9700. Type 2 diabetes mellitus (T2DM) was defined as serum glucose≥7.0mmol/L. Four types of dyslipidemia, hypercholesterolemia (HTC), hypertriglyceridemia (HTG), low high-density lipoprotein cholesterol level (LHDLC), and high low-density lipoprotein cholesterol level (HLDLC), were defined according to Chinese Guidelines on Prevention and Treatment of Dyslipidemia in Adults (Joint committee for developing Chinese guidelines on prevention and treatment of dyslipidemia in adults, 2007).

2.5. Statistical analysis

Statistical analyses were performed using SPSS 19.0. P values less than 0.05 were considered to be statistically significant. Age and BMI were presented as mean ± standard deviation (SD), and Cd level was presented as median (interquartile range). Comparisons of demographic characteristics between participants with 3 year follow-up and those lost to follow-up were performed by t-tests for continuous variables or chi-square tests for categorical variables. Comparisons of baseline characteristics among Cd quartile groups were performed by analysis of variance (ANOVA) and Chi-square test. Analysis of covariance models (ANCOVA) were used to compare the differences of baseline cognitive scores among Cd quartile groups controlling for age, gender, marriage status, house holding status, education, APOE genotype, BMI, smoking, alcohol consumption, prevalence of hypertension, T2DM, HTC, HTG, LHDLC, HLDLC. For changes in cognitive function, we included the baseline score of each test as an additional covariate in the ANCOVA models. The odds ratios (OR) were calculated using logistic regression models to assess the risk of high Cd levels to the cognitive decline, defined as more than 10% or 20% decrease from the baseline score of each test, respectively (Akbaraly et al., 2007; Niti et al., 2008).

3. Results

3.1. Description of participants

In this study, a total of 1867 participants were evaluated at baseline and 1554 subjects completed all follow-up tests 3 years later. For the 1867 participants, the average age was 73.79±5.92 years, 53.24% were females. The median blood cadmium concentration was 2.12 μg/L, and the interquartile range (IQR) was 1.42–4.64 μg/L. In Table 1, we present baseline characteristics of study participants by quartile groups of blood Cd levels. Participants did not differ in age among Cd quartile groups, but the differences in gender, BMI, education, house holding status, alcohol consumption, smoking, APOE genotype and history of HTC, HTG and LHDLC were statistically significant (P<0.05) among the four groups. In general, individuals in the highest Cd quartile group had more medical disorders than those in the lowest quartile group.

Table 1.

Baseline characteristics of study participants by cadmium quartile groups

Quartile groups by blood Cd level (μg/L)
Overall (n=1867) Q1 (≤1.42) (n=467) Q2 (1.42–2.12) (n=466) Q3 (2.12–4.64) (n=468) Q4 (≥4.64) (n=466) P values
Mean age, years 73.79±5.92 73.36±5.78 73.92±6.02 73.96±5.61 73.79±6.24 0.343a
Female (%) 53.24 66.81 58.58 45.94 41.63 <0.001
BMI, kg/m2 23.10±3.76 24.03±4.02 23.33±3.86 22.74±3.63 22.2±3.28 <0.001a
Attended school (%) 49.54 40.69 45.71 52.78 59.01 <0.001
Marital status (%) 0.145
 Married 63.69 66.60 61.59 63.03 63.52
 Unmarried 34.76 32.33 36.91 36.11 33.69
 Others 1.55 1.07 1.50 0.85 2.79
House holding status (%) <0.001
 Live with spouse 61.33 64.67 61.16 59.40 60.09
 Live with others 19.01 13.06 16.09 22.01 24.89
 Live alone 18.75 20.99 21.46 18.38 14.16
 Others 0.91 1.28 1.29 0.21 0.86
Alcohol consumption (%) 38.73 32.33 36.05 43.80 42.70 <0.001
Smoking (%) 40.01 22.06 27.90 53.85 56.22 <0.001
Prevalence of (%)
 T2DM 8.52 9.83 6.99 7.81 9.45 0.363
 Hypertension 63.36 66.60 62.66 62.82 61.51 0.399
 HTC 17.73 12.88 14.19 22.34 22.81 <0.001
 HTG 12.00 9.83 8.52 14.32 16.23 <0.001
 LHDLC 32.14 36.68 35.81 30.80 27.63 0.010
 HLDLC 13.02 11.35 11.57 13.88 16.23 0.101
APOE ε4 carriers (%) 15.85 12.63 13.95 19.44 17.38 0.017
Open in a new tab
a

oneway ANOVA

Comparisons of baseline demographic characteristics between the 1554 participants who completed 3-year follow-up and the 313 participants who did not were presented in Table 2. Of the 313 lost to follow-up participants, 213 were dead, 54 had moved, and 46 had severe hearing loss or other diseases during follow-up survey. No significant difference in blood Cd concentration was observed between the two groups. However, those who were lost to follow-up were significantly older, had lower BMI, education and lower composite cognitive score at baseline.

Table 2.

Comparisons of Demographic Characteristics between the 1554 Follow-up Participants and 313 Lost to Follow-up Participants

Follow-up (n=1554) Lost to Follow-up (n=313) P value
Mean age, years 73.10±5.55 77.25±6.47 <0.001
Female (%) 53.93% 49.84% 0.186
BMI, kg/m2 23.21±3.70 22.51±4.04 0.003
Ever attended school (%) 50.90% 42.81% 0.009
Alcohol consumption (%) 39.38% 35.46% 0.194
Smoking (%) 39.45% 42.81% 0.268
Blood Cd (μg/L) 5.76±18.34 5.07±8.30 0.512
Baseline Composite Z score 0.07±0.67 −0.30±0.75 <0.001
Open in a new tab

3.2. Cd level and baseline cognitive function

We examined the cross-sectional relationship between blood Cd level and cognitive functions by comparing baseline cognitive scores among Cd quartile groups. Crude (the upper row) and adjusted scores (the lower row) were presented in Table 3. Significant associations of higher Cd levels with lower cognitive scores were observed in five cognitive tests (Delayed Stick Design Test, Boston Naming Test, CERAD Word List Learning Test, Word List Recall Test and IU Story Recall Test) and the composite cognitive score.

Table 3.

Mean crude and adjusted baseline cognitive scores by cadmium quartile groups

Q1 (n=467) Q2 (n=466) Q3 (n=468) Q4 (n=466) F P values
Animal Fluency Test crude 13.23±4.84 12.92±4.80 13.51±4.64 13.61±5.36
adjusted 13.60±0.22 13.18±0.22 13.27±0.22 13.22±0.22 0.743 0.526
IU Story Recall Test crude 4.76±3.33 4.66±3.35 3.78±3.43 4.39±3.49
adjusted 4.84±0.15 4.76±0.15 4.73±0.15 4.26±0.15a,b,c 2.785 0.040
CSID Test crude 25.76±3.40 25.73±3.70 26.03±3.46 25.82±3.54
adjusted 25.97±0.15 25.92±0.15 25.88±0.15 25.55±0.15 1.577 0.193
Stick Design Test crude 10.34±2.56 10.42±2.38 10.60±2.35 10.45±2.55
adjusted 10.48±0.11 10.52±0.11 10.52±0.11 10.30±0.11 0.999 0.392
Delayed crude 5.03±2.59 4.92±2.59 5.33±2.80 4.90±2.75
Stick Design Test adjusted 5.18±0.12 5.04±0.12 5.23±0.12 4.73±0.12a,c 3.769 0.010
Boston Naming Test crude 13.72±3.60 13.75±3.63 13.71±3.36 13.74±3.45
adjusted 13.97±0.14 13.98±0.13 13.54±0.13a,b 13.42±0.14a,b 4.088 0.007
CERAD Word List crude 13.61±4.53 13.32±4.52 13.28±4.37 12.64±4.24
Learning Test adjusted 13.74±0.20 14.43±0.20 13.22±0.20 12.46±0.20a,b,c 6.673 <0.001
IU Token Test crude 16.67±5.29 16.60±5.18 16.84±5.19 16.77±5.02 0.321 0.810
adjusted 16.79±0.22 16.80±0.22 16.75±0.22 16.53±0.22
Word List Recall Test crude 4.69±2.32 4.49±2.88 4.51±2.20 4.03±2.33
adjusted 4.73±0.10 4.55±0.10 4.48±0.10 3.95±0.11a,b,c 9.626 <0.001
Composit e Z Score crude 0.03±0.71 0.00±0.69 0.05±0.68 0.00±0.72
adjusted 0.07±0.03 0.04±0.03 0.02±0.03 −0.09±0.03a,b,c 6.357 <0.001
Open in a new tab

Adjusted for age, gender, marriage status, house holding status, education, APOE genotype, BMI, smoking, alcohol consumption, history of T2DM, HBP, HTC, HTG, LHDLC and HLDLC.

a

compared with Q1, P<0.05

b

compared with Q2, P<0.05

c

compared with Q3, P<0.05

3.3. Cd level and cognitive decline

Table 4 and Figure 1 present results of ANCOVA models for changes in cognitive scores. Higher Cd levels were significantly associated with greater cognitive decline in Delayed Stick Design Test, Boston Naming Test, IU Token Test, Word List Recall and Composite cognitive score. For these cognitive tests, participants in the top two Cd quartile groups had significantly greater decline than those in the lowest quartile group, while the bottom two Cd quartile groups were not significantly different from each other.

Table 4.

Covariance analysis of the difference of cognitive function changes among different Cd level groups

Q1 (n=394) Q2 (n=383) Q3 (n=389) Q4 (n=388) F P values
Animal Fluency Test crude −1.51±4.70 −1.15±5.12 −1.20±5.27 −0.67±5.65
adjusted −1.24±0.22 −1.12±0.22 −1.36±0.22 −0.74±0.22 1.514 0.209
IU Story Recall Test crude −0.26±3.15 −0.20±3.42 −0.41±3.11 −0.23±3.59
adjusted −0.14±0.14 −0.13±0.14 −0.38±0.14 −0.45±0.15 1.138 0.332
CSID Test crude −0.30±3.14 −0.12±3.08 −0.46±3.16 −0.07±3.19
adjusted −0.20±0.14 −0.07±0.14 −0.41±0.14 −0.28±0.15 0.956 0.413
Stick Design Test crude 0.14±2.17 −0.05±2.61 −0.11±2.50 −0.05±2.18
adjusted 0.12±0.10 0.00±0.10 −0.11±0.10 −0.08±0.10 0.930 0.425
Delayed crude 0.33±2.45 0.21±2.76 −0.12±2.86 0.24±2.63
Stick Design Test adjusted 0.43±0.12 0.25±0.12 −0.02±0.12a 0.00±0.12a 2.808 0.038
Boston Naming Test crude 0.46±2.10 0.15±2.14 0.00±2.17 −0.22±1.98
adjusted 0.50±0.10 0.23±0.10 −0.07±0.10a,b 0.16±0.11a 4.770 0.003
CERAD crude 0.21±3.96 0.34±4.34 0.02±3.97 0.17±4.29
Word List Learning Test adjusted 0.43±0.19 0.38±0.19 0.10±0.19 −0.18±0.19 1.986 0.114
IU Token Test crude −0.17±4.33 −0.43±4.51 −0.58±4.80 −0.66±4.42
adjusted 0.00±0.20 −0.29±0.21 −0.70±0.20a −0.85±0.21a 3.156 0.024
Word List Recall Test crude 0.19±2.16 0.11±2.21 −0.19±2.13 0.11±2.27
adjusted 0.30±0.10 0.16±0.10 −0.13±0.10a,b −0.12±0.10a 4.078 0.007
Composite Z Score crude −0.04±0.40 −0.06±0.46 −0.12±0.45 −0.05±0.44
adjusted −0.02±0.02 −0.05±0.02 −0.12±0.02a,b −0.08±0.02 3.567 0.014
Open in a new tab

Adjusted for age, gender, marriage status, house holding status, education, APOE genotype, BMI, smoking, alcohol consumption, history ofT2DM, HBP, HTC, HTG, LHDLC, HLDLC and baseline score of each corresponding test.

a

compared with Q1, P<0.05

b

compared with Q2, P<0.05

c

compared with Q3, P<0.05

Figure 1.

Figure 1

Open in a new tab

Table 5 showed logistic regression analysis results after controlling for covariates. Cd level was found to be independently associated with cognitive decline in several tests. Compared to participants in the lowest quartile group (Q1), individuals in the Q2 group were more likely to show cognitive decline in Boston Naming Test using both 10% and 20% cutoffs (OR=1.464 for 10% decline, OR=1.894 for 20% decline, P<0.05). Individuals in the Q3 group were more likely to have cognitive decline on five tests (Stick Design Test, Delayed Stick Design Test, Boston Naming Test, IU Token Test and Composite Z score) at 10% cutoff and three tests (Delayed Stick Design Test, Boston Naming Test and Composite Z score) at 20% cutoff. Although non-significant, Q4 group was also likely to suffer from cognitive decline compared with Q1 group in most tests except the Animal Fluency Test at 10% and 20% cutoff, similar to Q2 and Q3.

Table 5.

Logistic regression analysis for the association between Cd level and 3-year cognitive decline

≥10% ≥20%
OR (95% confidence interval) P for trend OR (95% confidence interval) P for trend
N (%) Q1 Q2 Q3 Q4 N (%) Q1 Q2 Q3 Q4
Decline in Animal Fluency Test 751 (48.33%) - 0.95 (0.71, 1.27) 0.90 (0.66, 1.21) 0.77 (0.55, 1.05) 0.099 570 (36.68%) - 0.87 (0.65, 1.18) 0.94 (0.69, 1.28) 0.72 (0.52, 1.00) 0.084
Decline in IU Story Recall Test 788 (50.81%) - 1.15 (0.86, 1.54) 1.17 (0.86, 1.58) 1.17 (0.86, 1.60) 0.322 722 (46.55%) - 1.03(0.77, 1.38) 1.12 (0.83, 1.52) 1.22 (0.90, 1.67) 0.180
Decline in CSID Test 320 (20.59%) - 0.81 (0.56, 1.18) 1.38 (0.95, 2.00) 1.13 (0.77, 1.68) 0.193 87 (5.60%) - 0.69 (0.36, 1.35) 1.06 (0.54, 2.07) 1.49 (0.78, 2.85) 0.165
Decline in Stick Design Test 284 (18.59%) - 1.46 (0.99, 2.15) 1.62* (1.08, 2.43) 1.40 (0.92, 2.13) 0.108 205 (13.42%) - 1.47 (0.96, 2.26) 1.56 (0.99, 2.47) 1.07 (0.65, 1.75) 0.683
Decline in Delayed Stick Design Test 612 (40.10%) - 1.24 (0.92, 1.67) 1.48* (1.08, 2.02) 1.32 (0.96, 1.81) 0.060 492 (32.24%) - 1.24 (0.90, 1.71) 1.50* (1.08, 2.09) 1.25 (0.89, 1.76) 0.130
Decline in Boston Naming Test 300 (19.49%) - 1.46* (1.00, 2.14) 1.86* (1.26, 2.77) 1.33 (0.88, 2.02) 0.101 119 (7.73%) - 1.89* (1.07, 3.36) 2.08* (1.14, 3.80) 1.43 (0.75, 2.71) 0.273
Decline in CERAD Word List Learning Test 552 (35.96%) - 1.07 (0.79, 1.46) 1.08 (0.79, 1.48) 1.04 (0.75, 1.44) 0.814 362 (23.5 – 8%) - 0.99 (0.69, 1.40) 1.02 (0.71, 1.46) 1.22 (0.84, 1.73) 0.313
Decline in IU Token Test 532 (34.95%) - 1.05 (0.77, 1.44) 1.41* (1.02, 1.95) 1.30 (0.94, 1.81) 0.053 336 (22.08%) - 1.03 (0.71, 1.48) 1.42 (0.98, 2.06) 1.27 (0.87, 1.87) 0.107
Decline in Word List Recall Test 638 (41.56%) - 1.01 (0.75, 1.36) 1.29 (0.95, 1.76) 0.92 (0.67, 1.27) 0.966 493 (32.12%) - 0.93 (0.68, 1.29) 1.31 (0.95, 1.82) 1.07 (0.76, 1.50) 0.373
Decline in Composite Z Score 430 (28.63%) - 1.25 (0.89, 1.74) 1.42* (101, 2.02) 1.30 (0.91, 1.86) 0.121 173 (11.52%) - 1.46 (0.89, 2.39) 1.94* (1.18, 3.20) 1.31 (0.77, 2.25) 0.205
Open in a new tab
*

P<0.05. Adjusted for age, gender, marriage status, house holding status, education, APOE genotype, BMI, smoking, alcohol consumption, T2DM, HBP, HTC, HTG, LHDLC and HLDLC

4. Discussion

In this prospective cohort study, we examined the association between blood Cd level and cognitive decline in a 3-year follow up of rural Chinese adults aged 65 years or older. We observed that individuals in higher Cd quartile groups had lower baseline cognitive scores and greater cognitive decline 3 years later. Blood Cd level was associated with greater cognitive decline in composite cognitive scores and 4 tests (Delayed Stick Design Test, Boston Naming Test, IU token Test and Word List Recall Test) covering cognitive domains of visual-spatial, language, executive and working memory.

Our results are consistent with several previous studies. A cross-sectional study in Brazil examined Cd concentrations in blood samples from 125 adults aged 50–82 years old (Souza-Talarico et al., 2017). The Counting Span Test (CST) was used to evaluate working memory capacity. The mean blood Cd concentration in the study participants was 0.9±1.12 μg/L. The study reported that higher blood Cd concentration in older adults was associated with worse working memory function. Another cross-sectional study analyzed data from NHANES (2011–2014) in the US involving 2068 adults aged 60 years and older (Li et al., 2018). Four cognitive tests were used to evaluate cognitive function. An inverse association between blood Cd levels and composite cognitive score was observed controlling for potential confounding factors. The median blood Cd concentration in the study participants was 0.35 μg/L, and the IQR was 0.24~0.56 μg/L. These findings support the hypothesis that higher blood cadmium was associated with worse cognitive function in older adults.

However, other population-based studies have reported different results. Results from a cross-sectional study of 804 subjects with an average age of 87 years showed non-significant relationship between blood Cd and cognitive function using the Mini-Mental State Examination instrument (Nordberg et al., 2000). Interestingly, a case control study involving 89 Alzheimer’s Disease (AD) patients and 118 cognitively normal individuals reported that serum Cd level was significantly higher in the AD group without the adjustment of age, but no significant difference was observed between the two groups after age was adjusted (Park et al., 2014). Another hospital-based study with smaller sample size measured blood Cd concentrations in 9 patients with dementia and 19 patients with non-dementia, no significant difference in blood Cd levels between the two groups was observed (Basun et al., 1994).

The limited number of published studies on the relationship between Cd exposure and cognitive function in the elderly population suggest that there are large differences in study design, sample size, and Cd exposure levels among the studies. We note that median blood Cd level in our sample is higher than those of other large-scale cross-sectional studies (Adams & Newcomb 2014; Nisse et al., 2017; Saravanabhavan et al., 2017; Takeda et al., 2017; Zhang et al., 2015). Potential causes for differences in blood Cd levels among study populations may be the differences in age, gender, diets, smoking status and residential environments. In our study population, we hypothesize that the main source of Cd exposure may be from diet and tobacco smoke, since the smoking rate of our sample is much higher than those in other samples. Our study is the first longitudinal study on the association between Cd and cognitive decline. Hence results from our study need to be confirmed in future studies of other populations.

The mechanism underlying our finding that higher Cd levels are associated with greater cognitive decline may be supported by studies of Cd toxicity. It was observed that cerebral cortex and hippocampal neurons are the main targets of Cd toxicity (Shukla et al., 1996). Cd could penetrate neurons via voltage-gated calcium channels and activate the apoptotic pathway by inducing stress responses in cerebral cortical neurons (Chen et al., 2008; López et al., 2003; Xu et al., 2011). Cd exposure could also impair hippocampal neurogenesis both in vitro and in vivo by significantly inducing apoptosis, inhibiting proliferation, and impairing neuronal differentiation in primary cultured adult neural progenitor/stem cells derived from the subgranular zone. In human brains, the hippocampus is one of the most important networks of learning and memory acquisition playing important roles in the process of situational memory, autobiographical memory, learning, imagination, spatial information processing and perception (Maguire et al., 2016; Sheldon & Levine 2016; Zeidman & Maguire 2016). Damage to hippocampus has been observed to lead to cognitive impairment (Arushanyan and Beier, 2008). Thus, it is reasonable to hypothesize that Cd’s detrimental effect on cognitive decline might be caused by its toxic impact on the hippocampus.

Our study has several advantages. First, the longitudinal study design allowed us to examine Cd’s association with cognitive decline. Second, the sample size was relatively large and provided adequate statistical power. Third, multiple cognitive tests were used allowing comprehensive assessment of many cognitive domains. In addition, our analyses included important covariates including age, gender, education level, alcohol and tobacco consumption, social status, APOE genotype, metabolic syndromes to control for potential confounding effects.

There are also limitations in our study. Follow-up time of this study is only 3 years. Hence the association between Cd and cognitive decline needs to be examined in studies with longer observation. Another potential limitation is that our study was conducted in a population with a relatively low level of education and with BMI in the normal range in a rural Chinese elderly population.

5. Conclusion

Our findings suggest that higher Cd exposure is associated with greater cognitive decline in older Chinese adults. Further research is required to confirm our findings in other populations and to explore potential mechanisms.

Highlights.

  • First longitudinal study on the association between Cd and cognitive decline of older adults.

  • Higher Cd levels were associated with lower scores of several cognitive tests.

  • Higher Cd levels were associated with greater 3-year cognitive decline.

Acknowledgements

This research was supported by grants from the National Key Research and Development Program of China (2018YFC801102) and the National Institutes of Health of USA (R01AG019181). The authors thank all the participants and the involving staff from the local cooperative hospitals and centers for disease control and prevention for their efforts.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Declaration of competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  1. Adams SV & Newcomb PA, 2014. Cadmium blood and urine concentrations as measures of exposure: NHANES 1999–2010. J. Exposure Sci. Environ. Epidemiol 24, 163–170. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Akbaraly TN, Hininger-Favier I, Carrière I, Arnaud J, Gourlet V, Roussel AM, Berr C., 2007. Plasma selenium over time and cognitive decline in the elderly. Epidemiology. 18, 52–8. [DOI] [PubMed] [Google Scholar]
  3. Basun H, Lind B, Nordberg M, Nordström M, Björkstén KS & Winblad B, 1994. Cadmium in blood in Alzheimer’s disease and non-demented subjects: results from a population-based study. Biometals. 7, 130–134. [DOI] [PubMed] [Google Scholar]
  4. Branca JJV, Morucci G & Pacini A, 2018. Cadmium-induced neurotoxicity: still much ado. Neural Regener. Res 13, 1879–1882. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Cao Y, Chen A, Radcliffe J, Dietrich KN, Jones RL, Caldwell K & Rogan WJ, 2009. Postnatal cadmium exposure, neurodevelopment, and blood pressure in children at 2, 5, and 7 years of age. Environ. Health Perspect 117, 1580–1586. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cheng Y, Jin Y, Unverzagt FW, Su L, Yang L, Ma F, Hake AM, Kettler C, Chen C, Liu J, Bian J, Li P, Murrell JR, Hendrie HC & Gao S, 2014. The relationship between cholesterol and cognitive function is homocysteine-dependent. Clin. Interventions Aging 9, 1823–1829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Gao S, Jin Y, Hall KS, Liang C, Unverzagt FW, Ji R, Murrell JR, Cao J, Shen J, Ma F, Matesan J, Ying B, Cheng Y, Bian J, Li P & Hendrie HC, 2007. Selenium level and cognitive function in rural elderly Chinese. Am. J. Epidemiol 165, 955–965. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Gao S, Jin Y, Unverzagt FW, Cheng Y, Su L, Wang C, Ma F, Hake AM, Kettler C, Chen C, Liu J, Bian J, Li P, Murrell JR, Clark DO & Hendrie HC, 2014. Cognitive function, body mass index and mortality in a rural elderly Chinese cohort. Arch Public Health. 72, 9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Gao S, Jin Y, Unverzagt FW, Ma F, Hall KS, Murrell JR, Cheng Y, Shen J, Ying B, Ji R, Matesan J, Liang C & Hendrie HC, 2008. Trace element levels and cognitive function in rural elderly Chinese. J Gerontol A Biol Sci Med Sci. 63, 635–641. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Gustin K, Tofail F, Vahter M & Kippler M, 2018. Cadmium exposure and cognitive abilities and behavior at 10 years of age: A prospective cohort study. Environ Int. 113, 259–268. [DOI] [PubMed] [Google Scholar]
  11. Huat TJ, Camats-Perna J, Newcombe EA, Valmas N, Kitazawa M & Medeiros R, 2019. Metal Toxicity Links to Alzheimer’s Disease and Neuroinflammation. J Mol Biol. 431, 1843–1868. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Joint committee for developing Chinese guidelines on prevention and treatment of dyslipidemia in adults, 2007. Chinese guidelines on prevention and treatment of dyslipidemia in adults. Zhonghua xin xue guan bing za zhi 35, 390–419. [PubMed] [Google Scholar]
  13. Kim SW, Heo JH, Kim CH, Yoo DC, Won DH, Lee SG, Cho KJ, Song JH, Park SJ, Yang YG & Choi DW, 2010. Rapid and direct detection of apolipoprotein E genotypes using whole blood from humans. J Toxicol Environ Health A. 73, 1502–1510. [DOI] [PubMed] [Google Scholar]
  14. Li H, Wang Z, Fu Z, Yan M, Wu N, Wu H & Yin P, 2018. Associations between blood cadmium levels and cognitive function in a cross-sectional study of US adults aged 60 years or older. BMJ open 8, e020533. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. López E, Figueroa S, Oset-Gasque MJ & González MP, 2003. Apoptosis and necrosis: two distinct events induced by cadmium in cortical neurons in culture. Br J Pharmacol. 138, 901–911. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Maguire EA, Intraub H & Mullally SL, 2016. Scenes, Spaces, and Memory Traces: What Does the Hippocampus Do? Neuroscientist. 22, 432–439. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Mezynska M & Brzóska MM, 2018. Environmental exposure to cadmium-a risk for health of the general population in industrialized countries and preventive strategies. Environ Sci Pollut Res Int. 25, 3211–3232. [DOI] [PubMed] [Google Scholar]
  18. Min J-Y & Min K-B, 2016. Blood cadmium levels and Alzheimer’s disease mortality risk in older US adults. Environ Health. 15, 69. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Moyano P, García JM, Lobo M, Anadón MJ, Sola E, Pelayo A, García J, Frejo MT & Pino JD, 2018. Cadmium alters heat shock protein pathways in SN56 cholinergic neurons, leading to Aβ and phosphorylated Tau protein generation and cell death. Food Chem Toxicol. 121, 297–308. [DOI] [PubMed] [Google Scholar]
  20. Nisse C, Tagne-Fotso R, Howsam M, Richeval C, Labat L & Leroyer A, 2017. Blood and urinary levels of metals and metalloids in the general adult population of Northern France: The IMEPOGE study, 2008–2010. Int J Hyg Environ Health. 220, 341–363. [DOI] [PubMed] [Google Scholar]
  21. Niti M, Yap KB, Kua EH, Tan CH, Ng TP., 2008. Physical, social and productive leisure activities, cognitive decline and interaction with APOE-epsilon 4 genotype in Chinese older adults. Int Psychogeriatr. 20, 237–51. [DOI] [PubMed] [Google Scholar]
  22. Nordberg M, Winblad B & Basun H, 2000. Cadmium concentration in blood in an elderly urban population. Biometals. 13, 311–317. [DOI] [PubMed] [Google Scholar]
  23. Park J-H, Lee D-W, Park KS & Joung H, 2014. Serum trace metal levels in Alzheimer’s disease and normal control groups. Am J Alzheimers Dis Other Demen. 29, 76–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Peng Q, Bakulski KM, Nan B & Park SK, 2017. Cadmium and Alzheimer’s disease mortality in U.S. adults: Updated evidence with a urinary biomarker and extended follow-up time. Environ Res. 157, 44–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Rodríguez-Barranco M, Lacasaña M, Gil F, Lorca A, Alguacil J, Rohlman DS, González-Alzaga B, Molina-Villalba I, Mendoza R & Aguilar-Garduño C, 2014. Cadmium exposure and neuropsychological development in school children in southwestern Spain. Environ Res.134, 66–73. [DOI] [PubMed] [Google Scholar]
  26. Saravanabhavan G, Werry K, Walker M, Haines D, Malowany M & Khoury C, 2017. Human biomonitoring reference values for metals and trace elements in blood and urine derived from the Canadian Health Measures Survey 2007–2013. Int J Hyg Environ Health. 220, 189–200. [DOI] [PubMed] [Google Scholar]
  27. Satarug S & Moore MR, 2004. Adverse health effects of chronic exposure to low-level cadmium in foodstuffs and cigarette smoke. Environ Health Perspect. 112, 1099–1103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Satarug S, Vesey DA & Gobe GC, 2017. Current health risk assessment practice for dietary cadmium: Data from different countries. Food Chem Toxicol. 106, 430–445. [DOI] [PubMed] [Google Scholar]
  29. Sheldon S & Levine B, 2016. The role of the hippocampus in memory and mental construction. Ann N Y Acad Sci. 1369, 76–92. [DOI] [PubMed] [Google Scholar]
  30. Shukla A, Shukla GS & Srimal RC, 1996. Cadmium-induced alterations in blood-brain barrier permeability and its possible correlation with decreased microvessel antioxidant potential in rat. Hum Exp Toxicol. 15, 400–405. [DOI] [PubMed] [Google Scholar]
  31. Souza-Talarico JN, Marcourakis T, Barbosa F, Moraes Barros SB, Rivelli DP, Pompéia S, Caramelli P, Plusquellec P, Lupien SJ, Catucci RF, Alves AR & Suchecki D, 2017. Association between heavy metal exposure and poor working memory and possible mediation effect of antioxidant defenses during aging. Sci Total Environ. 575, 750–757. [DOI] [PubMed] [Google Scholar]
  32. Takeda SHK, Kuno R, Barbosa F & Gouveia N, 2017. Trace element levels in blood and associated factors in adults living in the metropolitan area of São Paulo, Brazil. J Trace Elem Med Biol. 44, 307–314. [DOI] [PubMed] [Google Scholar]
  33. Wang B, Du Y.,2013. Cadmium and its neurotoxic effects. Oxidative medicine and cellular longevity. 2013, 898034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Wang Y, Zhang P, Chen X, Wu W, Feng Y, Yang H, Li M, Xie B, Guo P, Warren JL, Shi X, Wang S & Zhang Y, 2019. Multiple metal concentrations and gestational diabetes mellitus in Taiyuan, China. Chemosphere. 237, 124412. [DOI] [PubMed] [Google Scholar]
  35. Zeidman P & Maguire EA, 2016. Anterior hippocampus: the anatomy of perception, imagination and episodic memory. Nat Rev Neurosci. 17, 173–182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Zhang L-L, Lu L, Pan Y-J, Ding C-G, Xu D-Y, Huang C-F, Pan X-F & Zheng W, 2015. Baseline blood levels of manganese, lead, cadmium, copper, and zinc in residents of Beijing suburb. Environ Res. 140, 10–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Zhao F-J, Ma Y, Zhu Y-G, Tang Z & McGrath SP, 2015. Soil contamination in China: current status and mitigation strategies. Environ Sci Technol. 49, 750–759. [DOI] [PubMed] [Google Scholar]
  38. Zhou T, Guo J, Zhang J, Xiao H, Qi X, Wu C, Chang X, Zhang Y, Liu Q & Zhou Z, 2020. Sex-Specific Differences in Cognitive Abilities Associated with Childhood Cadmium and Manganese Exposures in School-Age Children: a Prospective Cohort Study. Biol Trace Elem Res. 193, 89–99. [DOI] [PubMed] [Google Scholar]

RESOURCES