Abstract
Background
Neurodegenerative disorders, as the irreversible condition, have a long, silent preclinical period. Recognition of early physical signs of neurodegenerative disorders had important practical implications for identifying at‐risk population. The aim of this study was to investigate whether handgrip strength (HGS) asymmetry was associated with the incidence of neurodegenerative disorders among Chinese older adults.
Methods
This study used the data of participants aged 60 years and over from three waves (2011–2015) of China Health and Retirement Longitudinal Study. HGS asymmetry was measured with HGS ratio (maximal non‐dominant HGS/maximal dominant HGS), with the value less than 0.9 or more than 1.1 considered as HGS asymmetry. Physician‐diagnosed neurodegenerative disorders were identified by self‐reported or proxy‐reported information. Competing risk analysis was conducted to examine the association between HGS asymmetry and incident neurodegenerative disorders, with mortality treated as the competing event.
Results
A total of 4925 participants were included in the analysis [mean (SD) age: 68.1(6.68); female: 49.7%]. Eight hundred and eighty‐eight (18.0%) participants had low HGS and 2227 (45.2%) had HGS asymmetry. During the 4 years of follow‐up, there were 156 cases of neurodegenerative disorders and 422 cases of mortality. The incidence of neurodegenerative disorders was 8.7 per 1000 person‐years [95% confidence interval (CI): 7.4–10.2], and the incidence of mortality was 23.5 per 1000 person‐years (95% CI: 21.4–25.9). Both the cause‐specific model and the Fine–Gray subdistribution hazard model showed that participants with HGS asymmetry had increased hazard of neurodegenerative disorders [hazard ratio (HR) = 1.66, P = 0.002, 95% CI: 1.202–2.297; subdistribution hazard ratio (SHR) = 1.65, P = 0.002, 95% CI: 1.202–2.285]. Low HGS, but not HGS asymmetry, was related to the higher hazard of mortality (HR = 1.61, P < 0.001, 95% CI: 1.297–1.995; SHR = 1.58, P < 0.001, 95% CI: 1.286–1.951).
Conclusions
Handgrip strength asymmetry was associated with the future risk of neurodegenerative disorders among Chinese older adults. Public healthcare providers could consider examining HGS asymmetry along with the maximal HGS as a way to identify those at elevated risk of neurodegenerative disorders.
Keywords: Functional laterality, Muscle strength, Neurodegenerative disorders
Introduction
Age‐related neurodegenerative disorders, such as Parkinson's diseases (PD), Alzheimer's disease and other dementias, are posing increasing burdens to families and healthcare systems. 1 , 2 According to a systematic analysis for the Global Burden of Disease study 2016, neurodegenerative disorders ranked as the leading cause of disability‐adjusted life‐years (DALYs), contributing to nearly 11.6% of DALYs worldwide. 3 A recent national investigation reported the overall prevalence of dementia among Chinese adults aged 60 years and over was 6.0%, representing nearly 15 million people. 4 Based on an updated meta‐analysis, the prevalence of PD in China was nearly 1.1% among older adults. 5 Individuals with neurodegenerative disorders usually experience a long asymptomatic period prior to diagnosis. However, by the point at which the diagnosis is confirmed, severe and irreversible damage to the nervous system has already occurred. 6 , 7 At present, neurodegenerative disorders are incurable and irreversible. Therefore, early identification of vulnerable groups and interventions targeting modifiable risk factors, preferably at the asymptomatic stage, are urgently needed. This urgency is promoted by the importance not only of preventing or delaying the occurrence of neurogenerative disorders but also of postponing the progression of the disease, and thus avoiding a series of potential complications. 8
Many studies have investigated the modifiable risk factors of neurodegenerative disorders, such as lifestyle‐related behaviours and health conditions. 8 , 9 More recently, growing evidence has indicated an association between low handgrip strength (HGS) and accelerated decline in neurocognitive function among middle‐aged and older adults. 10 , 11 In terms of mechanism, the grip force generated during HGS assessments is partly regulated by the neural system that mediates the control of coordinated movements. 12 Therefore, the maximal HGS is postulated to partly reflect the functioning of the neural system or brain health. 12 Besides that, the difference in HGS between hands, as characterized by an asymmetrical deficit in HGS, has recently been proposed as a potential indicator of decreased cognitive function and functional disability. 13 , 14 It is generally known that the human body exhibits laterality, with the dominant and non‐dominant side. The difference in strength between the dominant and non‐dominant hand or leg exists even in healthy individuals. 15 Humans tend to use their dominant hands more frequently for daily activities, and the dominant side often demonstrates better motor performance than the non‐dominant side. 16 However, a large motor asymmetry in functional performance between hands may indicate the deficit in neural system functioning or imbalance in brain hemisphere activation. 13 , 16 , 17 For example, McGrath et al. found that HGS asymmetry was independently associated with future impairment of instrumental activities of daily living (IADLs) and decreased cognitive function. 13 , 14 These findings support a possible link between HGS asymmetry and declines in neurocognitive function.
Unlike maximal HGS, HGS asymmetry may reflect another aspect of strength capacity, namely strength imbalance. 18 According to McGrath et al., HGS asymmetry is defined based on the HGS ratio, the ratio between the maximal non‐dominant HGS and maximal dominant HGS. 13 Current protocols of HGS testing already recommend collecting information on hand dominance and multiple measures of HGS for both hands. 19 Therefore, HGS asymmetry can be easily evaluated by using the data obtained during standard HGS assessments. Recently, research on HGS asymmetry and its possible association with future health‐related outcomes, such as functional disability and cognitive decline, has been increasing. 13 , 14 , 18 However, little is known about whether HGS asymmetry was associated with the incidence of neurodegenerative disorders. Therefore, the purpose of this study was to examine whether HGS asymmetry was associated with the incidence of physician‐diagnosed neurodegenerative disorders among Chinese adults aged 60 years and over.
Methods
Data source
This study used data from the China Health and Retirement Longitudinal Study (CHARLS), an ongoing longitudinal survey of a nationally representative sample of Chinese adults aged 45 years and over. Details of the CHARLS have been documented elsewhere. 20 Briefly, the baseline survey (Wave 1) of the CHARLS was conducted in 2011, involving 17 708 respondents (response rate: 80.5%) from 28 Chinese provinces. Subsequent follow‐up with these participants has occurred every 2 years, with the first follow‐up survey (Wave 2) in 2013 and the second (Wave 3) in 2015. The original CHARLS researchers obtained ethical approval for the data collection from the Biomedical Ethics Review Committee of Peking University (IRB00001052–11015).
Participants
In this study, we included data from participants aged 60 years and over at Wave 1 (baseline). We excluded participants with a history of physician‐diagnosed psychiatric problems or neurodegenerative disorders at baseline. We also excluded those who had missing data on the main independent variable (HGS asymmetry) or the main outcome variable of interest (neurodegenerative disorders) at baseline. Furthermore, to reduce the uncertainty in identifying the onset time of events, we excluded participants who did not experience the competing event (mortality) but had missing information to determine whether the main outcome (neurodegenerative disorders) occurred at Wave 2.
Measures
Maximum handgrip strength
In the CHARLS, HGS is measured with the mechanical dynamometer (YuejianTM WL‐1000, Nantong, China). According to the manufacturer, the precision of this type of dynamometer was ±3% (http://www.ntyuejian.com/3e/4.htm). In the present study, this dynamometer showed satisfactory test–retest reliability for both hands (interclass correlation: 0.94 for the left hand; 0.95 for the right hand). Those who have experienced the surgery, inflammation, severe pain, or injury on one or both hands in the preceding 6 months are not asked to complete the HGS testing. After reporting their hand dominance, participants are instructed to bend the elbow at 90° angle and squeeze the dynamometer as hard as they can for a couple of seconds. Those unable to stand unassisted can complete the test while seated. The test is conducted twice on each hand, and the maximum reading from all four tests is used to reflect HGS. The HGS is coded as missing if the participants cannot complete the test for either hand (i.e., the maximal HGS on the one hand is zero). According to the Asian Work Group for Sarcopenia 2019 consensus, low HGS is defined as the maximum HGS less than 28 kg in men and less than 18 kg in women. 21
Handgrip strength asymmetry
The percent difference in HGS between hands is calculated as the ratio of the maximal non‐dominant HGS and maximal dominant HGS (non‐dominant HGS/dominant HGS). 22 Following previous literature, the ‘10% rule’, the general 10% difference in HGS between dominant and non‐dominant hand, was applied to define the HGS asymmetry. 13 , 15 That is, participants with the ratio less than 0.9 or more than 1.1 were classified as having HGS asymmetry.
Physician‐diagnosed neurodegenerative disorders
In CHARLS, physician‐diagnosed neurodegenerative disorders are determined through a self‐reported question asking participants whether they have been diagnosed with any neurodegenerative disorders (e.g. dementia, brain atrophy or Parkinson's disease) by a doctor. The onset time was estimated as the midpoint of the time interval between the last wave when the participant was free of neurodegenerative disorders and the next wave in which the new diagnosis was recorded.
Mortality
Due to increasing age and comorbidities among older populations, mortality is an important source of censoring in geriatric cohort studies. 23 Mortality occurring before the occurrence of the main outcome can be considered a competing event that precludes the onset of the main outcome. Therefore, to obtain more accurate estimates, the competing event of mortality was considered in our examination of the association between HGS asymmetry and the incidence of neurodegenerative disorders. Participants enrolled at baseline were followed up in two subsequent waves. At Wave 2, death information, including death status and death time, was collected. However, at Wave 3, only the death status was recorded. To be consistent with our method of estimating the onset time of neurodegenerative disorders, the time of mortality was calculated as the midpoint of the time interval between the previous wave and the one in which the mortality was recorded.
Demographic characteristics
Demographic characteristics from the baseline of the CHARLS were extracted. Age was divided into three subgroups: 60 to 69, 70 to 79, and 80 and above. We categorized participants' highest education level into three groups: primary school and below, secondary school, and high school and above. Marital status was dichotomized into either (1) married or (2) single, divorced, or widowed. Rural or urban residence was determined based on the administrative divisions established by the National Bureau of Statistics China. 24
Lifestyle‐related behaviours
Baseline data on lifestyle‐related behaviours, including smoking and drinking status, physical activity, social participation, and sleep duration, were considered. Smoking and drinking status were grouped into three categories (never, ever but quit and current usage). Participants were considered as having moderate to vigorous physical activity (MVPA) if they reported engaging in such activity for at least 10 min in a usual week. 13 Social participation was determined based on whether participants reported having taken part in any of the following activities in the past month: interacting with friends; playing mah‐jongg, chess, or cards; going to the sport, social, or other kind of club; participating in a community‐related organization; caring for sick or disabled adults; doing voluntary or charity work; attending the educational or training courses; investing in stocks; using the Internet and other social activities. Sleep duration was categorized into three groups: short (<7 h), normal (7–9 h) and long time (≥9 h). 25
Health conditions
Body mass index (BMI) was calculated as weight in kilogrammes divided by the square of height in metres. BMI status was categorized into four groups: obese (BMI ≥ 25), overweight (23 ≤ BMI < 25), normal (18.5 ≤ BMI < 23) and underweight (BMI < 18.5). 26 Chronic conditions, such as cardiovascular diseases (viz. heart disease and stroke), hypertension and metabolic disorders (viz. dyslipidaemia and diabetes), which have been reported as risk factors of neurodegenerative disorders, were considered in our study. 8 Cognitive status was assessed using 10 mental status items from the Telephone Interview of Cognitive Status (TICS‐10), which includes tasks about identifying today's date (month, day and year), the day of the week, and the season of the year, along with the serial 7 subtraction task that involves subtracting from 100 up to five times. The total score is the sum of correct answers (ranging from 0 to 10). 27 Depression was measured with the validated 10‐item Center for Epidemiologic Studies Depression Scale short form. 28 Participants were asked to rate the frequency with which each mood or symptom occurred in the preceding week. Each item score could range from 0 to 3. The total score was calculated by summing all item scores after reversing two items that were positively formulated (items 5 and 8). A total score greater than or equal to 12 indicated depression. 28 Furthermore, participants were asked to rate their general health status using a single item. Following the previous study, self‐rated health was dichotomized into two categories: (1) good and above or (2) poor/fair. 29
Statistical analysis
Descriptive analysis was used to present the baseline characteristics of the analytical sample. Baseline characteristics were compared between subjects with and without HGS asymmetry. The main outcome of interest in this study was the time to onset of physician‐diagnosed neurodegenerative disorders among older adults. However, in this study, some participants might die before receiving a diagnosis of a neurodegenerative disorders, thereby preventing the main event of interest from happening. These deaths would decrease the actual number of neurodegenerative disorders, which accordingly influenced the probability of the events of neurodegenerative disorders. 30 Hence, in this study, death before the diagnosis of a neurodegenerative disorder was regarded as the competing event. In survival analysis, ignoring competing risks may overestimate the incidence of the event of interest especially when the competing risk is high. 30 Accordingly, we employed competing risk analysis to examine the adjusted association between baseline HGS asymmetry and the risk of neurodegenerative disorders, taking account of the competing risk of mortality. In the context of competing event, the time of event was calculated as the midpoint of the time interval between the wave of no diagnosed neurodegenerative disorders and death until the wave of first event occurred, either the diagnosis of a neurodegenerative disorder or death without a neurodegenerative disorder diagnosis.
The cause‐specific hazard model and Fine–Gray model are commonly used approaches to analyse competing risk endpoints. The cause‐specific hazard model estimates the hazard of an event of interest by excluding individuals from the risk set from the moment of the onset of a competing event; that is, competing events are treated as censored observations. 31 The Fine–Gray model, also known as the subdistribution hazard model, estimates the effect of covariates on the cumulative incidence function for the event of interest while taking competing events into account. 30 Those who experienced a competing event would remain in the risk set, but with no chance of experiencing the event of interest. 30 This apparently unnatural setting is necessary to establish the direct association between covariates and the cumulative incidence function. 30 As suggested by Latouche et al., 31 in order to gain a complete understanding of the relationship between covariates and competing endpoints, both the cause‐specific hazard model and the Fine–Gray subdistribution hazard model should be fitted for both the outcome of interest and the competing event. Hence, in this study, both models were fitted for the incidence of neurodegenerative disorders and mortality separately, adjusted for socio‐demographic characteristics, lifestyle‐related behaviours, health conditions and maximal HGS. The cause‐specific hazard ratio (HR, for the cause‐specific model) and the subdistribution hazard ratio (SHR, for the Fine–Gray model) with 95% confidence intervals (CIs) were presented.
The proportional hazard assumption for these models was checked by testing the significance of the time‐by‐covariate interaction, with significant interactions indicating violations of the proportional hazard assumption for the covariate being interacted with time. 31 If there was evidence of non‐proportionality, the interaction term between time and the covariates violating the proportional assumption would be included in the model. 30
Missing data were present in some baseline covariates. Among the final analytical sample, missing data were present in the following covariates: education level (0.1%), smoking status (0.7%), drinking status (0.1%), MVPA participation (58.6%), social participation (1.4%), sleep duration (2.5%), BMI (1.8%), cardiovascular diseases (0.5%), hypertension (0.3%), metabolic disorders (1.7%), TICS‐10 score (1.9%) and depression (7.5%). The large proportion of missing data with regard to MVPA participation arose because data on physical activity were collected only from a random subgroup of almost two‐fifths of all CHARLS participants. Therefore, it was safe to assume that the data were missing at random (MAR). Previous evidence has suggested that when compared with complete data analysis, valid multiple imputation reduced bias even when the proportion of missingness was large (60% and above) for MAR data. 32 Accordingly, in this study, multiple imputation with 10 times using chained equation model was performed to impute those covariates with missing values. 33 We performed all above models using the imputed dataset as the main results of the study.
Two sensitivity analyses were conducted to test the robustness of the results. First, all above models were repeated using the complete dataset with no missing values in all baseline covariates. Second, due to the large proportion of missing data in MVPA participation, all above models, but with MVPA participation excluded, were repeated with the complete dataset. All analyses were performed using ming Significance was set at the 0.05 level, using two‐tailed tests.
Results
Sample characteristics
At baseline, there were 7669 participants aged 60 years and over. Among those, 124 had psychiatric problems, 308 had physician‐diagnosed neurodegenerative disorders or had missing information with regard to neurodegenerative disorders. After excluding those with missing data on HGS (n = 1553) and HGS asymmetry (n = 328, including 266 ambidextrous respondents and 62 persons with no hand dominance information), there remained 5356 participants at baseline. We further excluded those with missing information on neurodegenerative disorders but who did not experience the competing event (mortality) at Wave 2 (n = 431). Therefore, a total of 4925 participants were included in our analysis.
Baseline characteristics for the analytical sample were presented in Table 1. The mean age was 68.1 ± 6.68 years. Of the 4925 participants, 18.0% had low HGS, and 45.2% had HGS asymmetry. Participants with HGS asymmetry at baseline were more female (P = 0.001), generally older (P < 0.001), had a lower TICS‐10 score (P < 0.001), and had the higher proportion of low HGS (P < 0.001), no MVPA participation (P = 0.019), cardiovascular diseases (P = 0.005), metabolic disorders (P = 0.049), depression (P = 0.009), and poor or fair self‐rated health (P < 0.001) (Table 2).
Table 1.
Characteristics of study sample at baseline (n = 4925)
| Characteristics | Freq | (%) |
|---|---|---|
| HGS | ||
| Normal | 4037 | 82.0 |
| Low | 888 | 18.0 |
| HGS asymmetry | ||
| No | 2698 | 54.8 |
| Yes | 2227 | 45.2 |
| Gender | ||
| Female | 2446 | 49.7 |
| Male | 2479 | 50.3 |
| Age | ||
| 60–70 | 3152 | 64.0 |
| 70–80 | 1428 | 29.0 |
| 80 and over | 345 | 7.0 |
| Education | ||
| Primary school and below | 4100 | 83.3 |
| Secondary school | 552 | 11.2 |
| High school and above | 269 | 5.5 |
| Marriage | ||
| Married | 3875 | 78.7 |
| Single, divorced or widowed | 1050 | 21.3 |
| Residence | ||
| Urban | 1688 | 34.3 |
| Rural | 3237 | 65.7 |
| Smoking status | ||
| Never | 2791 | 57.0 |
| Ever but quit | 571 | 11.7 |
| Current smoke | 1530 | 31.3 |
| Drinking status | ||
| Never | 2858 | 58.1 |
| Ever but quit | 551 | 11.2 |
| Current drink | 1513 | 30.7 |
| MVPA participation | ||
| No | 836 | 41.0 |
| Yes | 1203 | 59.0 |
| Social participation | ||
| No | 2531 | 52.1 |
| Yes | 2323 | 47.9 |
| Sleep duration | ||
| Short | 2608 | 54.3 |
| Normal | 1788 | 37.3 |
| Long | 404 | 8.4 |
| BMI | ||
| Underweight | 509 | 10.5 |
| Normal | 2181 | 45.1 |
| Overweight | 908 | 18.8 |
| Obesity | 1242 | 25.6 |
| Cardiovascular diseases | ||
| No | 4072 | 83.1 |
| Yes | 830 | 16.9 |
| Hypertension | ||
| No | 3396 | 69.2 |
| Yes | 1513 | 30.8 |
| Metabolic disorders | ||
| No | 4166 | 86.1 |
| Yes | 675 | 13.9 |
| TICS‐10 (mean ± SD) | — | 6.7 ± 3.26 |
| Depression | ||
| No | 3135 | 68.8 |
| Yes | 1421 | 31.2 |
| Self‐rated heath | ||
| Good and above | 2097 | 42.6 |
| Poor/fair | 2828 | 57.4 |
BMI, body mass index; HGS, handgrip strength; MVPA, moderate to vigorous physical activity; TICS‐10, Telephone Interview of Cognitive Status (10 items).
Table 2.
Characteristics among participants with and without HGS asymmetry at baseline (n = 4925)
| Characteristics | HGS asymmetry (%) (n = 2227) | No HGS asymmetry (%) (n = 2698) | P value |
|---|---|---|---|
| HGS | <0.001 | ||
| Normal | 78.0 | 85.2 | |
| Low | 22.0 | 14.8 | |
| Gender | 0.001 | ||
| Female | 52.2 | 47.5 | |
| Male | 47.8 | 52.5 | |
| Age | <0.001 | ||
| 60–70 | 61.0 | 66.4 | |
| 70–80 | 30.9 | 27.5 | |
| 80 and over | 8.1 | 6.1 | |
| Education | 0.673 | ||
| Primary school and below | 83.6 | 83.0 | |
| Secondary school | 10.8 | 11.6 | |
| High school and above | 5.6 | 5.4 | |
| Marriage | 0.020 | ||
| Married | 77.2 | 79.9 | |
| Single, divorced or widowed | 22.8 | 20.1 | |
| Residence | 0.796 | ||
| Urban | 34.1 | 34.4 | |
| Rural | 65.9 | 65.6 | |
| Smoking status | 0.017 | ||
| Never | 59.3 | 55.2 | |
| Ever but quit | 11.1 | 12.1 | |
| Current smoke | 29.6 | 32.7 | |
| Drinking status | 0.003 | ||
| Never | 60.3 | 56.2 | |
| Ever but quit | 11.4 | 11.0 | |
| Current drink | 28.3 | 32.8 | |
| MVPA participation | 0.019 | ||
| No | 43.8 | 38.7 | |
| Yes | 56.2 | 61.3 | |
| Social participation | 0.819 | ||
| No | 52.0 | 52.3 | |
| Yes | 48.0 | 47.7 | |
| Sleep duration | 0.942 | ||
| Short | 54.5 | 54.2 | |
| Normal | 37.0 | 37.5 | |
| Long | 8.5 | 8.3 | |
| BMI | 0.066 | ||
| Underweight | 11.7 | 9.5 | |
| Normal | 44.8 | 45.3 | |
| Overweight | 18.0 | 19.4 | |
| Obesity | 25.5 | 25.8 | |
| Cardiovascular diseases | 0.005 | ||
| No | 81.4 | 84.4 | |
| Yes | 18.6 | 15.6 | |
| Hypertension | 0.071 | ||
| No | 67.9 | 70.3 | |
| Yes | 32.1 | 29.7 | |
| Metabolic disorders | 0.049 | ||
| No | 85.0 | 87.0 | |
| Yes | 15.0 | 13.0 | |
| TICS‐10 (mean ± SD) | 6.5 ± 0.07 | 6.9 ± 0.06 | <0.001 |
| Depression | 0.009 | ||
| No | 66.8 | 70.4 | |
| Yes | 33.2 | 29.6 | |
| Self‐rated heath | <0.001 | ||
| Good and above | 39.4 | 45.2 | |
| Poor/fair | 60.6 | 54.8 |
BMI, body mass index; HGS, handgrip strength; MVPA, moderate to vigorous physical activity; TICS‐10, Telephone Interview of Cognitive Status (10 items).
At the end of the follow‐up, there were 156 cases of neurodegenerative disorders and 422 cases of mortality. The incidence rate of neurodegenerative disorders was 8.7 (95% CI: 7.4–10.2) per 1000 person‐years. The incidence rate of mortality was 23.5 (95% CI: 21.4–25.9) per 1000 person‐years.
Cause‐specific hazard model
As shown in Table 3, HGS asymmetry was independently associated with a 66% increase in the hazard of neurodegenerative disorders (HR = 1.66, 95% CI: 1.202–2.297). Though participants with low HGS showed increased risk of neurodegenerative disorders (HR = 1.29, 95% CI: 0.886 – 1.886), the statistical significance was not reached.
Table 3.
Competing risk analysis of incident neurodegenerative diseases, with mortality as the competing event (n = 4925)
| Neurodegenerative disorders | Cause‐specific hazard model a | Subdistribution hazard model b | ||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P value | SHR | 95% CI | P value | |
| HGS asymmetry | 1.66 | 1.202–2.297 | 0.002 | 1.65 | 1.202–2.285 | 0.002 |
| Low HGS | 1.29 | 0.886–1.886 | 0.183 | 1.28 | 0.884–1.842 | 0.192 |
| Male | 1.20 | 0.751–1.920 | 0.444 | 1.19 | 0.719–1.973 | 0.498 |
| Age (ref: 60–70) | 0.283 | 0.280 | ||||
| 70–80 | 1.34 | 0.929–1.920 | 1.33 | 0.927–1.907 | ||
| 80 and over | 1.10 | 0.579–2.104 | 1.07 | 0.566–2.032 | ||
| Education (ref: primary and below) | 0.007 | 0.010 | ||||
| Secondary school | 2.09 | 1.300–3.352 | 2.07 | 1.281–3.351 | ||
| High school and above | 1.69 | 0.813–3.528 | 1.69 | 0.791–3.606 | ||
| Single, divorced or widowed | 0.98 | 0.653–1.472 | 0.925 | 0.97 | 0.645–1.467 | 0.894 |
| Rural residence | 1.41 | 0.912–2.184 | 0.122 | 1.42 | 0.913–2.195 | 0.120 |
| Smoking status (ref: never) | 0.285 | 0.281 | ||||
| Ever but quit | 1.46 | 0.875–2.447 | 1.44 | 0.867–2.391 | ||
| Current smoke | 1.05 | 0.667–1.646 | 1.04 | 0.646–1.670 | ||
| Drinking status (ref: never) | 0.115 | 0.153 | ||||
| Ever but quit | 1.41 | 0.862–2.320 | 1.40 | 0.843–2.337 | ||
| Current drink | 1.52 | 1.009–2.276 | 1.53 | 0.982–2.372 | ||
| MVPA participation | 0.88 | 0.577–1.327 | 0.527 | 0.88 | 0.582–1.336 | 0.550 |
| Social participation | 0.76 | 0.545–1.061 | 0.107 | 0.76 | 0.544–1.059 | 0.105 |
| Sleep duration (ref: normal) | 0.185 | 0.173 | ||||
| Short | 0.79 | 0.550–1.123 | 0.79 | 0.555–1.112 | ||
| Long | 0.59 | 0.297–1.179 | 0.59 | 0.294–1.171 | ||
| BMI (ref: underweight) | 0.775 | 0.774 | ||||
| Normal | 0.93 | 0.544–1.578 | 0.93 | 0.557–1.564 | ||
| Overweight | 0.99 | 0.531–1.854 | 1.00 | 0.548–1.840 | ||
| Obesity | 1.16 | 0.639–2.098 | 1.17 | 0.653–2.085 | ||
| Cardiovascular diseases | 1.40 | 0.953–2.044 | 0.087 | 1.39 | 0.952–2.022 | 0.089 |
| Hypertension | 1.42 | 1.007–2.014 | 0.046 | 1.41 | 1.008–1.981 | 0.045 |
| Metabolic disorders | 1.01 | 0.639–1.594 | 0.968 | 1.01 | 0.623–1.641 | 0.965 |
| TICS‐10 | 0.90 | 0.848–0.950 | 0.001 | 0.90 | 0.847–0.955 | <0.001 |
| Depression | 1.09 | 0.758–1.564 | 0.645 | 1.08 | 0.760–1.548 | 0.654 |
| Poor/fair self‐rated heath | 1.47 | 1.017–2.132 | 0.040 | 1.47 | 1.025–2.104 | 0.036 |
| Time‐by‐covariate interaction | ||||||
| Rural residence | 0.44 | 0.222–0.871 | 0.019 | 0.44 | 0.222–0.857 | 0.016 |
BMI, body mass index; CI, confidence interval; HGS, handgrip strength; HR, hazard ratio; MVPA, moderate to vigorous physical activity; SHR, subdistribution hazard ratio; TICS‐10, Telephone Interview of Cognitive Status (10 items).
Rural residence did not meet proportional hazard assumption (P = 0.019 for the interaction). Therefore, time by residence interaction term was added in the cause‐specific model.
Rural residence did not meet proportional hazard assumption (P = 0.016 for the interaction). Therefore, time by residence interaction term was added in the subdistribution hazard model.
For the competing event of mortality, among participants who were alive and free of neurodegenerative disorders, HGS asymmetry was not significantly associated with the hazard of mortality (HR = 1.04, 95% CI: 0.861–1.268), whereas low HGS was associated with a 61% increase in the hazard of mortality (HR = 1.61, 95% CI: 1.297–1.995) (Table 4).
Table 4.
Competing risk analysis of incident mortality, with neurodegenerative disorders as the competing event (n = 4925)
| Mortality | Cause‐specific hazard model | Subdistribution hazard model | ||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P value | SHR | 95% CI | P value | |
| HGS asymmetry | 1.04 | 0.861–1.268 | 0.658 | 1.04 | 0.860–1.246 | 0.716 |
| Low HGS | 1.61 | 1.297–1.995 | <0.001 | 1.58 | 1.286–1.951 | <0.001 |
| Male | 1.41 | 1.062–1.880 | 0.018 | 1.41 | 1.067–1.856 | 0.016 |
| Age (ref: 60–70) | <0.001 | <0.001 | ||||
| 70–80 | 1.93 | 1.519–2.445 | 1.92 | 1.525–2.424 | ||
| 80 and over | 3.62 | 2.655–4.938 | 3.66 | 2.729–4.911 | ||
| Education (ref: primary and below) | 1.101 | 0.079 | ||||
| Secondary school | 0.87 | 0.592–1.269 | 0.85 | 0.584–1.235 | ||
| High school and above | 0.50 | 0.262–0.961 | 0.49 | 0.258–0.934 | ||
| Single, divorced or widowed | 1.38 | 1.106–1.728 | 0.004 | 1.39 | 1.121–1.719 | 0.003 |
| Rural residence | 0.94 | 0.751–1.171 | 0.570 | 0.93 | 0.749–1.159 | 0.525 |
| Smoking status (ref: never) | <0.001 | <0.001 | ||||
| Ever but quit | 2.12 | 1.557–2.876 | 2.09 | 1.560–2.795 | ||
| Current smoke | 1.48 | 1.126–1.957 | 1.49 | 1.146–1.936 | ||
| Drinking status (ref: never) | 0.290 | 0.208 | ||||
| Ever but quit | 1.15 | 0.859–1.541 | 1.15 | 0.875–1.517 | ||
| Current drink | 0.90 | 0.703–1.164 | 0.89 | 0.705–1.133 | ||
| MVPA participation | 0.69 | 0.498–0.958 | 0.028 | 0.70 | 0.506–0.957 | 0.027 |
| No social participation | 0.84 | 0.686–1.023 | 0.082 | 0.84 | 0.692–1.018 | 0.075 |
| Sleep duration (ref: normal) | 0.338 | 0.269 | ||||
| Short | 1.15 | 0.922–1.424 | 1.15 | 0.933–1.411 | ||
| Long | 1.20 | 0.869–1.671 | 1.22 | 0.895–1.675 | ||
| BMI (ref: underweight) | 0.026 | 0.013 | ||||
| Normal | 0.71 | 0.543–0.932 | 0.71 | 0.554–0.914 | ||
| Overweight | 0.59 | 0.410–0.851 | 0.59 | 0.416–0.833 | ||
| Obesity | 0.67 | 0.471–0.946 | 0.66 | 0.477–0.924 | ||
| Cardiovascular diseases | 1.18 | 0.924–1.511 | 0.185 | 1.17 | 0.927–1.483 | 0.184 |
| Hypertension | 1.44 | 1.163–1.785 | 0.001 | 1.43 | 1.165–1.751 | 0.001 |
| Metabolic disorders | 0.93 | 0.682–1.260 | 0.628 | 0.93 | 0.694–1.255 | 0.650 |
| TICS‐10 | 0.97 | 0.939–1.006 | 0.110 | 0.98 | 0.944–1.008 | 0.142 |
| Depression | 1.09 | 0.869–1.373 | 0.447 | 1.10 | 0.888–1.366 | 0.379 |
| Poor/fair self‐rated heath | 1.23 | 0.991–1.529 | 0.061 | 1.22 | 1.000–1.499 | 0.050 |
BMI, body mass index; CI, confidence interval; HGS, handgrip strength; HR, hazard ratio; MVPA, moderate to vigorous physical activity; SHR, subdistribution hazard ratio; TICS‐10, Telephone Interview of Cognitive Status (10 items).
Fine–Gray subdistribution hazard model
For the Fine–Gray subdistribution hazard model of neurodegenerative disorders, with mortality as the competing risk, the results were similar to those of the cause‐specific model. HGS asymmetry was associated with a 65% increase in the subdistribution hazard of neurodegenerative disorders (SHR = 1.65, 95% CI: 1.202–2.285), and low HGS was not significantly associated with increased risk of neurodegenerative disorders (SHR = 1.28, 95% CI: 0.884–1.842) (Table 3).
For the subdistribution hazard of mortality, with neurodegenerative disorders as the competing risk, the results showed that HGS asymmetry was not significantly associated with the incidence of mortality (SHR = 1.04, 95% CI: 0.860–1.246). However, low HGS was associated with a 58% increase in the subdistribution hazard of mortality (SHR = 1.58, 95% CI: 1.286–1.951) (Table 4).
Sensitivity analysis
Both two sensitivity analyses, (1) using the complete dataset without missing data in all covariates (n = 1841) and (2) using the complete dataset but excluding MVPA participation from the models (n = 4363), showed consistent results with our main findings that HGS asymmetry was independently associated with increased hazard of neurodegenerative disorders (Tables S1 and S3), and low HGS, but not HGS asymmetry, was related to the higher risk of mortality (Tables S2 and S4). Although the first sensitivity analysis (Table S1) showed insignificant association between HGS asymmetry and the incidence of neurodegenerative disorders (HR = 1.45, 95% CI: 0.863–2.438; SHR = 1.44, 95% CI: 0.863–2.410), the estimates (HR and SHR) were similar with our main results. The insignificance might manifest the insufficient statistical power due to the complete case analysis. Nevertheless, these two sensitivity analyses further supported the robustness of our results.
Discussion
In this study, we examined the association between HGS asymmetry and the incidence of neurodegenerative disorders among Chinese older adults using a 4‐year population‐based longitudinal study. Our results suggested that when compared with participants without HGS asymmetry, older adults with HGS asymmetry had an increased hazard of neurodegenerative disorders, after considering the competing event of mortality and controlling for socio‐demographic characteristics, lifestyle‐related behaviours, health conditions and maximal HGS. This was the first study to investigate the association between HGS asymmetry and the incidence of neurodegenerative disorders.
Our study showed that low HGS was not associated with an increased hazard of future neurodegenerative disorders among older population. The significant association between HGS and neurodegenerative disorders was reported by some, 10 , 11 but not all prospective studies. 34 Furthermore, most research focused on the dementia risk, while limited evidence was available for other types of neurodegenerative disorders. Decline in HGS has been considered primarily as an age‐related change in the musculoskeletal system, whereas it has increasingly been viewed as associated with diminished neurological function and brain health. 12 The grip force and muscle coordination involved in the maximum HGS assessment are heavily controlled by the neural system. 12 Therefore, HGS might be a potential indicator of nervous system function and reflect brain health. 12 The absence of significance in our study might show HGS asymmetry as a stronger predictor than low HGS. Yet, future studies are needed to vigorously compare the predictive performance of them.
Our study found that HGS asymmetry was significantly associated with the future risk of neurodegenerative disorders in Chinese older population, after taking mortality into account and controlling for other covariates including HGS. Although no published research has investigated the longitudinal association between HGS asymmetry and the incidence of neurodegenerative disorders, some studies have examined the association of HGS asymmetry with other outcomes that also involved neurocognitive function. 13 , 14 For example, a recent prospective cohort study found that older adults with HGS asymmetry had higher odds of future limitations in IADL, which might reflect impaired neurophysiological function. 14 Furthermore, a panel study showed HGS asymmetry was related to lower cognitive function among older adults. 13 Based on previous evidence, maximum grip force is a complex coordinated behaviour involving the engagement of numerous motor units which requires the activation across brain networks. 12 Disorders in the neural system or brain functioning might affect upper limb functions, which could manifest in decreased coordination when completing various tasks, such as the difficulty in holding an object, needing more time to accomplish a task, decreased strength capacity and asymmetrical performance between hands. 35 Given that strength asymmetry between limbs might indicate the imbalance in the brain hemisphere activation and neurological function, that might explain our findings that HGS asymmetry was associated with the higher risk of neurodegenerative disorders and underpin why previous research found that older adults with HGS asymmetry had higher odds of IADL limitations and lower cognitive function. However, the mechanism of and rationale for the association between HGS asymmetry and neurodegenerative disorders remain unclear and need further research. Moreover, our supplementary analysis showed only the participants with the HGS ratio < 0.9 showed the increased hazard of neurodegenerative disorders (HR = 1.71, 95% CI: 1.217–2.415; SHR = 1.71, 95% CI: 1.223–2.402), while estimates for the HGS ratio > 1.1 were not significant (HR = 1.51, 95% CI: 0.924–2.466; SHR = 1.49, 95% CI: 0.914–2.446) (Table S5). The difference in the risk of neurodegenerative disorders between the HGS ratio < 0.9 and >1.1 needed to be explored further.
Considering that common neurodegenerative disorders develop insidiously, with a long, silent preclinical period, recognition of early physical signs of neurodegenerative disorders had important practical implications for identifying at‐risk population, especially when the measures (e.g. HGS and HGS asymmetry) could be easily assessed in primary care settings. Our findings suggested that HGS asymmetry might be a potential screening tool to detect the risk of neurodegenerative disorders. Older adults with HGS asymmetry might benefit from early cognitive or neuroprotective intervention programmes. Current evidence has shown that exercise intervention, especially the progressive strength training, could not only improve the overall physical functioning but could also slow the degeneration in brain areas that are particularly susceptible to dementia. 36 However, it is unclear whether the strength training programmes that target asymmetric deficits would have additional protective effect on brain function. Because our study is the first preliminary examination of the association between HGS asymmetry and the risk of neurodegenerative disorders, future studies could consider to verify this association in other populations.
In agreement with previous evidence, our study also found low HGS was associated with an increased risk of mortality. 37 , 38 Especially, Leong et al. generalized the association between HGS and future mortality across countries with diverse socio‐economic circumstances. 37 Therefore, the low HGS might serve as a good marker of future risk of mortality. Although the underlying mechanism was uncertain, it was speculated that the loss of HGS might be a sensitive indicator of the ageing process. 39 Additionally, our findings suggested that HGS asymmetry was not associated with the risk of mortality. Although McGrath et al. found older adults with HGS asymmetry showed increased hazard of mortality based on a large sample of older Americans from the Health and Retirement Study (n = 19 325), the small effect size might not be clinically significant (HR = 1.10 95% CI: 1.03–1.17). 38 Future research could continue examining whether HGS asymmetry was associated with all‐cause and/or cause‐specific mortality among older population.
Among the main strengths of this study, we used a prospective design based on a nationwide representative sample of Chinese older population, and a sophisticated statistical model was applied to investigate the association between HGS asymmetry and the incidence of neurodegenerative disorders, taking into account the competing risk of mortality. However, our study was still subject to several limitations. Firstly, to define HGS asymmetry, we excluded participants who were ambidextrous or unable to complete the HGS test on either hand. Therefore, the observed association might not be generalized to the general older population. Secondly, due to data availability, physician‐diagnosed neurodegenerative disorders were self‐reported or proxy‐reported, which did not provide the detailed information about diagnostic criteria and might underestimate the incidence. Nevertheless, this self‐reported measure of neurodegenerative disorders was more feasible in large population‐based surveys and was shown generally valid in previous research. 30 , 41 Although the case ascertainment based on medical records was a more objective method, reliance on this approach also resulted in the underdiagnosis due to the socio‐economic status, disabling stigma, or reimbursement scheme and so forth. 41 Previous research suggested that the usage of any single information source (e.g. self‐reports, medical claims, cognitive tests or prescribed medication) in the case determination might produce underestimates. 42 , 43 Future studies could consider using multiple sources of information to define the neurodegenerative disorders. Besides, due to the self‐reported nature of the outcome, it was reasonable to expect that people with HGS asymmetry would be more likely to see a doctor as their condition progressed and became serious enough to influence their daily life. Therefore, differential misclassification of the outcome might exist. Thirdly, it was unknown about whether the neurodegenerative disorders existed at the time of death. Therefore, some cases of neurodegenerative disorders might be under‐reported. Nevertheless, our results still found the significant association between the HGS asymmetry and neurodegenerative disorders. Moreover, it was possible that the impact of HGS asymmetry on the incidence might vary across different types of neurodegenerative disorders. However, due to data availability, we could not conduct the subgroup analyses to examine the association between HGS asymmetry and the risk of different types of neurodegenerative disorders separately (e.g. Alzheimer's disease and Parkinson's disease). Future study could further examine the risk of different neurodegenerative disorders among older adults with HGS asymmetry. Furthermore, we only observed the 4‐year risk of neurodegenerative disorders. However, the follow‐up time might not be long enough for the development of neurodegenerative disorders. Future studies could consider to verify the impact of HGS asymmetry on the longer term risk of neurodegenerative disorders. In addition, we had missing data in some baseline covariates. In particular, nearly 60% of the baseline MVPA participation data were missing due to the survey design. However, we conducted sensitivity analyses to support the robustness of the main results using the imputed dataset. Furthermore, some covariates (e.g. HGS, depression and BMI) might change during the follow‐up. However, we only considered all covariates at the baseline level. Future studies could examine the impact of changes in these variables on the risk of neurodegenerative disorders. Besides, although we adjusted for numerous potential covariates, residual confounding from unmeasured covariates still existed, for example, family history of neurodegenerative disorders and history of traumatic brain injury. Future studies should consider including these potential covariates. Finally, although this study was prospective, only association rather than causal relationship could be claimed due to its observational nature.
Conclusions
This study indicated that HGS asymmetry was associated with the increased risk of neurodegenerative disorders among Chinese older adults. Public healthcare providers could consider examining HGS asymmetry along with the maximal HGS to help in early detection of vulnerable groups at risk of future neurodegenerative disorders. Because neurodegenerative disorders develop insidiously at the asymptomatic stage, regular assessment of neurocognitive function and early prevention strategies for at‐risk populations are encouraged.
Conflicts of interest
The authors declare that they have no competing interests.
Supporting information
Table S1 Sensitivity analysis of the incident neurodegenerative disorders using the complete dataset (n = 1,841), with mortality as the competing event
Table S2 Sensitivity analysis of the incident mortality using the complete dataset (n = 1,841), with neurodegenerative disorders as the competing event
Table S3 Sensitivity analysis of the incident neurodegenerative disorders using the complete dataset (MVPA participation was excluded, n = 4,363), with mortality as the competing event
Table S4 Sensitivity analysis of the incident mortality using the complete dataset (MVPA participation was excluded, n = 4,363), with neurodegenerative disorders as the competing event
Table S5 Supplementary analysis of the incident neurodegenerative disorders for the HGS ratio <0.9 or >1.1, with mortality as the competing event (n = 4,925)
Acknowledgements
We thank the Peking University National Center for Economic Research for providing the data of the China Health and Retirement Longitudinal Study. All procedures performed in this study have been approved by the appropriate ethics committee and have therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. All authors certify in their manuscript that they comply with the Ethical guidelines for authorship and publishing in the Journal of Cachexia, Sarcopenia and Muscle. 44
Chen Z., Ho M., and Chau P. H. (2022) Handgrip strength asymmetry is associated with the risk of neurodegenerative disorders among Chinese older adults, Journal of Cachexia, Sarcopenia and Muscle, 13, 1013–1023, 10.1002/jcsm.12933
References
- 1. Rajiah K, Maharajan MK, Yeen SJ, Lew S. Quality of life and caregivers' burden of Parkinson's disease. Neuroepidemiology 2017;48:131–137. [DOI] [PubMed] [Google Scholar]
- 2. Henderson C, Knapp M, Nelis SM, Quinn C, Martyr A, Wu YT, et al. Use and costs of services and unpaid care for people with mild‐to‐moderate dementia: baseline results from the IDEAL cohort study. Alzheimers Dement (N Y) 2019;5:685–696. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Feigin VL, Nichols E, Alam T, Bannick MS, Beghi E, Blake N, et al. Global, regional, and national burden of neurological disorders, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol 2019;18:459–480. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Jia L, Du Y, Chu L, Zhang Z, Li F, Lyu D, et al. Prevalence, risk factors, and management of dementia and mild cognitive impairment in adults aged 60 years or older in China: a cross‐sectional study. Lancet Public Health 2020;5:e661–e671. [DOI] [PubMed] [Google Scholar]
- 5. Cui L, Hou NN, Wu HM, Zuo X, Lian YZ, Zhang CN, et al. Prevalence of Alzheimer's disease and Parkinson's disease in China: an updated systematical analysis. Front Aging Neurosci 2020;12:603854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Villemagne VL, Burnham S, Bourgeat P, Brown B, Ellis KA, Salvado O, et al. Amyloid β deposition, neurodegeneration, and cognitive decline in sporadic Alzheimer's disease: a prospective cohort study. Lancet Neurol 2013;12:357–367. [DOI] [PubMed] [Google Scholar]
- 7. Bloem BR, Okun MS, Klein C. Parkinson's disease. Lancet 2021;397:2284–2303. [DOI] [PubMed] [Google Scholar]
- 8. Livingston G, Huntley J, Sommerlad A, Ames D, Ballard C, Banerjee S, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet 2020;396:413–446. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Sabia S, Fayosse A, Dumurgier J, van Hees VT, Paquet C, Sommerlad A, et al. Association of sleep duration in middle and old age with incidence of dementia. Nat Commun 2021;12:2289. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Zammit AR, Piccinin AM, Duggan EC, Koval A, Clouston S, Robitaille A, et al. A coordinated multi‐study analysis of the longitudinal association between handgrip strength and cognitive function in older adults. J Gerontol B Psychol Sci Soc Sci 2021;76:229–241. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Boyle PA, Buchman AS, Wilson RS, Leurgans SE, Bennett DA. Association of muscle strength with the risk of Alzheimer disease and the rate of cognitive decline in community‐dwelling older persons. Arch Neurol 2009;66:1339–1344. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Carson RG. Get a grip: individual variations in grip strength are a marker of brain health. Neurobiol Aging 2018;71:189–222. [DOI] [PubMed] [Google Scholar]
- 13. McGrath R, Cawthon PM, Cesari M, Al Snih S, Clark BC. Handgrip strength asymmetry and weakness are associated with lower cognitive function: a panel study. J Am Geriatr Soc 2020;68:2051–2058. [DOI] [PubMed] [Google Scholar]
- 14. Parker K, Rhee Y, Tomkinson GR, Vincent BM, O'Connor ML, McGrath R. Handgrip weakness and asymmetry independently predict the development of new activity limitations: results from analyses of longitudinal data from the US health and retirement study. J Am Med Dir Assoc 2021;22:821–826.e1. [DOI] [PubMed] [Google Scholar]
- 15. Armstrong CA, Oldham JA. A comparison of dominant and non‐dominant hand strengths. J Hand Surg Br 1999;24:421–425. [DOI] [PubMed] [Google Scholar]
- 16. Wallden M. Stimulus and response. J Bodyw Mov Ther 2011;15:525–527. [DOI] [PubMed] [Google Scholar]
- 17. Prichard E, Propper RE, Christman SD. Degree of handedness, but not direction, is a systematic predictor of cognitive performance. Front Psychol 2013;4:9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. McGrath R, Vincent BM, Jurivich DA, Hackney KJ, Tomkinson GR, Dahl LJ, et al. Handgrip strength asymmetry and weakness together are associated with functional disability in aging Americans. J Gerontol A Biol Sci Med Sci 2021;76:291–296. [DOI] [PubMed] [Google Scholar]
- 19. McGrath R. Are we maximizing the utility of handgrip strength assessments for evaluating muscle function? Aging Clin Exp Res 2021;33:1721–1723. [DOI] [PubMed] [Google Scholar]
- 20. Zhao Y, Hu Y, Smith JP, Strauss J, Yang G. Cohort profile: the China Health and Retirement Longitudinal Study (CHARLS). Int J Epidemiol 2014;43:61–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Chen L‐K, Woo J, Assantachai P, Auyeung T‐W, Chou M‐Y, Iijima K, et al. Asian Working Group for Sarcopenia: 2019 consensus update on sarcopenia diagnosis and treatment. J Am Med Dir Assoc 2020;21:300–307.e2. [DOI] [PubMed] [Google Scholar]
- 22. Carabello RJ, Reid KF, Clark DJ, Phillips EM, Fielding RA. Lower extremity strength and power asymmetry assessment in healthy and mobility‐limited populations: reliability and association with physical functioning. Aging Clin Exp Res 2010;22:324–329. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Weuve J, Proust‐Lima C, Power MC, Gross AL, Hofer SM, Thiébaut R, et al. Guidelines for reporting methodological challenges and evaluating potential bias in dementia research. Alzheimers Dement 2015;11:1098–1109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. National Bureau of Statistics China . Urban‐rural division codes for statistics in 2011. 2011. http://www.stats.gov.cn/tjsj/tjbz/tjyqhdmhcxhfdm/2011/index.html. Accessed 4th September 2021.
- 25. Li J, Cacchione PZ, Hodgson N, Riegel B, Keenan BT, Scharf MT, et al. Afternoon napping and cognition in Chinese older adults: findings from the China Health and Retirement Longitudinal Study Baseline Assessment. J Am Geriatr Soc 2017;65:373–380. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. World Health Organization . Regional Office for the Western P. The Asia‐Pacific Perspective: Redefining Obesity and its Treatment. Sydney: Health Communications Australia; 2000. [Google Scholar]
- 27. Lei X, Smith JP, Sun X, Zhao Y. Gender differences in cognition in China and Reasons for Change over Time: Evidence from CHARLS. J Econ Ageing 2014;4:46–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Chen H, Mui AC. Factorial validity of the Center for Epidemiologic Studies Depression Scale short form in older population in China. Int Psychogeriatr 2014;26:49–57. [DOI] [PubMed] [Google Scholar]
- 29. Beydoun HA, Beydoun MA, Weiss J, Hossain S, Huang S, Alemu BT, et al. Insomnia as a predictor of diagnosed memory problems: 2006‐2016 Health and Retirement Study. Sleep Med 2021;80:158–166. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 1999;94:496–509. [Google Scholar]
- 31. Latouche A, Allignol A, Beyersmann J, Labopin M, Fine JP. A competing risks analysis should report results on all cause‐specific hazards and cumulative incidence functions. J Clin Epidemiol 2013;66:648–653. [DOI] [PubMed] [Google Scholar]
- 32. Madley‐Dowd P, Hughes R, Tilling K, Heron J. The proportion of missing data should not be used to guide decisions on multiple imputation. J Clin Epidemiol 2019;110:63–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Royston P, White IR. Multiple imputation by chained equations (MICE): implementation in Stata. J Stat Softw 2011;45:1–20. [Google Scholar]
- 34. Cui C, Mackey RH, Shaaban CE, Kuller LH, Lopez OL, Sekikawa A. Associations of body composition with incident dementia in older adults: Cardiovascular Health Study‐Cognition Study. Alzheimers Dement. 2020;16(10):1402–1411. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Carmeli E, Patish H, Coleman R. The aging hand. J Gerontol A Biol Sci Med Sci 2003;58:M146–M152. [DOI] [PubMed] [Google Scholar]
- 36. Broadhouse KM, Singh MF, Suo C, Gates N, Wen W, Brodaty H, et al. Hippocampal plasticity underpins long‐term cognitive gains from resistance exercise in MCI. NeuroImage Clin 2020;25:102182, 10.1016/j.nicl.2020.102182 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Leong DP, Teo KK, Rangarajan S, Lopez‐Jaramillo P, Avezum A Jr, Orlandini A, et al. Prognostic value of grip strength: findings from the Prospective Urban Rural Epidemiology (PURE) study. Lancet 2015;386:266–273. [DOI] [PubMed] [Google Scholar]
- 38. McGrath R, Tomkinson GR, LaRoche DP, Vincent BM, Bond CW, Hackney KJ. Handgrip strength asymmetry and weakness may accelerate time to mortality in aging Americans. J Am Med Dir Assoc 2020;21:2003–2007.e1. [DOI] [PubMed] [Google Scholar]
- 39. Syddall H, Cooper C, Martin F, Briggs R, Aihie Sayer A. Is grip strength a useful single marker of frailty? Age Ageing 2003;32:650–656. [DOI] [PubMed] [Google Scholar]
- 40. McGrath R, Robinson‐Lane SG, Clark BC, Suhr JA, Giordani BJ, Vincent BM. Self‐reported dementia‐related diagnosis underestimates the prevalence of older Americans living with possible dementia. J Alzheimers Dis 2021;82:373–380. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Chen Y, Tysinger B, Crimmins E, Zissimopoulos JM. Analysis of dementia in the US population using Medicare claims: insights from linked survey and administrative claims data. Alzheimers Dement (N Y) 2019;5:197–207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Lin PJ, Kaufer DI, Maciejewski ML, Ganguly R, Paul JE, Biddle AK. An examination of Alzheimer's disease case definitions using Medicare claims and survey data. Alzheimers Dement 2010;6:334–341. [DOI] [PubMed] [Google Scholar]
- 43. Pressley JC, Trott C, Tang M, Durkin M, Stern Y. Dementia in community‐dwelling elderly patients: a comparison of survey data, medicare claims, cognitive screening, reported symptoms, and activity limitations. J Clin Epidemiol 2003;56:896–905. [DOI] [PubMed] [Google Scholar]
- 44. von Haehling S, Morley JE, Coats AJS, Anker SD. Ethical guidelines for publishing in the Journal of Cachexia, Sarcopenia and Muscle: update 2021. J Cachexia Sarcopenia Muscle 2021;12:2259–2261. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1 Sensitivity analysis of the incident neurodegenerative disorders using the complete dataset (n = 1,841), with mortality as the competing event
Table S2 Sensitivity analysis of the incident mortality using the complete dataset (n = 1,841), with neurodegenerative disorders as the competing event
Table S3 Sensitivity analysis of the incident neurodegenerative disorders using the complete dataset (MVPA participation was excluded, n = 4,363), with mortality as the competing event
Table S4 Sensitivity analysis of the incident mortality using the complete dataset (MVPA participation was excluded, n = 4,363), with neurodegenerative disorders as the competing event
Table S5 Supplementary analysis of the incident neurodegenerative disorders for the HGS ratio <0.9 or >1.1, with mortality as the competing event (n = 4,925)