Introduction

Measures of handgrip strength (HGS) have the convenient ability to assess overall muscle strength, and low HGS is associated with a variety of clinically relevant health outcomes [1]. Decreased HGS during aging has typically been ascribed to age-related reductions in muscle mass and strength from diminished muscular system functioning; however, some have highlighted that declines in HGS are primarily attributed to reduced nervous system functioning [2, 3]. For example, aging adults produce a muscle force during a maximal HGS measurement that is approximately half of what would be expected if the skeletal musculature were fully activated by the nervous system, due to reduced neural drive to the muscles [3, 4]. Such nervous system dysfunction during aging may help to explain why decreased HGS is associated with cognitive impairment [5, 6]. Given that weakness and cognitive morbidity each accelerate the disabling process [7, 8], the presence of both these health conditions may further limit autonomous living and basic self-care during aging.

Tasks included in assessments of instrumental activities of daily living (IADL) are considered necessary for independent living and dependent on neuropsychological functioning [9]; tasks included in assessments of activities of daily living (ADL) are considered necessary for basic self-care and dependent on physical functioning [10]. When someone reports that they have difficulty or an inability in performing any IADL or ADL, they are usually considered as having an IADL or ADL disability, respectively [11]. Many adults have an IADL or ADL disability, and those living with a functional disability experience lower quality of life and poorer health [12]. Although the presence of a functional disability is linked to several health consequences [10], the conventional binary definitions for IADL and ADL disability have been challenged. For example, IADL disability is fluid, and limitations in individual IADLs are differentially associated with an increased risk for future self-care disability [13]. Similarly, ADL disability is mutable, and limitations in each ADL are uniquely associated with increased mortality risk [14]. Given that it is possible for aging adults to experience worsening, persistence, or recovery from a functional disability, it is also possible for individuals to experience transitions between “have no difficulty”, “have difficulty”, and “are unable to do” for IADLs and ADLs.

Decreased HGS and cognitive morbidity during aging are risk factors for functional disability [10, 15]. However, it remains unclear how weakness and cognitive impairment, independently and jointly, impact declines in autonomous living or basic self-care tasks. Providing further clarity to such a paradigm will allow for a more detailed understanding of how certain health factors potentiate functional declines, thereby informing healthcare providers working with aging adult patients. Moreover, targeted interventions of community-dwelling aging adults aiming to preserve functional capacity through muscle-strengthening activities and cognitive skill tasks may benefit from the development and usage of comprehensive and sensitive measures that better capture the scope of functional outcomes, such as HGS and cognitive assessments [16]. The purpose of this study was to determine the independent and joint associations of HGS and cognitive function on IADL and ADL disability decline in aging Americans.

Methods

Participants

Adults aged 50 years and older who participated in at least one wave of the 2006–2014 waves of the Health and Retirement Study (HRS) were included. Publicly available RAND HRS data were joined with other HRS data files [17]. The HRS is designed to monitor the health and financial status of middle-aged and older Americans [18]. Since 1998, the HRS has supplied data for a national sample of aging Americans. A steady-state design allows the HRS to replenish the sample every 6 years with younger cohorts who were not previously represented [18]. Baseline interviews, which were conducted face-to-face, were only completed with community-dwelling persons, and those who moved into an institutionalized setting after baseline were retrained and interviewed [18]. More details for the HRS have been published elsewhere [19].

Beginning in the 2006 wave, a mixed-mode design for follow-up was utilized, whereby half of participants alternated completion of the core interviews (sometimes by telephone) and detailed face-to-face interviews, which included physical and biological measures. Half samples in the HRS alternated completion of the detailed interviews at each wave to minimize participant burden. Response rates for interviews at each wave of the HRS were > 80% [18]. Written informed consent was provided by participants and the university’s Behavioral Sciences Committee Instructional Review Board approved study protocols. Participant anonymity was ensured, because data used in this secondary analysis contained no direct identifiers.

Measures

Outcome variables

Performance in six IADLs was reported at each wave: use a map, prepare hot meals, take medications, manage money, use a telephone, and shop for groceries. Respondents indicated if they “have no difficulty”, “have difficulty”, or “are unable to do (can’t do or don’t do)” for each IADL. Participants also reported performance in six ADLs: dressing, eating, transferring in or out of bed, toileting, bathing, and walking across a small room. Likewise, respondents indicated if they “have no difficulty”, “have difficulty”, or “are unable to do (can’t do or don’t do)” for each ADL.

Exposure variables

Cognitive functioning was assessed at each wave with a series of tests that were modified from the Telephone Interview of Cognitive Status, a validated screening tool from the Mini-Mental State Examination that was designed for population-based studies [20]. For those aged under 65 years, a 27-point composite scale that included delayed word recall from a list of 10 words (0–20 points), serial sevens subtraction test beginning from 100 (0–5 points), and counting backwards at maximum speed for 10 continuous numbers starting with the number 20 (0–2 points). Respondents with scores ≤ 11 were considered as having a cognitive impairment [21].

For persons aged ≥ 65 years, a 35-point composite scale was used. Additional assessments on the 35-point scale were object naming (0–2 points), date naming (0–4 points), and correctly identifying the current president and vice president of the United States (0–2 points). Respondents with scores ≤ 10 were considered as having a cognitive impairment [22]. Assessments of cognitive functioning from proxy respondents were excluded, because IADLs were included in proxy assessments.

Maximal HGS was measured with a Smedley spring-type hand-held dynamometer (Scandidact, Denmark). The dynamometer was adjusted to fit the hand size of each participant, and then, a practice trial was performed in the standing position with the arm at the side and elbow flexed at 90°. Beginning on the non-dominant hand, participants squeezed the dynamometer with maximal effort, and then released the muscle contractions. Two measures were performed on each hand, alternating between hands. The highest recorded HGS measurement on either hand was included in the analyses.

Data for HGS from either the 2006 and 2008 waves, or 2010 and 2012 waves were ad hoc imputed. For example, if a participant had their HGS measured in the 2006 wave, the same HGS value was used for the 2008 wave; whereas, if an individual had HGS measured in the 2008 wave, the same HGS value was used for the 2006 wave. Utilizing an ad hoc imputation method is common for HGS in the HRS, because measures were part of the detailed face-to-face interviews. Sex- and race-specific HGS cut-points were used for determining weakness (< 40 kg for Black men, < 35 kg for White men, < 31 kg for Black women, and < 22 kg for White women) [23]. Those with no measures for HGS at any wave were excluded. Details for the HGS test protocols are explained elsewhere [24].

Covariates

Age, sex, race, height, and body weight were self-reported at each wave. Exclusions occurred for unknown sex and race that was not Black or White, because we used sex- and race-specific weakness cut-points. Body mass index (BMI) was calculated as body weight in kilograms divided by height in meters-squared. Participants reported healthcare provider diagnosed hypertension, diabetes, cancer, lung disease, heart condition, stroke, psychiatric problems, and arthritis. The number of affirmative morbid diagnoses was summed at each wave and included in the analyses. Mental health was examined at each wave with the 8-item Center for the Epidemiologic Studies Depression (CES-D) scale and continuous scores were included in the analyses [25]. Participants indicated at each wave if they had ever smoked at least 100 cigarettes in their lifetime (smoking history), and if they were current cigarette smokers. A single-item measure of self-rated health was collected at each wave. Participants reported if they thought their health was “excellent”, “very good”, “good”, “fair”, or “poor”. A study sample flowchart for those included is presented in Fig. 1.

Fig. 1
figure 1

Cohort selection diagram. HRS health and retirement study

Full size image

Statistical analysis

All analyses were conducted with SAS 9.4 software (SAS Institute; Cary, NC). Participants were stratified into four groups based on their HGS and cognitive function categorization at each wave: (1) not-weak and no cognitive impairment, (2) weak only, (3) cognitive impairment only, and (4) both weak and has cognitive impairment. Responses from IADL and ADL assessments (“have no difficulty”, “have difficulty”, “are unable to perform”) were categorized as ordinal data. These data were organized with each row representing a single wave for a participant. Separate multilevel ordinal logistic regression models examined the independent and joint associations of weakness and cognitive impairment on IADL and ADL disability decline. Each multilevel ordinal logit model was adjusted for age, sex, race, BMI, morbidity, CES-D score, current smoking status, smoking history, self-rated health, and time since first HGS measurement until last known interview or death. While the estimates acknowledged how data for participants changed at each wave, participants entered and exited the study at different waves (e.g., non-respondent at a wave).

Additional analyses were performed for the independent and joint associations of weakness and cognitive impairment on IADL and ADL disability decline by sex and age. The multilevel ordinal logit models were stratified separately by sex and age (middle-aged, 50–64 years; older adult, ≥ 65 years). Sex and age-stratified analyses were performed as additional analyses, because examining the moderating effects of sex and age for our models was not part of our a priori study purposes. An alpha level of 0.05 was used for all analyses.

Results

Table 1 shows the descriptive characteristics of the 18,391 participants included in the analyses. To make comparisons of descriptive characteristics between HGS and cognitive function groups, the means and 95% confidence intervals (CI) are in Appendix 1 in ESM. Sankey Bar Charts are presented in Figs. 2 and 3 to portray how IADL and ADL status fluctuated across waves, respectively. Sankey Bar Charts allow for fluctuations within categorical groups to be viewed over time [26]. Participants in our study may have experienced improvements, continuance, or declines in their IADL and ADL status at each wave.

Table 1 Descriptive characteristics of the participants
Full size table
Fig. 2
figure 2

Sankey bar chart for illustrating changes in IADL disability status at each wave. 0, have no difficulty; 1, have difficulty; 2, are unable to do; IADL, instrumental activities of daily living. The Sankey bar chart displays a stacked bar graph of the overall percent for those categorized as "have no difficulty", "have difficulty", or "are unable to do" for one or more IADLs at each wave (2006, 2008, 2010, 2012, and 2014). This diagram is overlaid to illustrate the flow of participants’ IADL status between waves. The width of the flow between bar categories is proportional to the number of participants with that particular transition between waves. The figure illustrates that although the overall proportions of participants categorized as "have no difficulty", "have difficulty", and "are unable to do" remain relatively stable over time, transitions occur within participants

Full size image
Fig. 3
figure 3

Sankey bar chart for illustrating changes in ADL disability status at each wave. 0, have no difficulty, 1, have difficulty, 2, are unable to do, ADL, activities of daily living. The Sankey bar chart displays a stacked bar graph of the overall percent for those categorized as "have no difficulty", "have difficulty", or "are unable to do" for one or more ADLs at each wave (2006, 2008, 2010, 2012, and 2014). This diagram is overlaid to illustrate the flow of participants’ ADL status between waves. The width of the flow between bar categories is proportional to the number of participants with that particular transition between waves. The figure illustrates that although the overall proportions of participants categorized as "have no difficulty", "have difficulty", and "are unable to do" remain relatively stable over time, transitions occur within participants

Full size image

The results for the independent and joint associations of HGS and cognitive function on IADL disability decline are shown in Table 2. Relative to aging Americans who were not-weak and did not have a cognitive impairment, those who were weak only, had a cognitive impairment only, and were both weak and had a cognitive impairment had 1.70 (CI 1.57, 1.84), 1.97 (CI 1.74, 2.23), and 3.13 (CI 2.73, 3.59) greater odds for IADL disability decline, respectively.

Table 2 Independent and joint associations of handgrip strength and cognitive function on instrumental activities of daily living disability decline
Full size table

Table 3 displays the results of the independent and joint associations of HGS and cognitive function on ADL disability decline. Compared to those who were not-weak and did not have a cognitive impairment, persons who were weak only had 2.26 (CI 2.03, 2.51) greater odds for ADL disability decline, whereas persons who only had cognitive impairment had 1.26 (CI 1.05, 1.51) greater odds for ADL decline. Aging Americans who were both weak and had a cognitive impairment had 4.48 (CI 3.72, 5.39) greater odds for ADL disability decline relative to those who were not-weak and did not have a cognitive impairment. Results for the moderating effects of sex and age for our models are presented in Appendix 2 in ESM.

Table 3 Independent and joint associations of handgrip strength and cognitive function on activities of daily living disability decline
Full size table

Discussion

The principal results of this investigation demonstrated that weakness and cognitive impairment, independently and jointly, were associated with IADL or ADL disability decline among aging Americans. Specifically, weakness only, cognitive impairment only, and the presence of both weakness and a cognitive impairment in aging Americans were associated with 70%, 97%, and 213% increased odds for declining in an IADL disability compared to those who were not-weak and did not have a cognitive impairment, respectively. Relative to those who were not-weak and did not have a cognitive impairment, aging Americans who were weak had 126% increased odds for declining in an ADL disability; whereas persons who had a cognitive impairment had 26% increased odds for declining in an ADL disability. Interestingly, those who were both weak and had a cognitive impairment had 348% increased odds for declining in an ADL disability.

Tasks included in assessments of IADLs are mostly neuropsychological in nature and do not require high levels of muscle force to be generated for completion. Although HGS has generally been viewed as a clinically viable screening tool for evaluating muscle mass, strength, and function [1], emerging evidence suggests that declines in HGS are linked to the neural systems mediating the control of coordinated movement [2]. Our findings demonstrate weakness and cognitive impairment, independently and jointly, were associated with IADL decline. Given that age-related reductions in neural system functioning contribute to decreased HGS [3], weakness should be viewed as a risk factor for future autonomous living deficits. Research should continue examining how HGS, cognition, and functional capacity are connected, including the development of innovative diagnostic tools and methodologies for identifying cognitive morbidity and functional deficits.

Although higher levels of muscle function are generally needed for completing ADLs, certain tasks also require coordination and fine motor skills. For example, dressing oneself involves both motor (e.g., reaching, lifting, buttoning) and process skills (e.g., locate clothes and dress in a sequential order) [27], and limitations in dressing often occur at the onset of cognitive declines, before other ADL limitations [10, 28]. When aging adults reach the latter stages of the disabling process and experience difficulties in their ability to ambulate or feed themselves, early stage deficits in basic self-care such as dressing have likely regressed into an inability to complete the task.

Weakness may particularly be an important risk factor for diminished ADL capacity, because it is indicative of muscle and motor functioning [2]. This may help to explain why the odds for declining in an ADL disability in our study drastically increased for aging Americans who were both weak and had a cognitive impairment. Healthcare providers working with aging adult patients should routinely assess HGS and cognitive function as a risk-stratifying method for functional declines. Targeted interventions, such as those described by others [10, 29, 30], aiming to prevent or delay functional decline through healthy behavior adherence and cognitive stimulation in aging adults should also include measures of HGS and cognitive function for determining functional decline risk and as possible outcomes. For example, promising interventions, including that performed by Liu et al. [27], which supplement moderate-intensity muscle-strengthening activities with task-oriented training for preserving ADLs may want to consider assessments of HGS and cognitive function for screening and determining an intervention’s efficacy.

Some limitations should be noted. Self-report information is common in large epidemiological studies such as the HRS, but bias from self-report may have created underestimations for our results. Data for HGS in the 2014 wave could not be merged with the 2016 wave, because those data were not yet released at the time of the analyses. Participants intermittently entered and exited the HRS during the 2006–2014 waves. This probably explains why a shift in the proportions of participants who reported difficulty and an inability to perform an IADL occurred at the 2010 wave (Sankey Bar Chart). The small proportion of participants indicating that they were unable to perform an ADL at each wave is likely underestimated, because being unable to perform an ADL typically represents the end stages of the disabling process, and these individuals often have proxy respondents and are institutionalized. Although our sample were aging Americans, our findings may be generalizable to other global aging populations.

Conclusion

This investigation revealed that weakness and cognitive impairment were independently and jointly associated with increased odds for declining in an IADL or ADL disability for aging Americans. The odds of declining in a functional disability were particularly high for those with both weakness and a cognitive impairment, especially for ADL decline. Healthcare providers and interventions should include measures of HGS and cognitive functioning in their assessments of aging adults for detecting functional declines. Further investigating how HGS, the muscular system, and neural system are linked may help to uncover new biomarkers that prevent unsuccessful aging. Such research may help to decelerate the disabling process while lengthening autonomous living and basic self-care.