Highlights
-
•
New weakness cut-points are associated with activities of daily living limitations.
-
•
Fluidity in weakness status may inform odds for limitations in basic self-care.
-
•
Healthcare providers should discuss the importance of strength with older patients.
Keywords: Aging, Geriatric assessment, Longevity, Muscle strength, Muscle strength dynamometer
Abstract
Aims
To examine the associations of (1) absolute and normalized weakness cut-points, (2) collective weakness categories, and (3) changes in weakness status on future activities of daily living (ADL) limitations in older Americans.
Methods
The analytic sample included 11,656 participants aged ≥65 years from the 2006-2018 waves of the RAND Health and Retirement Study. ADL were self-reported. A handgrip dynamometer was used to measure handgrip strength (HGS). Males were classified as weak if their HGS was <35.5 kg (absolute), <0.45 kg/kg (body mass normalized), or <1.05 kg/kg/m2 (body mass index [BMI] normalized); females were considered weak if their HGS was <20.0 kg, <0.337 kg/kg, or <0.79 kg/kg/m2. Participants were similarly categorized as being below 1, 2, or all 3 absolute and normalized cut-points. These collective categories were also used to classify observed changes in weakness status over time (onset, persistent, progressive, recovery).
Results
Older Americans below absolute and normalized weakness cut-points had greater future ADL limitations odds: 1.34 (95% confidence interval [CI]: 1.22-1.47) for absolute, 1.36 (CI: 1.24-1.50) for BMI normalized, and 1.56 (CI: 1.41-1.73) for body mass normalized. Persons below 1, 2, or 3 cut-points had 1.36 (CI: 1.19-1.55), 1.60 (CI: 1.41-1.80), and 1.70 (CI: 1.50-1.92) greater odds for future ADL limitations, respectively. Those in each changing weakness classification had greater future ADL limitation odds: 1.28 (CI: 1.01-1.62) for onset, 1.53 (CI: 1.22-1.92) for persistent, 1.72 (CI: 1.36-2.19) for progressive, and 1.34 (CI: 1.08-1.66) for recovery.
Conclusions
The presence of weakness, regardless of cut-point and change in status over time, was associated with greater odds for future ADL limitations.
Introduction
Handgrip strength (HGS) is a feasible and reliable measure of strength capacity, and a biomarker of health that should be routinely collected as a vital sign during patient healthcare visits.1,2 Age-related changes in strength can be observed over time with HGS measurements.3 Weakness, as determined from HGS, reflects muscle dysfunction and the onset of the disabling process.4 While the cascade of functional loss may first be identified with weakness, the end-stages often include limitations in performing basic self-care tasks.4 Therefore, it is unsurprising that weakness is robustly associated with activities of daily living (ADL) limitations.5, 6, 7
Although weakness is associated with many clinically relevant health conditions including ADL limitations,8 several cut-points for defining weakness exist.9 The utilization of multiple cut-points for determining weakness creates heterogeneity in HGS protocols, thereby threatening internal validity and generating challenges in making comparisons across investigations examining weakness status.9 Body size is also a factor that may influence muscle strength.10 Normalizing absolute HGS measures to a body size factor (e.g., body mass or body mass index [BMI]) may improve the clinical utility of strength measures for predicting the development of future mobility limitations. Moreover, working groups on sarcopenia from Asia and Europe created separate weakness cut-points.11,12 As such, having uniform population-specific absolute and body size–normalized weakness cut-points may help to reduce inconsistencies in HGS protocols while also enabling cut-point selection that accommodates different quantifications of strength.
The Sarcopenia Definition and Outcomes Consortium (SDOC) project, which was a large collaborative effort of content experts who conducted a set of comprehensive data analyses (e.g., receiving operating characteristic curves, classification and regression trees) from different investigations in the United States (e.g., Osteoporotic Fractures in Men [MrOS] Study; Study of Osteoporotic Fractures; Health, Aging and Body Composition Study; Cardiovascular Health Study; Johnston County Arthritis Study), Sweden (e.g., MrOS Sweden), China (e.g., Mr&MsOS Hong Kong), and Australia (e.g., Concord Health and Aging in Men Project), generated new absolute and normalized (body mass and BMI) weakness cut-points for distinguishing slowness in community-dwelling older adults.13,14 While these novel cut-points provide guidance for improving how we define weakness, the proportions of Americans below the respective absolute and normalized cut-points are not equal.15 Moreover, it is plausible that these absolute and body size–normalized cut-points could be used together for better defining weakness, which may lead to fewer weakness misclassifications and stronger predictive value for health conditions such as ADL limitations. However, it remains unknown how these cut-points, individually or collectively, are linked to basic self-care. The purposes of this study were to determine the associations of (1) both absolute and body size–normalized weakness cut-points, (2) collective weakness categories, and (3) changes in weakness status on future ADL limitations in older Americans.
Methods
Participants
We performed a secondary analysis of 11,799 participants aged ≥65 years from the 2006-2018 waves of the RAND Health and Retirement Study (HRS), with data for at least one wave of HGS and succeeding waves of ADLs. The HRS provides data for a national sample of Americans aged over 50 years. To determine eligibility, a brief household screening interview is conducted with each sampled housing unit.16 The HRS also utilizes a longitudinal panel design for monitoring health-related factors in Americans during aging.17 HRS participants are surveyed biennially until death, and new participant cohorts are added occasionally for helping the HRS maintain a national sample. Interview response rates for the HRS have regularly been >80%.16,18 Additional details about the HRS are available elsewhere.19
The HRS began performing detailed face-to-face interviews with participants starting in the 2006 wave, which included physical measures such as HGS.17 These face-to-face interviews occurred on a random half sample of HRS participants and alternated in completions at each wave, while the other random half sample engaged in core interviews to mitigate participant burden. Written informed consent was provided by all HRS participants before study entry and the University's Behavioral Sciences Committee Institutional Review Board approved study protocols.
Measures
Activities of Daily Living
At each wave, respondents told interviewers about their ability to complete ADLs including dressing, eating, transferring in or out of bed, toileting, bathing, and walking across a small room. Persons suggesting difficulty or an inability in completing any basic self-care task were considered as having an ADL limitation.
Handgrip Strength
Participants were eligible for HGS assessments if they did not report having surgery, swelling, severe pain, or an injury to both hands in the 6 months before the interview.20 If such symptoms were present in a single hand, then only the other hand was tested. A Smedley spring-type handgrip dynamometer (Scandidact; Odder, Denmark) was used to measure HGS. Procedures for measuring HGS in the HRS started with fitting the dynamometer to the hand size of the participant and allowing for a practice trial on the reported dominant hand while standing with their arm positioned on the side at a 90° angle. Each participant was encouraged to squeeze the dynamometer with maximal effort for a couple of seconds and then release their grasp. The handgrip dynamometer was placed in the nondominant hand for the start of testing, and HGS values were collected twice, alternating on each hand. If HGS was only performed on a single hand, then a 30-second rest period passed between measures. Procedure alternatives were allowed for participants unable to stand or grasp the dynamometer, whereby persons could be seated and place their arm on a supporting object.20 More details about the HGS procedures in the HRS are available elsewhere.20
The highest recorded HGS regardless of hand was included in the analyses. Males with HGS <35.5 kg (absolute), <0.45 kg/kg (normalized to body mass), or <1.05 kg/kg/m2 (normalized to BMI) were considered weak, while females with HGS <20.0 kg, <0.337 kg/kg, or <0.79 kg/kg/m2 were classified as weak.13, 14, 15 Collective weakness was categorized as participants being below 1, 2, or all 3 cut-points. Likewise, the collective weakness categories were used to classify changes in weakness status across the biennial waves in which HGS was collected in the HRS. Persons considered never weak had HGS that was below none of the weakness cut-points at each wave. Onset weakness was defined as being below no weakness cut-points and then transitioning into being below ≥1 cut-point. Those with progressive weakness were below at least 1 cut-point and then transitioned into being below additional cut-points. Persistent weakness was operationalized as persons being below the same number of cut-points across waves. Those experiencing weakness recovery transitioned across waves from being below at least 1 cut-point into fewer or no cut-points.
Covariates
Age, sex (male, female), race (black, white, other), ethnicity (Hispanic, non-Hispanic), educational achievement (less than high school, graduate education development test passed, high school graduate, some college, college graduate), marital status (married or partnered, not married or partnered), height, and body mass were reported at each wave. Persons with a BMI ≥30 kg/m2 were categorized as obese. Participants also reported if they consume alcohol, were currently smoking cigarettes, or had smoked at least 100 cigarettes in their lifetime (previous smoker). Respondents told interviewers if a healthcare provider diagnosed them with a heart condition, stroke, hypertension, chronic lung disease, emotional or psychiatric problems, arthritis or rheumatism, diabetes, and cancer (excluding minor skin cancer). Those affirming the presence of 2 or more conditions were classified as having multimorbidity. Self-rated health was collected with a single-item indicator whereby participants perceived their health as either “excellent,” “very good,” “good,” “fair,” or “poor.”
A modified version of the Telephone Interview of Cognitive Status (TICS) was used to examine cognitive function. The modified TICS is a well-utilized and validated tool that was developed from the Mini-Mental State Examination for population-based studies such as the HRS.21 A 35-point composite scale was used which included assessments for immediate and delayed word recall from a list of 10 words, serial sevens subtraction test beginning with the number 100, counting backward at maximal speed for 10 consecutive numbers starting from the number 10, object naming, date naming, and reporting the current president and vice president of the United States. Persons with scores <11 were classified as having a cognitive impairment.22 Depressive symptomology was evaluated with the 8-item Center for the Epidemiologic Studies Depression scale.23 Participants indicated if they experienced certain positive or negative emotions from the week before the interview date. Scores ranged from 0 to 8, with higher scores suggesting more depressive symptoms. Persons with scores ≥3 were considered depressed.23 Persons reported participation in moderate-to-vigorous physical activity (MVPA) as “never,” “1-3 occasions per month,” “once per week,” “more than once per week,” or “daily.” Those with missing covariates were excluded (n = 143).
Statistical Analysis
All analyses were performed with SAS 9.4 software (SAS Institute; Cary, NC, USA). Participants entered our analyses when HGS was first measured, and ADL status, along with the other covariates, were evaluated at each wave in which HGS was recorded. The outcome was ADL limitation status at the next available wave. Follow-up time between waves wherein HGS was ascertained and the ADL status outcome were accounted for in the analyses. Appendix 1 outlines when participants first entered our analyses and when ADL status was subsequently assessed. The descriptive characteristics of the participants are presented as mean ± standard deviation and frequency (percentage) for categorical variables. Differential classification (sensitivity, specificity, positive predictive value, negative predictive value) of individual (absolute, BMI normalized, body mass normalized) and collective weakness categories (≥1 cut-point, ≥2 cut-points, 3 cut-points) for future ADL limitations were similarly quantified.
We employed a series of generalized estimating equations (GEEs) for our analyses because we were interested in overall average effects instead of subject-specific effects (e.g., mixed effects model) with large repeated measures data such as HRS.24,25 Therefore, individual GEEs determined the associations of (1) absolute weakness (reference: no absolute weakness), (2) BMI-normalized weakness (reference: no BMI-normalized weakness), and (3) body mass–normalized weakness (reference: no body mass–normalized weakness) on future ADL limitations. Further, another GEE was used to analyze the association between persons below 1, 2, or all 3 weakness cut-points (reference: below 0 cut-points) for future ADL disability. Each GEE was first only adjusted for ADL limitation status at current wave and follow-up years (crude). Thereafter, the fully adjusted GEEs (principal findings) controlled for ADL limitation status at current wave, follow-up years, sex, race, multimorbidity, age, cigarette smoking status, depression, MVPA, obesity, cognitive impairment, alcohol consumption, marital status, and educational achievement. All GEEs accounted for repeated measurements and the outcome for the next wave participated was used. Covariates included in the fully adjusted models were prespecified by investigators because they were considered influential for the association between weakness and functional limitations.
A Sankey bar chart was created to display the fluidity in the collective weakness categories. Given that the HRS collects HGS on random half samples of participants at alternating waves, we merged such waves whereby HGS was measured on the whole sample for the Sankey bar chart (i.e., 2006-2008; 2010-2012; 2014-2016). In brief, Sankey bar charts allow for observation of changes and magnitude of flow within categorical groups over time.26 We next used a GEE to determine whether change in weakness category (i.e., onset, progression, persistence, recovery) was associated with future ADL limitations (reference: never weak). The binary outcome was again ADL limitation status at the next wave. This specific GEE required at least two waves of HGS measurements to capture changes in weakness status, and therefore, the analytic sample for the model included 9,878 participants. For example, in a participant with HGS measured at 2006 and 2010 waves, the weakness status category was determined using the change from wave 2006 and 2010, and the outcome (future ADL limitation) was assessed at 2012, or the next available wave. The model likewise accounted for repeated measures and the same covariate modeling procedures were utilized. An α-level of 0.05 was used for the analyses.
Results
The descriptive characteristics of the 11,656 participants are shown in Table 1. Overall, participants were aged 72.4 ± 6.6 years and there were slightly more females (57.8%). Of the participants included, 5,112 (43.9%) met no weakness cut-point criteria, 2,191 (18.8%) were below 1 cut-point, 2,289 (19.6%) were below 2 cut-points, and 2,064 (17.7%) were below all 3 weakness cut-points. Figure 1 presents the proportions of participants classified as weak for each cut-point individually and together. For the overall sample, 5,112 (43.9%) were not weak, while 499 (22.1%) participants with a future ADL limitation were not weak.
Table 1.
Baseline descriptive characteristics of the participants.
| Overall (n = 11,656) | 0 Weakness categories (n = 5,112) | 1 Weakness category (n = 2,191) | 2 Weakness categories (n = 2,289) | 3 Weakness categories (n = 2,064) | |
|---|---|---|---|---|---|
| Age (years) | 72.4 ± 6.6 | 70.8 ± 5.6 | 72.1 ± 6.2 | 73.0 ± 6.8 | 75.8 ± 7.8 |
| Female (n [%]) | 6,741 (57.8) | 2,898 (56.7) | 999 (45.6) | 1,443 (63.0) | 1,401 (67.8) |
| Race and ethnicity (n [%]) | |||||
| Hispanic Black | 29 (0.2) | 8 (0.2) | 3 (0.1) | 15 (0.7) | 3 (0.2) |
| Hispanic White | 746 (6.4) | 280 (5.5) | 119 (5.4) | 158 (6.9) | 189 (9.1) |
| Non-Hispanic Black | 1,623 (13.9) | 667 (13.0) | 324 (14.8) | 376 (16.4) | 256 (12.4) |
| Non-Hispanic White | 8,759 (75.2) | 3,955 (77.4) | 1,673 (76.4) | 1,636 (71.5) | 1,495 (72.4) |
| Other | 499 (4.3) | 202 (3.9) | 72 (3.3) | 104 (4.5) | 121 (5.9) |
| Multimorbidity (n [%]) | 7,985 (68.5) | 2,975 (58.2) | 1,551 (70.8) | 1,798 (78.5) | 1,661 (80.4) |
| Cigarette smoking status (n [%]) | |||||
| Current smoker | 1,176 (10.1) | 630 (12.3) | 224 (10.2) | 190 (8.3) | 132 (6.4) |
| Previous smoker | 5,386 (46.2) | 2,280 (44.6) | 1,110 (50.7) | 1,053 (46.0) | 943 (45.7) |
| Never smoked | 5,094 (43.7) | 2,202 (43.1) | 857 (39.1) | 1,046 (45.7) | 989 (47.9) |
| Self-rated health (n [%]) | |||||
| Excellent | 1,066 (9.1) | 666 (13.0) | 180 (8.2) | 131 (5.7) | 89 (4.3) |
| Very good | 3,643 (31.3) | 1,931 (37.8) | 661 (30.2) | 587 (25.6) | 464 (22.5) |
| Good | 3,900 (33.5) | 1,649 (32.3) | 774 (35.3) | 828 (36.2) | 649 (31.4) |
| Fair | 2,356 (20.2) | 722 (14.1) | 451 (20.6) | 567 (24.8) | 616 (29.9) |
| Poor | 691 (5.9) | 144 (2.8) | 125 (5.7) | 176 (7.7) | 246 (11.9) |
| Depressed (n [%]) | 2,129 (18.2) | 708 (13.8) | 360 (16.4) | 505 (22.0) | 556 (26.9) |
| Obesity (n [%]) | 3,717 (31.9) | 736 (14.4) | 992 (45.2) | 1,186 (51.8) | 803 (38.9) |
| Consumes alcohol (n [%]) | 5,888 (50.5) | 2,835 (55.4) | 1,164 (53.1) | 1,047 (45.7) | 842 (40.8) |
| Married or partnered (n [%]) | 7,525 (64.5) | 3,527 (69.0) | 1,499 (68.4) | 1,359 (59.3) | 1,140 (55.2) |
| Moderate-to-vigorous physical activity participation (n [%]) | |||||
| Daily | 2,437 (20.9) | 655 (12.8) | 230 (10.5) | 233 (10.2) | 193 (9.3) |
| More than once per week | 529 (4.6) | 2,799 (54.8) | 996 (45.4) | 851 (37.2) | 722 (35.0) |
| Once per week | 3,698 (31.7) | 697 (13.6) | 350 (16.0) | 359 (15.7) | 275 (13.3) |
| 1-3 Occasions per month | 2,542 (21.8) | 403 (7.9) | 226 (10.3) | 236 (10.3) | 187 (9.1) |
| Never | 2,450 (21.0) | 558 (10.9) | 389 (17.8) | 610 (26.6) | 687 (33.3) |
| Cognitive impairment (n [%]) | 232 (2.0) | 48 (0.9) | 43 (1.9) | 54 (2.3) | 87 (4.2) |
| Educational achievement (n [%]) | |||||
| Less than high school | 2,437 (20.9) | 915 (17.9) | 423 (19.3) | 545 (23.8) | 554 (26.8) |
| Graduate education development test passed | 529 (4.5) | 234 (4.6) | 123 (5.6) | 102 (4.5) | 70 (3.4) |
| High school graduate | 3,698 (31.8) | 1,572 (30.8) | 658 (30.0) | 742 (32.4) | 726 (35.2) |
| Some college | 2,542 (21.8) | 1,137 (22.2) | 502 (22.9) | 495 (21.6) | 408 (19.8) |
| College graduate | 2,450 (21.0) | 1,254 (24.5) | 485 (22.2) | 405 (17.7) | 306 (14.8) |
| Follow-up years | 2.1 ± 0.5 | 2.1 ± 0.5 | 2.1 ± 0.4 | 2.1 ± 0.5 | 2.1 ± 0.5 |
| Activities of daily living limitation (n [%]) | 1,847 (15.8) | 392 (7.6) | 310 (14.1) | 493 (21.5) | 652 (31.6) |
| Activities of daily living limitation at next wave (n [%]) | 2,261 (19.4) | 499 (9.7) | 395 (18.0) | 609 (26.6) | 758 (36.7) |
Note: Results are presented as mean ± standard deviation or frequency (percentage) as indicated.
Figure 1.
Venn diagram for the proportions of participants classified as weak for each cut-point. Note: (A) Overall (n = 11,656); (B) Future activities of daily living limitation (n = 2,261). Gray = Absolute weakness (<35.5 kg in males and <20 kg in females), Green = Body mass–normalized weakness (<0.45 kg/kg in males and <0.337 kg/kg in females), Yellow = Body mass index–normalized weakness (<1.05 kg/kg/m2 in males and <0.79 kg/kg/m2 in females).
The differential classification characteristics of the individual weakness categories on future ADL limitations are in Table 2. The sensitivity for the absolute, BMI-normalized, and body mass–normalized weakness cut-points was 69.8% (95% CI: 69.1, 70.4), 71.1% (CI: 70.4, 71.7), and 47.8% (CI: 47.1, 48.6), respectively. The specificity was 51.6% (CI: 50.1, 53.1) for those below the absolute weakness cut-point, 54.4% (CI: 53.0, 55.9) for persons below the BMI-normalized weakness cut-point, and 76.9% (CI: 75.6, 78.1) for participants below the body mass–normalized weakness cut-point. For the collective weakness categories, sensitivity elevated for future ADL limitations as being below more cut-points increased: 43.8% (CI: 43.0, 44.5) for ≥1 cut-point, 62.7% (CI: 62.0, 63.4) for ≥2 cut-points, and 82.2% (CI: 81.6, 82.7) for all 3 cut-points. Alternatively, specificity declined for future ADL limitations as being beneath more cut-points increased: 80.7% (CI: 79.5, 81.8) for ≥1 cut-point, 62.7% (CI: 62.7, 65.5) for ≥2 cut-points, and 38.2% (CI: 36.7, 39.6) for all 3 cut-points.
Table 2.
Classification characteristics of the weakness categories on future activities of daily living limitations.
| Estimate (%) | 95% Confidence interval | |
|---|---|---|
| Individual Categories | ||
| Absolute weakness | ||
| Sensitivity | 69.8 | 69.1, 70.4 |
| Specificity | 51.6 | 50.1, 53.1 |
| PPV | 84.9 | 84.3, 85.4 |
| NPV | 30.5 | 29.4, 31.5 |
| Body mass index–normalized weakness | ||
| Sensitivity | 71.1 | 70.4, 71.7 |
| Specificity | 54.4 | 53.0, 55.9 |
| PPV | 85.8 | 85.3, 86.4 |
| NPV | 32.6 | 31.5, 33.6 |
| Body mass–normalized weakness | ||
| Sensitivity | 47.8 | 47.1, 48.6 |
| Specificity | 76.9 | 75.6, 78.1 |
| PPV | 88.9 | 88.3, 89.6 |
| NPV | 27.4 | 26.6, 28.2 |
| Collective Categories | ||
| Below ≥1 threshold | ||
| Sensitivity | 43.8 | 43.0, 44.5 |
| Specificity | 80.7 | 79.5, 81.8 |
| PPV | 89.8 | 89.2, 90.4 |
| NPV | 26.9 | 26.1, 27.7 |
| Below ≥2 thresholds | ||
| Sensitivity | 62.7 | 62.0, 63.4 |
| Specificity | 64.1 | 62.7, 65.5 |
| PPV | 87.1 | 86.6, 87.7 |
| NPV | 30.6 | 29.7, 31.5 |
| Below 3 thresholds | ||
| Sensitivity | 82.2 | 81.6, 82.7 |
| Specificity | 38.2 | 36.7, 39.6 |
| PPV | 83.8 | 83.2, 84.3 |
| NPV | 35.5 | 34.1, 36.9 |
Note: (A) First HGS measured and subsequent activities of daily living outcomes assessed; (B) all waves where HGS was measured and subsequent activities of daily living outcomes were assessed. HGS = handgrip strength; NPV = negative predictive value; PPV = positive predictive value.
The results for the associations of the individual weakness categories on future ADL status are shown in Table 3. Those below each cut-point had greater odds for future ADL limitations: 1.34 (CI: 1.22, 1.47) for absolute weakness, 1.36 (CI: 1.24, 1.50) for BMI-normalized weakness, and 1.56 (CI: 1.41, 1.73) for body mass–normalized weakness. Table 4 presents the results for the associations of the compounding weakness categories on future ADL limitation status. Individuals below 1, 2, and all 3 weakness cut-points had 1.36 (CI: 1.19, 1.55), 1.60 (CI: 1.42, 1.80), and 1.70 (CI: 1.50, 1.92) greater odds for future ADL limitations, respectively.
Table 3.
Associations of the individual weakness categories with future activities of daily living limitations.
| Crude |
Fully adjusted |
|||
|---|---|---|---|---|
| Odds ratio | 95% Confidence interval | Odds ratio | 95% Confidence interval | |
| Absolute weakness* | 1.96 | 1.81, 2.12 | 1.34 | 1.22, 1.47 |
| Body mass index–normalized weakness† | 2.02 | 1.86, 2.19 | 1.36 | 1.24, 1.50 |
| Body mass–normalized weakness‡ | 2.23 | 2.04, 2.43 | 1.56 | 1.41, 1.73 |
Note: Crude models controlled for activities of daily living limitation at current wave and follow-up years. Fully adjusted models controlled for activities of daily living limitation at current wave, follow-up years, sex, race, multimorbidity, obesity, age, smoking status, alcohol consumption, marital status, self-rated health, depressive status, moderate-to-vigorous physical activity participation, cognitive impairment, and educational achievement.
Reference: no absolute weakness;
Reference: no body mass index–normalized weakness;
Reference: no body mass–normalized weakness.
Table 4.
Associations of the compounding weakness categories on future activities of daily living limitations.
| Crude |
Fully adjusted |
|||
|---|---|---|---|---|
| Odds ratio | 95% Confidence interval | Odds ratio | 95% Confidence interval | |
| 1 Weakness category | 1.71 | 1.52, 1.93 | 1.36 | 1.19, 1.55 |
| 2 Weakness categories | 2.31 | 2.07, 2.59 | 1.60 | 1.41, 1.80 |
| 3 Weakness categories | 3.05 | 2.74, 3.40 | 1.70 | 1.50, 1.92 |
Reference: 0 weakness categories.
Note: Crude models controlled for activities of daily living limitation at current wave and follow-up years. Fully adjusted models controlled for activities of daily living limitation at current wave, follow-up years, sex, race, multimorbidity, obesity, age, smoking status, alcohol consumption, marital status, self-rated health, depressive status, moderate-to-vigorous physical activity participation, cognitive impairment, and educational achievement.
Figure 2 depicts the variability in the collective weakness categories with a Sankey bar chart. The proportions of participants classified as being below 1, 2, or all 3 weakness cut-points were fluid over time. Table 5 displays the results for the associations of the changes in weakness categories on future ADL limitations. Compared to persons who were never weak across waves, those in each changing weakness category had greater odds for future ADL limitations: 1.28 (CI: 1.01, 1.62) for onset weakness, 1.53 (CI: 1.22, 1.92) for persistent weakness, 1.72 (CI: 1.36, 2.19) for progressive weakness, and 1.34 (CI: 1.08, 1.66) for weakness recovery.
Figure 2.
Sankey bar chart for depicting fluidity in the collective weakness categories. Note: Weakness definitions denote participants below 0, 1, 2, or all 3 weakness cut-points. The absolute and normalized cut-points for males are <35.5 kg, <0.45 kg/kg, and <1.05 kg/kg/m2, respectively, while those for females are <20.0 kg, <0.337 kg/kg, and <0.79 kg/kg/m2, respectively.
Table 5.
Associations of the changing weakness categories on future activities of daily living limitations.
| Crude |
Fully adjusted |
|||
|---|---|---|---|---|
| Odds ratio | 95% Confidence interval | Odds ratio | 95% Confidence interval | |
| Onset weakness | 1.73 | 1.39, 2.16 | 1.28 | 1.01, 1.62 |
| Persistent weakness | 2.52 | 2.04, 3.11 | 1.53 | 1.22, 1.92 |
| Weakness progression | 2.83 | 2.27, 3.54 | 1.72 | 1.36, 2.19 |
| Weakness recovery | 1.60 | 1.30, 1.96 | 1.34 | 1.08, 1.66 |
Reference: never weak.
Note: Model included n = 9,878. Crude models controlled for activities of daily living limitation at current wave and follow-up years. Fully adjusted models controlled for activities of daily living limitation at current wave, follow-up years, sex, race, multimorbidity, obesity, age, smoking status, alcohol consumption, marital status, self-rated health, depressive status, moderate-to-vigorous physical activity participation, cognitive impairment, and educational achievement.
Discussion
The principal findings from our investigation revealed that older Americans below each weakness cut-point had increased odds for future ADL limitations. Specifically, those below the absolute, BMI-normalized, and body mass–normalized weakness cut-points had 34%, 36%, and 56% increased odds for future ADL limitations, respectively. When treating weakness cut-points collectively, persons below 1, 2, or all 3 cut-points had 36%, 60%, and 70% greater odds for future ADL limitations, respectively. Older Americans experiencing changes in their weakness status also had greater odds for future ADL limitations when compared to those classified as not weak. The SDOC cut-points provide a unique resource for categorizing weakness using absolute and body size–normalized thresholds, including capturing fluidity in weakness status.
Indeed, previous studies have demonstrated that being categorized as weak with absolute or body size–normalized cut-points has an increased risk for ADL limitations.5,7 For example, Delinocente et al.27 showed that HGS <32 kg (49.1% sensitivity and 79.8% specificity) in males was the best cut-point for identifying mobility limitations, while HGS <22 kg (58.6% sensitivity and 72.9% specificity) in females was likewise an optimal threshold. While the cut-points from Delinocente et al.27 are compatible with those from the SDOC, our findings indicate that the sensitivity and specificity of the SDOC cut-points for ADL limitations may differ. Factors related to low strength and changes in body composition are associated with ADL abilities.28, 29, 30 Previous work has similarly found that both absolute and body size–normalized HGS are linked to chronic disease and disability.8,31,32 Given that absolute strength and body composition influence ADLs but are subject to age-related changes, monitoring such changes in these characteristics over time may help to inform absolute and body size–normalized strength capacity assessments.33 Nonetheless, our findings suggest that using the absolute or body size–normalized weakness cut-points from the SDOC has utility for predicting ADL limitations.
Single cut-points are frequently utilized for examining weakness status,9 but misclassification of weakness and related health conditions from using stand-alone cut-points may exist. For example, the proportions of participants below the absolute and normalized cut-points used in this investigation are not equivalent, which aligns with another investigation that also showed imbalances in the proportion of persons below SDOC cut-points, including similar imbalances with respect to persons below certain weakness cut-points.15 Misclassifications for future ADL limitations were also different regarding individual weakness cut-points, such that sensitivity was particularly high for the absolute and BMI-normalized cut-points, but specificity was high for the body mass–normalized cut-points. Further, sensitivity for future ADL limitations increased as being below more weakness cut-points increased, but specificity decreased. These findings suggest that consideration should be given to how individual and collective weakness cut-points may predict basic self-care.
New absolute and body size–normalized weakness cut-points have been provided by the SDOC,13,14 which in turn will help to mitigate heterogeneity across HGS protocols. Although low HGS is associated with ADL limitations,5, 6, 7 the findings from our investigation provide advancements on topic of weakness and ADL limitations by presenting (1) individual absolute and body size–normalized weakness cut-points that come from a single source, (2) collective weakness classifications, and (3) how weakness status may change over time. While changes in weakness status may occur over time, which subsequently influences the magnitude of the odds for future ADL limitations, the absence of weakness (reference groups) is crucial for ADL limitation prevention. Accordingly, healthcare providers should measure HGS during routine visits as a vital sign, discuss the importance of strength capacity for health with their patients, and encourage preservation or restoration of strength through intervention when applicable.
When selecting weakness cut-points, awareness should be provided as to how such cut-points are used. For example, interventions that are relatively inexpensive and lower risk but higher benefit to participants (e.g., exercise) might be willing to inherit more weakness false positives as part of their screening and assessment of strength capacity because there could be few limitations to having older adults engage in such interventions when they may not truly have weakness. Alternatively, high-cost and burdensome interventions would benefit from cut-points with few false positives as part of screening and assessment of muscle strength for improving precision in study criteria. Future research should continue examining the role of the SDOC weakness cut-points, individually and collectively, alongside asymmetry for better classifying muscle dysfunction and subsequent age-related health conditions.
Some study limitations should be acknowledged. Participants were included if they had at least two observations. Other variables in the HRS that were not included in our investigation may have influenced the study outcomes. Validity may have been threatened in the HRS because different interviewers performed HGS assessments, and persons unable to perform HGS testing may have created biases in our estimates (probably underestimations). Self-report information may have also influenced classification of participants, which in turn, may generate residual confounding. However, these minor limitations are part of performing secondary analyses with population-based data. Additionally, those that may have died shortly after being interviewed likely had declining HGS and a functional limitation. We performed GEEs that typically had a 2-year mean follow-up period to better reflect changes in weakness and ADLs, but other modeling procedures may have yielded different estimates. Changes in weakness categorization from the body size–normalized cut-points could have been driven by HGS, body mass, or both. Other body size measurements such as stature10 were not considered in our analyses because normalized weakness cut-points for these anthropometric characteristics were not part of the SDOC.
Conclusions
This investigation revealed that older Americans below each absolute and body mass–normalized weakness cut-point had greater odds for future ADL limitations. Further, when these cut-points were utilized collectively, the odds for future limitations elevated as more persons below the weakness cut-points increased. Weakness status is fluid, and while preventing weakness is optimal for health, observing changes in weakness over time may guide intervention and inform prognosis of basic self-care. Our findings overall support the use of the SDOC weakness cut-points as a predictor for ADL limitations. Although being below each cut-point may help to detect problems with ADLs, diagnostic approaches could be strengthened by utilizing these absolute and body size–normalized cut-points together, including the observation of changes in weakness status over time. While determining weakness with these cut-points has advantages, feasibly enhancing muscle function assessments by collecting more attributes of muscle function may improve timeliness of intervention referrals and guide deployment of precise therapeutic strategies.
Funding
Research reported in this publication was supported by the National Institute on Aging of the National Institutes of Health under Award Number R15AG072348 (to RM) and R01AG067758 (to BCC). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
CRediT authorship contribution statement
Ryan McGrath: Conceptualization, Formal analysis, Methodology, Visualization, Writing – original draft. Brenda M. McGrath: Formal analysis, Methodology, Visualization, Writing – review & editing. Soham Al Snih: Methodology, Writing – review & editing. Peggy M. Cawthon: Methodology, Writing – review & editing. Brian C. Clark: Methodology, Writing – review & editing. Halli Heimbuch: Methodology, Writing – review & editing. Mark D. Peterson: Methodology, Writing – review & editing. Yeong Rhee: Methodology, Writing – review & editing.
Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered potential competing interests: PMC is a consultant and owns stock in Myocorps for work unrelated to this research. In the past 5 years, BCC has received grant support from NMD Pharma, Myolex Inc., and Astella Pharma Global Development for contracted studies related to aging and muscle health. BCC is also cofounder with equity of OsteoDx Inc.
Acknowledgments
The Health and Retirement Study is sponsored by the National Institute on Aging (U01AG009740) and is conducted by the University of Michigan.
Appendix 1. Participant Breakdown for Wave in Which Handgrip Strength and Future Activities of Daily Living Limitations Status Were Measured.
Appendix 1.
Note: A) first HGS measured and subsequent activities of daily living outcomes assessed; B) all waves where HGS was measured, and subsequent activities of daily living outcomes were assessed. HGS=handgrip strength.
References
- 1.Bhasin S, Travison TG, Manini TM, et al. Sarcopenia definition: the position statements of the sarcopenia definition and outcomes consortium. J Am Geriatr Soc. 2020;68(7):1410–1418. doi: 10.1111/jgs.16372. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.McGrath R, Johnson N, Klawitter L, et al. What are the association patterns between handgrip strength and adverse health conditions? A topical review. SAGE Open Med. 2020;8 doi: 10.1177/2050312120910358. 2050312120910358. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Dufner TJ, Fitzgerald JS, Lang JJ, Tomkinson GR. Temporal trends in the handgrip strength of 2,592,714 adults from 14 countries between 1960 and 2017: a systematic analysis. Sports Med. 2020;50:2175–2191. doi: 10.1007/s40279-020-01339-z. [DOI] [PubMed] [Google Scholar]
- 4.Beaudart C, Rolland Y, Cruz-Jentoft AJ, et al. Assessment of muscle function and physical performance in daily clinical practice. Calcif Tissue Int. 2019;105(1):1–14. doi: 10.1007/s00223-019-00545-w. [DOI] [PubMed] [Google Scholar]
- 5.Duchowny KA, Clarke PJ, Peterson MD. Muscle weakness and physical disability in older Americans: longitudinal findings from the U.S. Health and Retirement Study. J Nutr Health Aging. 2018;22(4):501–507. doi: 10.1007/s12603-017-0951-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.McGrath RP, Vincent BM, Lee IM, Kraemer WJ, Peterson MD. Handgrip strength, function, and mortality in older adults: a time-varying approach. Med Sci Sports Exerc. 2018;50(11):2259–2266. doi: 10.1249/MSS.0000000000001683. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.McGrath RP, Ottenbacher KJ, Vincent BM, Kraemer WJ, Peterson MD. Muscle weakness and functional limitations in an ethnically diverse sample of older adults. Ethn Health. 2020;25(3):342–353. doi: 10.1080/13557858.2017.1418301. [DOI] [PubMed] [Google Scholar]
- 8.Bohannon RW. Grip strength: an indispensable biomarker for older adults. Clin Interv Aging. 2019;14:1681. doi: 10.2147/CIA.S194543. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.McGrath R, Cawthon P, Clark B, Fielding R, Lang J, Tomkinson G. Recommendations for reducing heterogeneity in handgrip strength protocols. J Frailty Aging. 2022;11(2):143–150. doi: 10.14283/jfa.2022.21. [DOI] [PubMed] [Google Scholar]
- 10.Nevill AM, Tomkinson GR, Lang JJ, Wutz W, Myers TD. How should adult handgrip strength be normalized? Allometry reveals new insights and associated reference curves. Med Sci Sports Exerc. 2022;54(1):162–168. doi: 10.1249/MSS.0000000000002771. [DOI] [PubMed] [Google Scholar]
- 11.Cruz-Jentoft AJ, Bahat G, Bauer J, et al. Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing. 2019;48(1):16–31. doi: 10.1093/ageing/afy169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chen LK, Woo J, Assantachai P, et al. Asian Working Group for Sarcopenia: 2019 consensus update on sarcopenia diagnosis and treatment. J Am Med Dir Assoc. 2020;21(3):300–307. doi: 10.1016/j.jamda.2019.12.012. [DOI] [PubMed] [Google Scholar]
- 13.Manini TM, Patel SM, Newman AB, et al. Identification of sarcopenia components that discriminate slow walking speed: a pooled data analysis. J Am Geriatr Soc. 2020;68(7):1419–1428. doi: 10.1111/jgs.16524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Cawthon PM, Manini T, Patel SM, et al. Putative cut-points in sarcopenia components and incident adverse health outcomes: an SDOC analysis. J Am Geriatr Soc. 2020;68(7):1429–1437. doi: 10.1111/jgs.16517. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Patel SM, Duchowny KA, Kiel DP, et al. Sarcopenia definition & outcomes consortium defined low grip strength in two cross-sectional, population-based cohorts. J Am Geriatr Soc. 2020;68(7):1438–1444. doi: 10.1111/jgs.16419. [DOI] [PubMed] [Google Scholar]
- 16.Sonnega A, Faul JD, Ofstedal MB, Langa KM, Phillips JW, Weir DR. Cohort profile: the health and retirement study (HRS) Int J Epidemiol. 2014;43(2):576–585. doi: 10.1093/ije/dyu067. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Fisher GG, Ryan LH. Overview of the health and retirement study and introduction to the special issue. Work Aging Retire. 2018;4(1):1–9. doi: 10.1093/workar/wax032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sample Sizes and Response Rates. Health and Retirement Study. Available at: https://hrs.isr.umich.edu/sites/default/files/biblio/ResponseRates_2017.pdf. Accessed June 20, 2023.
- 19.HRS Data Book. Health and Retirement Study. Available at: https://hrs.isr.umich.edu/about/data-book. Accessed June 20, 2023.
- 20.Crimmins E, Guyer H, Langa K, Ofstedal MB, Wallace R, Weir D. Documentation of Physical Measures, Anthropometrics and Blood Pressure in the Health and Retirement Study. Available at: https://hrs.isr.umich.edu/sites/default/files/biblio/dr-011.pdf. Accessed June 20, 2023.
- 21.Plassman BL, Newman TT, Welsh KA, Helms M. Properties of the telephone interview for cognitive status: application in epidemiological and longitudinal studies. Cogn Behav Neurol. 1994;7(3):235–241. [Google Scholar]
- 22.Langa KM, Larson EB, Karlawish JH, et al. Trends in the prevalence and mortality of cognitive impairment in the United States: is there evidence of a compression of cognitive morbidity? Alzheimers Dement. 2008;4(2):134–144. doi: 10.1016/j.jalz.2008.01.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Turvey CL, Wallace RB, Herzog R. A revised CES-D measure of depressive symptoms and a DSM-based measure of major depressive episodes in the elderly. Int Psychogeriatr. 1999;11(2):139–148. doi: 10.1017/s1041610299005694. [DOI] [PubMed] [Google Scholar]
- 24.Naseri P, Majd HA, Kariman N, Sourtiji A. Comparison of generalized estimating equations (GEE), mixed effects models (MEM) and repeated measures ANOVA in analysis of menorrhagia data. Arch Adv Biosci. 2016;7(1):32–40. [Google Scholar]
- 25.Wang M. Generalized estimating equations in longitudinal data analysis: a review and recent developments. Adv Stat. 2014; 2014( 1): 1– 11. [Google Scholar]
- 26.Rosanbalm S. Getting Sankey With Bar Charts. Available at: https://www.pharmasug.org/proceedings/2015/DV/PharmaSUG-2015-DV07.pdf. Accessed June 20, 2023.
- 27.Delinocente MLB, de Carvalho DHT, de Oliveira Máximo R, et al. Accuracy of different handgrip values to identify mobility limitation in older adults. Arch Gerontol Geriatr. 2021;94 doi: 10.1016/j.archger.2021.104347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Lv YB, Yuan JQ, Mao C, et al. Association of body mass index with disability in activities of daily living among Chinese adults 80 years of age or older. JAMA Netw Open. 2018;1(5) doi: 10.1001/jamanetworkopen.2018.1915. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Ozturk GZ, Egici MT, Bukhari MH, Toprak D. Association between body mass index and activities of daily living in homecare patients. Pak J Med Sci. 2017;33(6):1479. doi: 10.12669/pjms.336.13748. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Snih SA, Markides KS, Ottenbacher KJ, Raji MA. Hand grip strength and incident ADL disability in elderly Mexican Americans over a seven-year period. Aging Clin Exp Res. 2004;16:481–486. doi: 10.1007/BF03327406. [DOI] [PubMed] [Google Scholar]
- 31.Peterson MD, Duchowny K, Meng Q, Wang Y, Chen X, Zhao Y. Low normalized grip strength is a biomarker for cardiometabolic disease and physical disabilities among US and Chinese adults. J Gerontol Ser Biomed Sci Med Sci. 2017;72(11):1525–1531. doi: 10.1093/gerona/glx031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.McGrath RP, Kraemer WJ, Snih SA, Peterson MD. Handgrip strength and health in aging adults. Sports Med. 2018;48(9):1993–2000. doi: 10.1007/s40279-018-0952-y. [DOI] [PubMed] [Google Scholar]
- 33.McGrath R. Comparing absolute handgrip strength and handgrip strength normalized to body weight in aging adults. Aging Clin Exp Res. 2019;31(12):1851–1853. doi: 10.1007/s40520-019-01126-5. [DOI] [PubMed] [Google Scholar]