- Email: [email protected]
Is strength training associated with mortality benefits? A 15 year cohort study of US older adults
Is strength training associated with mortality benefits? A 15 year cohort study of US older adults Jennifer L. Kraschnewski, Christopher N. Sciamanna, Jennifer M. Poger, Liza S. Rovniak, Erik B. Lehman, Amanda B. Cooper, Noel H. Ballentine, Joseph T. Ciccolo PII: DOI: Reference:
S0091-7435(16)30016-0 doi: 10.1016/j.ypmed.2016.02.038 YPMED 4561
To appear in:
Preventive Medicine
Received date: Revised date: Accepted date:
6 November 2015 10 February 2016 23 February 2016
Please cite this article as: Kraschnewski Jennifer L., Sciamanna Christopher N., Poger Jennifer M., Rovniak Liza S., Lehman Erik B., Cooper Amanda B., Ballentine Noel H., Ciccolo Joseph T., Is strength training associated with mortality benefits? A 15 year cohort study of US older adults, Preventive Medicine (2016), doi: 10.1016/j.ypmed.2016.02.038
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Title Page Is strength training associated with mortality benefits? A 15 year cohort study of US older adults.
PT
Jennifer L. Kraschnewski, M.D., MPH1,2,3, Christopher N. Sciamanna, M.D.,1,2,3, MPH, Jennifer M. Poger, M.Ed.1, Liza S. Rovniak, Ph.D., MPH1,3, Erik B. Lehman, MS3, Amanda B. Cooper, M.D.2, Noel H. Ballentine,
RI
M.D., F.A.C.P.2, & Joseph T. Ciccolo, Ph.D.4 1
Department of Medicine, Penn State College of Medicine, 2Department of Medicine, Penn State
4
NU
Department of Applied Physiology, Columbia University
SC
Hershey Medical Center, 3Department of Public Health Sciences, Penn State College of Medicine,
Corresponding Author:
MA
Jennifer L. Kraschnewski, MD, MPH
Assistant Professor of Medicine and Public Health Sciences
D
Associate Director, MD-MPH Program
Department of Medicine
TE
Medical Director of Research, Penn State Hershey PRO Wellness Center
Division of General Internal Medicine
AC CE P
Penn State Hershey Medical Center 500 University Drive, HO34 Hershey, PA 17033 (717) 531-8161
FAX: (717) 531-7726
[email protected]
Abstract Word Count: 246 Article Word Count: 2,950 Number of References: 55 Number of Tables: 4 Number of Figures: 0
1
ACCEPTED MANUSCRIPT Abstract Background:
PT
The relationship between strength training (ST) behavior and mortality remains understudied in large, national samples, although smaller studies have observed that greater amounts of muscle strength are
SC
guidelines and future mortality in an older US adult population.
RI
associated with lower risks of death. We aimed to understand the association between meeting ST
Methods:
NU
Data were analyzed from the 1997-2001 National Health Interview Survey (NHIS) linked to death certificate data in the National Death Index. The main independent variable was guideline-concordant
MA
ST (i.e. twice each week) and dependent variable was all-cause mortality. Covariates identified in the literature and included in our analysis were demographics, past medical history, and other health behaviors (including other physical activity). Given our aim to understand outcomes in older adults,
AC CE P
Results:
TE
multiple logistic regression analysis.
D
analyses were limited to adults age 65 years and older. Multivariate analysis was conducted using
During the study period, 9.6% of NHIS adults age 65 and older (N=30,162) reported doing guidelineconcordant ST and 31.6% died. Older adults who reported guideline-concordant ST had 46% lower odds of all-cause mortality than those who did not (adjusted odds ratio: 0.64; 95% CI: 0.57, 0.70; p<0.001). The association between ST and death remained after adjustment for past medical history and health behaviors.
Conclusions: Although a minority of older US adults met ST recommendations, guideline-concordant ST is significantly associated with decreased overall mortality. All-cause mortality may be significantly reduced through the identification of and engagement in guideline-concordant ST interventions by older adults.
Keywords: strength training; mortality; resistance training; older adults
2
ACCEPTED MANUSCRIPT Introduction
Engaging in regular physical activity reaps a multitude of health benefits for adults of all ages; it is
PT
perhaps of greatest value in preventing premature mortality associated with all-cause, cardiovascular disease, diabetes, and even some cancers.1-11 Significant quality of life improvements and decreased risk
RI
of mortality have also been observed in physically active, older adults.4,9,12-16 Unlike aerobic exercise’s well-established effects, strength training has only recently garnered attention for its benefits in
SC
regaining muscle mass and strength often depleted with age and disability.17-20 While guidelineconcordant physical activity has consistent and powerful relationships with longer life expectancy,
NU
meeting strength training guidelines may also play an important role in decreasing premature mortality
MA
and warrants further investigation.
Although physical activity guidelines have been widely available for decades from the American College
D
of Sports Medicine (ACSM),21 specific recommendations regarding strength training have only been developed more recently; in 2007, the American Heart Association (AHA) together with the ACSM
TE
established recommendations encouraging all adults to participate in strength training activities at least twice each week,19 with a similar emphasis on muscle-strengthening activity for older adults.22 In
AC CE P
addition, the Behavioral Risk Factor Surveillance System (BRFSS), which began assessing aerobic activity with the first survey in 1984, only began assessing strength training in 2011.23 Over this time, studies have demonstrated robust effects of strength training on strength, muscle mass and physical function as well as improvement in a range of chronic conditions, including diabetes, osteoporosis, low back pain, and obesity.7,18,24-26 Recently, Krist and colleagues observed considerable improvements in muscle strength and mobility among elderly nursing home residents with impaired mobility at baseline after they engaged in a resistance exercise program twice weekly over a two month period.17 Similarly, Mayer and colleagues found that progressive strength training in the elderly helped retain motor function and reduce sarcopenia.20 Older adults with functional limitations can achieve significant improvements in physical function and muscle strength by engaging in strength training activity.27
In addition to improvements in physical function, several clinical studies have demonstrated that higher levels of muscular strength are associated with lower all-cause, cardiovascular, and cancer mortality risk, among both men and the elderly.28-30 Findings from 15 epidemiological and 8 clinical studies summarized in the 2015 systematic review by Volaklis et al. provides evidence that muscular strength is 3
ACCEPTED MANUSCRIPT inversely and independently associated with all-cause and cardiovascular mortality.31 Although there is increasing evidence illuminating the substantive benefits of muscle strength on decreased mortality risk,31 no studies to our knowledge describe the relationship between strength training behavior and
PT
mortality in a large, nationally representative sample over an extended time period, particularly in older
RI
adults.
SC
The aim of this investigation was to understand the association between meeting strength training guidelines and future mortality in older adults. We hypothesized that guideline-concordant strength
NU
training is significantly associated with decreased overall mortality in older US adults.
MA
Methods
In the present cohort study, longitudinal data from the 1997-2001 National Health Interview Survey (NHIS), linked to death certificate data from the National Center for Health Statistics National Death
D
Index,32 were analyzed. Survey sampling methods and data collection procedures have been described
TE
extensively elsewhere.33 Briefly, the NHIS employs a multi-stage stratified sampling method to collect overall health, disease, and disability data of the US population from a nationally representative
AC CE P
sampling of all 50 states and the District of Columbia .33 Personal household interviews conducted by Census interviewers collect basic demographic and family relationship information about all household members. Additionally, one randomly selected adult in each household is interviewed in detail regarding their health and health-related behaviors through the Sample Adult questionnaire, which was analyzed in this study. Our analysis was limited to those individuals age 65 years and older, given our interest in assessing strength training activity in the Medicare-eligible population. Three years of the study were used (1997, 1998, 1999) to provide enough data, including deaths (through 2011), to allow for significant relationships to emerge after adjusting for potential confounders. This study was exempt from Institutional Review Board approval.
The main independent variable, guideline-concordant strength training, was assessed using the following question: “How often do you do leisure-time physical activities specifically designed to strengthen your muscles, such as lifting weight or doing calisthenics?” Participant responses included both the number of times strength training was performed and the unit of time (i.e.“per week,” “per month”). These responses were categorized to signify whether the individual performed these activities 4
ACCEPTED MANUSCRIPT at least twice each week, consistent with the ACSM/AHA Guidelines.34 Additionally, in order to assess how often moderate and vigorous physical activity was performed, participants were asked the following questions: How often do you do light or moderate activities for at least 10 minutes that cause
PT
only light sweating or a slight to moderate increase in breathing or heart rate?”; “About how long do you do these light or moderate activities each time?”; “How often do you do vigorous activities for at
RI
least 10 minutes that cause heavy sweating or large increases in breathing or heart rate?”; “About how
SC
long do you do these vigorous activities each time?” Physical activity was categorized according to the 2008 Physical Activity Guidelines for Americans: highly active (≥300 min per week of light- to moderateintensity aerobic activity, ≥150 min per week of vigorous-intensity aerobic activity or in combination);
NU
sufficiently active (150–300 min per week of light- to moderate-intensity aerobic activity, 75–150 min per week of vigorous-intensity aerobic activity or in combination); insufficiently active (some activity but
MA
not enough to meet the more active definitions); and inactive (no light-to moderate- or vigorousintensity activity of ≥10 min per session).35-37 Minutes of moderate and vigorous intensity activity were
D
re-categorized, based on whether the individual met the ACSM/AHA Guidelines for moderate intensity aerobic activity (≥ 150 minutes per week) and vigorous intensity aerobic activity (≥ 75 minutes per
TE
week). The main outcome variable was mortality (all-cause, cancer-related, and cardiac-related) as identified from the National Death Index.32 The cause of death classification was based on the
AC CE P
International Classification of Disease codes (ICD-10). The following codes were included to define each cause: Cancer (19-44); Cardiac (53-75).
The following demographic variables were included as covariates in the analysis: age, gender, race (Caucasian, African-American, Other), ethnicity (Hispanic vs. non-Hispanic), education level (less than high school, high school or GED, some college/ associate's degree, bachelor degree or higher), and marital status (married/living with partner, widowed, divorced, other). Additionally, we included health behaviors (body mass index (BMI), smoking, alcohol intake, physical activity) and comorbid conditions (hypertension, diabetes, heart disease, non-skin cancer), consistent with variables shown to have associations with the outcome in other studies.35,38 Based on cut-points from the National Institutes of Health, BMI was re-categorized into normal, overweight and obese.39
A weighted analysis employing procedures in SAS 9.4 (SAS Institute, Cary, NC) which take into consideration the complex sampling stratification and clustering incorporated by the NHIS study was used for all analyses. Using weighted logistic regression, bivariate comparisons were made between all5
ACCEPTED MANUSCRIPT cause mortality, cancer-related mortality, or cardiac-related mortality and strength training as well as demographic variables, health behaviors, and comorbid conditions. Weighted multivariable logistic regression models were then utilized to further investigate the association between strength training
PT
and the types of mortality while adjusting for the three groups of covariates: demographic variables only, both demographic variables and health behaviors, and finally demographic variables, health
RI
behaviors, and comorbid conditions. Odds ratios were calculated for all logistic regression analyses to
SC
quantify the direction and magnitude of the associations.
A confirmatory analysis was then performed using propensity scores.40,41 Three sets of propensity
NU
scores, one set for each group of covariates from the primary analysis, were calculated based on the predicted probabilities from a weighted logistic regression incorporating strength training as the
MA
outcome and the covariates as the predictors. The three sets of propensity scores where then applied in two ways. First, they were added as the only covariate for adjustment as a substitute for the set of
D
predictors used in their construction in a weighted multivariable logistic regression model with strength training as the predictor and all-cause mortality, cancer-related mortality, or cardiac-related mortality as
TE
the outcomes. Second, they were used in a 1-to-1 method of matching subjects with strength training <2 times/week to subjects with strength training ≥2 times/week. A weighted logistic regression was
AC CE P
then performed using this matched sample including strength training as the only predictor in the model, given that the sample was already adjusted for the covariates through the matching process.
Results
Overall, 9.6% of NHIS adults 65 years of age and older (N=30,162) reported doing strength training at least twice weekly, consistent with guidelines from the ASCM/AHA,19 and 31.6% of the study population died during the 15 year follow-up period. Table 1 displays demographic, health behavior, and comorbid condition characteristics associated with strength training frequency. Older adults who met guidelines for strength training were, on average, slightly younger and were more likely to be married white males with higher levels of education. Individuals who had a normal body weight, engaged in aerobic exercise, and abstained from alcohol and tobacco use were also more likely to meet strength training guidelines. Finally, older adults who had guideline-concordant strength training were less likely to have diabetes and hypertension, but did not differ in terms of coronary heart disease or cancer status from their lessactive counterparts. 6
ACCEPTED MANUSCRIPT
The bivariate relationships between mortality and demographic (Group 1), health behavior (Group 2) and comorbid condition (Group 3) variables are presented in Table 2. The bivariate relationships
PT
between meeting strength training guidelines and the other covariates were also examined. Older adults who performed strength training at least twice a week had 45% lower odds of all-cause mortality, 19%
SC
RI
lower odds of death from cancer, and 41% lower odds of cardiac death.
As shown in Table 3, after adjusting for demographic covariates (Group 1), those who reported guideline-concordant strength training had 36% lower odds of all-cause mortality than those who did
NU
not (adjusted odds ratio: 0.64; 95% CI: 0.57, 0.70; p<0.001). There remained a statistically significant decrease in odds of dying from cancer and cardiac etiologies as well; 17% (adjusted odds ratio: 0.83;
MA
95% CI: 0.70, 0.98; p<0.05) and 28% (adjusted odds ratio: 0.72; 95% CI: 0.62, 0.83; p<0.001) respectively. After adjusting for demographic variables, health behaviors, and comorbid conditions (Groups 1, 2, & 3),
D
all-cause mortality remained significant, with a 19% decreased odds of dying (see Table 2). However, the point estimate of the relationship between meeting strength training guidelines and cancer and cardiac
AC CE P
remained in the same direction.
TE
mortality after controlling for these covariates was no longer statistically significant, although it
Analyses adjusted for the propensity scores in Table 4 show similar results. There was 37% lower odds of all-cause mortality after adjusting for demographic covariates (Group 1) (adjusted odds ratio: 0.63; 95% CI: 0.57, 0.70; p<0.001). Similarly, there remained a statistically significant decrease in all-cancer and allcardiac deaths, although cancer odds were not as strongly associated, as also evident in Table 3 analyses; 19% (adjusted odds ratio: 0.81; 95% CI: 0.68, 0.96; p<0.05) and 31% (adjusted odds ratio: 0.69; 95% CI: 0.59, 0.80; p<0.001) respectively. A similar pattern remained when adjusting for demographic, health behavior, and comorbid condition variables (Groups 1, 2, & 3) with the adjusted propensity scores; there was 18% lower odds of all-cause mortality (adjusted odds ratio: 0.82; 95% CI: 0.73, 0.92; p<0.001), while all-cancer and all-cardiac mortalities showed a similar trend but were no longer statistically significant. When matched by propensity scores, there was 31% lower odds of all-cause mortality and 24% lower odds of cardiac death when adjusting for demographic variables (Group 1). Additionally, there was a statistically significant decrease in all-cause mortality when controlling for demographic, health behavior, and comorbid condition variables (Groups 1, 2, & 3) using propensity
7
ACCEPTED MANUSCRIPT score matching (adjusted odds ratio: 0.84; 95% CI: 0.72, 0.96; p<0.05). All-cancer deaths were no longer statistically significant when controlling for any of the covariates in the matching analysis.
PT
Discussion
RI
The purpose of the present study was to understand the relationship between meeting strength training guidelines and future mortality. Overall, the results showed a consistent and inverse association in a
SC
large, national sample;42 those who performed guideline-concordant strength training had significantly lower odds of all cause, cancer, and cardiac death than their less-active counterparts. This association
NU
remained significant for all-cause mortality after adjusting for each of the covariate groups (demographic, health behavior, and comorbid condition variables, respectively). In addition, these
MA
patterns remained significant, albeit less strongly, after adjusting specifically for demographic variables in cancer and cardiac mortalities. Therefore, older adults who perform strength training activities not
D
only improve their physical functioning as previously demonstrated, 17,20,43 but their survival rate as well.
TE
Although not statistically significant in cancer and cardiac deaths, point estimates suggest that guidelineconcordant strength training behaviors are associated with decreased risk of cause-specific mortality.
AC CE P
This weaker association may be the result of sample size, as a lesser percentage of individuals died from cancer (N=2,192; 7.2%) or cardiac (N=4,275; 13.4%) etiologies than from all-causes.
Results from the propensity score analyses confirmed our findings from the weighted logistic regression models. After matching subjects who strength trained <2 times/week to subjects who strength trained ≥2 times/week in the propensity score matching, there remained a significantly lower odds of all-cause mortality when controlling for all variables (Groups 1, 2, & 3). Although sample size was greatly reduced in the propensity score matching, as those who met strength training guidelines was a much smaller group than those who were not guideline concordant (N=5,408), this reduction in power is balanced by the efficiency in matching.44
Although few studies have examined the impact of strength training on future mortality, existing outcomes are similar to the results of the present study.45,46 For example, Singh and colleagues performed a randomized controlled trial involving 12 months of high-intensity weight-lifting exercise as part of a multidisciplinary intervention for patients recovering from hip fractures and found an 81% 8
ACCEPTED MANUSCRIPT reduction in mortality (N=4) compared to the usual care control condition (N=8) (adjusted odds ratio: 0.19; 95% CI; 0.04, 0.91; p < .04).45 Similarly, Hardee and colleagues conducted a prospective epidemiological investigation of self-reported physical activity and resistance training among cancer
PT
survivors and observed a 33% lower risk of all-cause mortality among those who participated in resistance training ≥ 1 day per week (N=39) (95% CI, 0.45-0.99) after adjusting for potential
RI
confounders, including physical activity. No correlation was found with those who engaged in physical activity only.46 While these studies are limited by sample size, the present study benefits from using a
SC
large, nationally representative sample of a range of health outcomes for the US population.42
NU
Consistent with prior findings, meeting strength training guidelines was strongly associated with younger age, among other demographic and health behavior variables, including meeting physical
MA
activity guidelines and not smoking.34,35,47 Ciccolo and colleagues (2010) in a separate survey of 9,651 US adults observed that only 21.7% of older adults met strength training guidelines versus 37.5% of adults
D
aged 35–54.38 This was consistent with our small percentage of older adults performing guidelineconcordant strength training activities (N=2,725; 9.6%). After controlling for physical activity level, which
TE
was strongly associated with reduced mortality, specifically reporting strength exercises appeared to
AC CE P
confer additional benefit beyond reporting physical activity alone.
This study has several important limitations to consider. First, the study design is that of a cohort study, limiting ability to determine cause and effect. In addition, participation in strength training was selfreported without information regarding session quality or length. Self-report measures remain widely used for physical activity assessment, however, typically overestimate activity level when compared with objective measurement.48 Unmeasured confounders, including diet and self-report of health status could have obscured the data. Unfortunately, dietary habits and objective physical and biological data are difficult to assess and beyond the scope of this secondary data analysis. Although not ideal, selfreport remains a reliable and valid indicator of health.49 Another limitation is that respondents may have misinterpreted the survey questions or may not have fully understood the definitions of strength training and calisthenics, which may have weakened the accuracy of strength training prevelance.34 Additionally, given that calisthenics isn’t always categorized as progressive training, the survey question used from the NHIS combining strength training with calisthenics may have undesirably included some aerobic activity as well. Since respondents were not asked to provide details regarding the type of strength training activities performed, some activities such as stair climbing could have been missed.34 9
ACCEPTED MANUSCRIPT Although it is possible that a small percentage of older adults were physically unable to perform strength training activities, potentially confounding the data, the literature suggests otherwise; Drey and colleagues (2012), for example, randomized older adults (>65 years) into either a strength training (ST),
PT
power training (PT) or control group and found the highest fidelity in the ST group, with 87% attendance rates over 12 weeks.50 A final limitation is this study relied on death certificate data, which has been
RI
previously questioned for its reliability and accuracy in properly identifying cause of death.51,52 However, when used for observational studies the National Death Index has been shown to be accurate in
SC
ascertaining deaths, with high sensitivity (96%) and specificity (100%).53
NU
Conclusion
MA
Despite guidelines, only 21.7% of older adults currently meet recommendations of strength training at least twice each week, substantially lower than the 2010 national objective of 30%.54 This underscores the need for additional programs to increase strength training among older adults to help engage
D
patients in the “Exercise is Medicine” campaign.34 This study further demonstrates the importance of
TE
encouraging doctors to recommend physical activity to patients as they would any other effective treatment, particularly since exercise has been shown to be as effective as some medications.55
AC CE P
Identifying interventions to successfully engage older adults in guideline-concordant strength training has the potential to significantly reduce all-cause mortality in this population.
Conflict of Interest Statement
The authors declare that there are no conflicts of interest.
10
ACCEPTED MANUSCRIPT References 1.
Moore SC, Patel AV, Matthews CE, et al. Leisure time physical activity of moderate to vigorous intensity and mortality: a large pooled cohort analysis. PLoS Med. 2012;9(11):e1001335. Autenrieth CS, Baumert J, Baumeister SE, et al. Association between domains of physical activity
PT
2.
and all-cause, cardiovascular and cancer mortality. Eur J Epidemiol. 2011;26(2):91-99. Samitz G, Egger M, Zwahlen M. Domains of physical activity and all-cause mortality: systematic
RI
3.
SC
review and dose-response meta-analysis of cohort studies. Int J Epidemiol. 2011;40(5):13821400. 4.
Wen CP, Wai JP, Tsai MK, et al. Minimum amount of physical activity for reduced mortality and
5.
NU
extended life expectancy: a prospective cohort study. Lancet. 2011;378(9798):1244-1253. Woodcock J, Franco OH, Orsini N, Roberts I. Non-vigorous physical activity and all-cause
MA
mortality: systematic review and meta-analysis of cohort studies. Int J Epidemiol. 2011;40(1):121-138.
Lollgen H, Bockenhoff A, Knapp G. Physical activity and all-cause mortality: an updated meta-
D
6.
analysis with different intensity categories. Int J Sports Med. 2009;30(3):213-224. Davidson LE, Hudson R, Kilpatrick K, et al. Effects of exercise modality on insulin resistance and
TE
7.
functional limitation in older adults: a randomized controlled trial. Arch Intern Med.
8.
AC CE P
2009;169(2):122-131.
Nocon M, Hiemann T, Muller-Riemenschneider F, Thalau F, Roll S, Willich SN. Association of physical activity with all-cause and cardiovascular mortality: a systematic review and metaanalysis. Eur J Cardiovasc Prev Rehabil. 2008;15(3):239-246.
9.
Wannamethee SG, Shaper AG. Physical activity in the prevention of cardiovascular disease: an epidemiological perspective. Sports Med. 2001;31(2):101-114.
10.
Davey Smith G, Shipley MJ, Batty GD, Morris JN, Marmot M. Physical activity and cause-specific mortality in the Whitehall study. Public Health. 2000;114(5):308-315.
11.
Haapanen N, Miilunpalo S, Vuori I, Oja P, Pasanen M. Characteristics of leisure time physical activity associated with decreased risk of premature all-cause and cardiovascular disease mortality in middle-aged men. Am J Epidemiol. 1996;143(9):870-880.
12.
Brown WJ, McLaughlin D, Leung J, et al. Physical activity and all-cause mortality in older women and men. Br J Sports Med. 2012;46(9):664-668.
13.
Ueshima K, Ishikawa-Takata K, Yorifuji T, et al. Physical activity and mortality risk in the Japanese elderly: a cohort study. Am J Prev Med. 2010;38(4):410-418. 11
ACCEPTED MANUSCRIPT 14.
Leitzmann MF, Park Y, Blair A, et al. Physical activity recommendations and decreased risk of mortality. Arch Intern Med. 2007;167(22):2453-2460.
15.
Sundquist K, Qvist J, Sundquist J, Johansson SE. Frequent and occasional physical activity in the
16.
PT
elderly: a 12-year follow-up study of mortality. Am J Prev Med. 2004;27(1):22-27. Wannamethee SG, Shaper AG, Walker M. Physical activity and mortality in older men with
Krist L, Dimeo F, Keil T. Can progressive resistance training twice a week improve mobility,
SC
17.
RI
diagnosed coronary heart disease. Circulation. 2000;102(12):1358-1363.
muscle strength, and quality of life in very elderly nursing-home residents with impaired mobility? A pilot study. Clin Interv Aging. 2013;8:443-448.
Gennuso KP, Zalewski K, Cashin SE, Strath SJ. Resistance training congruent with minimal
NU
18.
guidelines improves function in older adults: a pilot study. J Phys Act Health. 2013;10(6):769-
19.
MA
776.
Haskell WL, Lee IM, Pate RR, et al. Physical activity and public health: updated recommendation
D
for adults from the American College of Sports Medicine and the American Heart Association. Circulation. 2007;116(9):1081-1093.
Mayer F, Scharhag-Rosenberger F, Carlsohn A, Cassel M, Muller S, Scharhag J. The intensity and
TE
20.
effects of strength training in the elderly. Dtsch Arztebl Int. 2011;108(21):359-364. American College of Sports Medicine position statement on the recommended quantity and
AC CE P
21.
quality of exercise for developing and maintaining fitness in healthy adults. Med Sci Sports. 1978;10(3):vii-x. 22.
Nelson ME, Rejeski WJ, Blair SN, et al. Physical activity and public health in older adults: recommendation from the American College of Sports Medicine and the American Heart Association. Circulation. 2007;116(9):1094-1105.
23.
Centers for Disease Control and Prevention. Adult Participation in Aerobic and MuscleStrengthening Physical Activities — United States, 2011. MMWR. 2013;62(17):326-330.
24.
Nelson ME, Fiatarone MA, Morganti CM, Trice I, Greenberg RA, Evans WJ. Effects of highintensity strength training on multiple risk factors for osteoporotic fractures. A randomized controlled trial. JAMA. 1994;272(24):1909-1914.
25.
Fiatarone MA, O'Neill EF, Ryan ND, et al. Exercise training and nutritional supplementation for physical frailty in very elderly people. N Engl J Med. 1994;330(25):1769-1775.
12
ACCEPTED MANUSCRIPT 26.
Pollock MV, KR. Resistance training for health. The President's Council on Physical Fitness and Sports. https://www.presidentschallenge.org/informed/digest/docs/199612digest.pdf. Accessed August 27, 2015. Kraschnewski JL, Sciamanna CN, Ciccolo JT, et al. Is exercise used as medicine? Association of
PT
27.
meeting strength training guidelines and functional limitations among older US adults. Prev
Ortega FB, Silventoinen K, Tynelius P, Rasmussen F. Muscular strength in male adolescents and
SC
28.
RI
Med. 2014;66:1-5.
premature death: cohort study of one million participants. BMJ. 2012;345:e7279. 29.
Ruiz JR, Sui X, Lobelo F, et al. Muscular strength and adiposity as predictors of adulthood cancer
30.
NU
mortality in men. Cancer Epidemiol Biomarkers Prev. 2009;18(5):1468-1476. Ruiz JR, Sui X, Lobelo F, et al. Association between muscular strength and mortality in men:
31.
MA
prospective cohort study. BMJ. 2008;337:a439.
Volaklis KA, Halle M, Meisinger C. Muscular strength as a strong predictor of mortality: A
32.
D
narrative review. Eur J Intern Med. 2015;26(5):303-310. Centers for Disease Control and Prevention. National Death Index.
33.
TE
http://www.cdc.gov/nchs/ndi.htm. Accessed August 27, 2015. Centers for Disease Control and Prevention and U.S. Department of Health and Human Services.
2012.
AC CE P
2011 National Health Interview Survey (NHIS) Public Use Data Release NHIS Survey Description.
ftp://ftp.cdc.gov/pub/Health_statistics/NCHs/Dataset_Documentation/NHIS/2011/srvydesc.pdf. Accessed July 6, 2015. 34.
Centers for Disease Control and Prevention. Trends in Strength Training --- United States, 1998-2004. MMWR. 2006;55(28):769-772.
35.
Carlson SA, Fulton JE, Schoenborn CA, Loustalot F. Trend and prevalence estimates based on the 2008 Physical Activity Guidelines for Americans. Am J Prev Med. 2010;39(4):305-313.
36.
Schoenborn CA, Adams PE. Health behaviors of adults: United States, 2005-2007. Vital Health Stat 10. 2010(245):1-132.
37.
U. S. Department of Health and Human Services. Physical Activity Guidelines for Americans. 2008. http://www.health.gov/paguidelines/guidelines/default.aspx. Accessed July 31, 2015.
13
ACCEPTED MANUSCRIPT 38.
Ciccolo JT, Pettee Gabriel KK, Macera C, Ainsworth BE. Association between self-reported resistance training and self-rated health in a national sample of U.S. men and women. J Phys Act Health. 2010;7(3):289-298. National Institutes of Health. Clinical Guidelines on the Identification, Evaluation, and Treatment
PT
39.
of Overweight and Obesity in Adults--The Evidence Report. Obes Res. 1998;6(2):51S-209S. Austin PC. An Introduction to Propensity Score Methods for Reducing the Effects of Confounding
RI
40.
41.
Rosenbaum P, Rubin, DB. The Central Role of the Propensity Score in Observational Studies for Causal Effects. Biometrika. 1983;70(1):41-55.
Miller D, Fernandez CA, Lee DJ. National Health Interview Survey. In: Gellman MD, Turner JR,
NU
42.
SC
in Observational Studies. Multivariate Behav Res. 2011;46(3):399-424.
eds. Encyclopedia of Behavioral Medicine. New York, NY: Springer New York; 2013:1286-1288. Kennis E, Verschueren SM, Bogaerts A, Van Roie E, Boonen S, Delecluse C. Long-term impact of
MA
43.
strength training on muscle strength characteristics in older adults. Arch Phys Med Rehabil.
44.
D
2013;94(11):2054-2060.
Polkinghorne KR, McDonald SP, Atkins RC, Kerr PG. Vascular access and all-cause mortality: a
45.
TE
propensity score analysis. J Am Soc Nephrol. 2004;15(2):477-486. Singh NA, Quine S, Clemson LM, et al. Effects of high-intensity progressive resistance training
AC CE P
and targeted multidisciplinary treatment of frailty on mortality and nursing home admissions after hip fracture: a randomized controlled trial. J Am Med Dir Assoc. 2012;13(1):24-30. 46.
Hardee JP, Porter RR, Sui X, et al. The effect of resistance exercise on all-cause mortality in cancer survivors. Mayo Clin Proc. 2014;89(8):1108-1115.
47.
Chevan J. Demographic determinants of participation in strength training activities among U.S. adults. J Strength Cond Res. 2008;22(2):553-558.
48.
Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, McDowell M. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc. 2008;40(1):181-188.
49.
Dalen JD, Huijts T, Krokstad S, Eikemo TA. Are there educational differences in the association between self-rated health and mortality in Norway? The HUNT Study. Scand J Public Health. 2012;40(7):641-647.
50.
Drey M, Zech A, Freiberger E, et al. Effects of strength training versus power training on physical performance in prefrail community-dwelling older adults. Gerontology. 2012;58(3):197-204.
14
ACCEPTED MANUSCRIPT 51.
Trepka MJ, Maddox LM, Lieb S, Niyonsenga T. Utility of the National Death Index in ascertaining mortality in acquired immunodeficiency syndrome surveillance. Am J Epidemiol. Jul 1 2011;174(1):90-98. Smith Sehdev AE, Hutchins GM. Problems with proper completion and accuracy of the cause-ofdeath statement. Arch Intern Med. 2001;161(2):277-284.
Stampfer MJ, Willett WC, Speizer FE, et al. Test of the National Death Index. Am J Epidemiol.
RI
53.
SC
1984;119(5):837-839. 54.
PT
52.
Centers for Disease Control and Prevention. Adult participation in aerobic and musclestrengthening physical activities — United States, 2011. MMWR. 2013;62(17):326-330.
NU
Naci H, Ioannidis JP. Comparative effectiveness of exercise and drug interventions on mortality
TE
D
MA
outcomes: metaepidemiological study. BMJ. 2013;347:f5577.
AC CE P
55.
15
ACCEPTED MANUSCRIPT Table 1. Demographic, health behavior, and comorbid condition characteristics associated with strength training frequency.a All
Strength Training Frequency
Characteristic 2+ times/week
< 2 times/week
P-value
74.2
73.0 (72.7, 73.2)
74.3 (74.2, 74.4)
<0.001
42.4 57.6
52.3 47.7
41.3 58.7
<0.001
88.7 8.2 3.1
91 5.2 3.8
88.5 8.6 3
<0.001
4 96
5.8 94.2
<0.001
18 26.8 25.5 29.8
34.9 33.7 17.9 13.5
<0.001
56.1 32.7 6.5 4.7
63.4 25.7 6.9 4
55.3 33.4 6.5 4.8
<0.001
42.4 37 20.6
46.9 39.7 13.4
41.9 36.7 21.4
14.7 12.1 18 55.2
35.8 25.9 22.2 16
12.5 10.7 17.5 59.3
19.4 77.6 3
23.6 72.6 3.8
18.9 78.1 2.9
PT
(N=29,975)b
Marital Status
33.3 33.1 18.6 15.1
Married Widowed Divorced Other Group 2: Health Behaviors Body Mass Index <25 (Normal) 25-30 (Overweight) 30+ (Obese) Physical Activitye Highly Active Sufficiently Active Insufficiently Active Inactive Alcohol use Never Moderate regular Unhealthy regular
SC
NU MA D
5.6 94.4
AC CE P
Gender Male Female Race Caucasian African American Other Ethnicity Hispanic Non-Hispanic Educational Status Less than HSc HS or GEDd Some college College degree
TE
Age (5 year increase)
RI
Group 1: Demographics
<0.001
<0.001
<0.001 16
ACCEPTED MANUSCRIPT Table 1. Demographic, health behavior, and comorbid condition characteristics associated with strength training frequency (continued).
AC CE P
TE
D
MA
NU
SC
RI
PT
Smoking status Current 10.7 6.1 11.2 <0.001 Former 39.6 48.3 38.7 Never 49.6 45.6 50 Group 3: Comorbid Conditions Diabetes Yes 14.1 9.9 14.5 <0.001 No 85.9 90.1 85.5 Hypertension Yes 47.7 41.2 48.4 <0.001 No 52.3 58.8 51.6 Coronary heart disease Yes 12.7 13.3 12.6 0.378 No 87.3 86.7 87.4 Non-skin cancer Yes 15.9 16.8 15.8 0.238 No 84.1 83.2 84.2 a Numbers reported are percentages. b There are only 29,975 subjects out of 30,162 subjects overall who have non-missing data for the strength training variable from the 1997-2001 National Health Interview Survey. c HS = high school. d GED = general educational development. e 2008 Physical Activity Guidelines for Americans: highly active (≥300 min per week of light- to moderateintensity aerobic activity, ≥150 min per week of vigorous-intensity aerobic activity or in combination); sufficiently active (150–300 min per week of light- to moderate-intensity aerobic activity, 75–150 min per week of vigorous-intensity aerobic activity or in combination); insufficiently active (some activity but not enough to meet the more active definitions); and inactive (no light-to moderate- or vigorousintensity activity of ≥10 min per session).
17
ACCEPTED MANUSCRIPT Table 2. Strength training and characteristics associated with mortality in US older adults (N=30,162).a
27,250 (90.4) 2,725 (9.6)
Reference 0.55 (0.51, 0.61)
Reference 0.81 (0.69, 0.96)
74.2 (74.1, 74.3)
1.80 (1.75, 1.85)
11,412 (42.4) 18,750 (57.6) 25,842 (88.7) 3,396 (8.2) 924 (3.1)
RI
PT
All Cause OR (95% CI)
All Cardiac OR (95% CI) Reference 0.59 (0.51, 0.69)
1.86 (1.80, 1.93)
1.36 (1.28, 1.45) Reference
1.72 (1.57, 1.90) Reference
1.17 (1.09, 1.27) Reference
Reference 1.17 (1.08, 1.28) 0.57 (0.47, 0.69)
Reference 1.11 (0.94, 1.30) 0.78 (0.55, 1.08)
Reference 1.22 (1.08, 1.36) 0.61 (0.45, 0.81)
0.73 (0.65, 0.80) Reference
0.78 (0.64, 0.95) Reference
0.87 (0.74, 1.02) Reference
Reference 0.63 (0.59, 0.68) 0.60 (0.55, 0.66) 0.51 (0.46, 0.56)
Reference 0.88 (0.78, 0.99) 0.82 (0.71, 0.94) 0.77 (0.65, 0.90)
Reference 0.66 (0.60, 0.72) 0.63 (0.56, 0.70) 0.55 (0.48, 0.63)
12,410 (56.1) 12,949 (32.7) 2,794 (6.5) 1,962 (4.7)
Reference 1.75 (1.65, 1.86) 1.21 (1.09, 1.34) 1.40 (1.26, 1.58)
Reference 0.97 (0.87, 1.07) 0.99 (0.84, 1.16) 0.92 (0.74, 1.15)
Reference 1.82 (1.68, 1.97) 1.14 (0.98, 1.32) 1.55 (1.35, 1.79)
12,883 (42.3) 10,880 (36.8) 6,399 (20.9)
1.46 (1.35, 1.57) 0.93 (0.87, 1.01) Reference
1.35 (1.19, 1.55) 1.15 (0.99, 1.34) Reference
1.34 (1.20, 1.48) 0.92 (0.84, 1.02) Reference
3,541 (14.7) 3,032 (12.1) 4,653 (18.0) 14,952 (55.2)
0.32 (0.28, 0.35) 0.39 (0.35, 0.44) 0.59 (0.54, 0.64) Reference
0.73 (0.61, 0.87) 0.76 (0.64, 0.91) 0.80 (0.69, 0.94) Reference
0.33 (0.28, 0.39) 0.40 (0.34, 0.47) 0.66 (0.58, 0.74) Reference
5,717 (19.3) 23,579 (77.7) 866 (3.0)
Reference 1.46 (1.36, 1.58) 1.35 (1.13, 1.62)
Reference 1.08 (0.95, 1.23) 1.70 (1.32, 2.18)
Reference 1.51 (1.36, 1.67) 0.91 (0.70, 1.20)
3,436 (10.7) 11,236 (39.6)
1.79 (1.63, 1.95) 1.39 (1.31, 1.48)
2.94 (2.56, 3.42) 1.86 (1.67, 2.08)
1.05 (0.92, 1.19) 1.01 (0.93, 1.11)
NU
MA
D
2,631 (5.6) 27,522 (94.4)
SC
1.12 (1.08, 1.16)
TE
10,890 (33.3) 9,436 (33.0) 5,466 (18.6) 4,063 (15.0)
AC CE P
Characteristic Strength Trainingb < 2 times/week ≥ 2 times/week Group 1: Demographics Age (5-year increase) Gender Male Female Race Caucasian African American Other Ethnicity Hispanic Non-Hispanic Educational Status Less than HSc HS or GEDd Some College College Degree Marital Status Married Widowed Divorced Other Group 2: Health Behaviors Body Mass Index <25 (Normal) 25-30 (Overweight) 30+ (Obese) Physical Activitye Highly Active Sufficiently Active Insufficiently Active Inactive Alcohol Use Never Moderate Regular Unhealthy Regular Smoking Status Current Former
N (%) or Mean (95% CI)
Mortality All Cancer OR (95% CI)
18
ACCEPTED MANUSCRIPT Table 2. Strength training and characteristics associated with mortality in US older adults (N=30,162) (continued).
AC CE P
TE
D
MA
NU
SC
RI
PT
Never 15,490 (49.7) Reference Reference Reference Group 3: Comorbid Conditions Diabetes Yes 4,283 (14.1) 1.72 (1.60, 1.84) 0.92 (0.79, 1.06) 1.63 (1.49, 1.79) No 25,839 (85.9) Reference Reference Reference Hypertension Yes 14,585 (47.7) 1.31 (1.23, 1.38) 0.95 (0.87, 1.05) 1.48 (1.38, 1.60) No 15,577 (52.3) Reference Reference Reference Coronary Heart Disease Yes 3,577 (12.7) 1.92 (1.77, 2.08) 1.13 (0.98, 1.31) 2.32 (2.10, 2.53) No 26,434 (87.3) Reference Reference Reference Non-Skin Cancer Yes 4,586 (16.0) 1.60 (1.49, 1.70) 3.10 (2.80, 3.46) 1.01 (0.91, 1.12) No 25,496 (84.0) Reference Reference Reference a All odds ratios (OR) and confidence intervals (CI) are from a logistic regression analysis which was weighted based on the complex sampling stratification and clustering of the 1997-2001 National Health Interview Survey data. All percentages, means and confidence limits are also weighted based on the complex sampling stratification and clustering of the NHIS data. b Responses were categorized to signify whether the individual performed strength training activities at least twice each week, consistent with the ACSM/AHA Guidelines. c HS = high school. d GED = general educational development. e 2008 Physical Activity Guidelines for Americans: highly active (≥300 min per week of light- to moderateintensity aerobic activity, ≥150 min per week of vigorous-intensity aerobic activity or in combination); sufficiently active (150–300 min per week of light- to moderate-intensity aerobic activity, 75–150 min per week of vigorous-intensity aerobic activity or in combination); insufficiently active (some activity but not enough to meet the more active definitions); and inactive (no light-to moderate- or vigorousintensity activity of ≥10 min per session).
19
ACCEPTED MANUSCRIPT Table 3: Strength training associated with mortality adjusted for demographics, health behaviors, and comorbid conditions in US older adults.a
All-Cause
Total %
ST 2+ times/ week %
ST <2 times/ week %
31.6
21.3
32.7
Group 1 (N=29,637) OR (95% CI)
Groups 1 & 2 (N=25,848) OR (95% CI)
PT
Mortality
Groups 1, 2, & 3 (N=25,663) OR (95% CI)
AC CE P
TE
D
MA
NU
SC
RI
0.64 0.82 0.81 (0.57, 0.70) (0.73, 0.92) (0.71, 0.92) All Cancer 13.4 8.8 13.9 0.83 0.87 0.84 (0.70, 0.98) (0.70, 1.06) (0.68, 1.03) All Cardiac 7.2 6.1 7.4 0.72 0.88 0.88 (0.62, 0.83) (0.74,1.04) (0.73,1.04) a All percentages are weighted based on the complex sampling stratification and clustering of the 1997-2001 National Health Interview Survey data. All odds ratios (OR) and confidence intervals (CI) are from a logistic regression analysis which was weighted based on the complex sampling stratification and clustering of the NHIS data and was adjusted for the covariates corresponding the group as listed in Table 1.
20
ACCEPTED MANUSCRIPT Table 4: Strength training associated with mortality adjusted or matched via a propensity score based on demographics, health behaviors, and comorbid conditions in US older adults.a
AC CE P
TE
D
MA
NU
SC
RI
PT
Adjusted for Propensity Scoreb Matched by Propensity Scorec Group 1 Groups 1 & 2 Groups 1, 2, & 3 Group 1 Groups 1 & 2 Groups 1, 2, & 3 Mortality (N=29,637) (N=25,848) (N=25,663) (N=5,408) (N=4,622) (N=4,600) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) All-Cause 0.63 0.82 0.82 0.69 0.90 0.84 (0.57, 0.70) (0.73, 0.91) (0.73, 0.92) (0.60, 0.79) (0.78, 1.04) (0.72, 0.96) All Cancer 0.81 0.84 0.83 0.85 1.03 0.84 (0.68, 0.96) (0.69, 1.03) (0.67, 1.01) (0.68, 1.06) (0.78, 1.36) (0.66, 1.08) All Cardiac 0.69 0.88 0.89 0.76 0.94 0.84 (0.59, 0.80) (0.74,1.05) (0.75,1.06) (0.63, 0.93) (0.75, 1.19) (0.66, 1.04) a All odds ratios (OR) and confidence intervals (CI) for strength training are from a logistic regression analysis which was weighted based on the complex sampling stratification and clustering of the 1997-2001 National Health Interview Survey data and was either adjusted for the propensity score derived from the covariates corresponding the groups as listed in Table 1 or matched strength training groups 1-to-1 based on the same propensity score. b All data from the sample adjusted for propensity score without any covariates included since propensity score was generated from those covariates. c Reduced sample matched by propensity score without any covariates included.
21
ACCEPTED MANUSCRIPT Is strength training associated with mortality benefits? A 15 year cohort study of US older adults. Jennifer L. Kraschnewski, M.D., MPH1,2,3, Christopher N. Sciamanna, M.D.,1,2,3, MPH, Jennifer M. Poger,
PT
M.Ed.1, Liza S. Rovniak, Ph.D., MPH1,3, Erik B. Lehman, MS3, Amanda B. Cooper, M.D.2, Noel H. Ballentine, M.D., F.A.C.P.2, & Joseph T. Ciccolo, Ph.D.4
Department of Medicine, Penn State College of Medicine, 2Department of Medicine, Penn State
RI
1
4
Department of Applied Physiology, Columbia University
TE
D
MA
Guideline-concordant strength training reduces all cause, cancer, and cardiac death. Only a minority of older adults meet current strength training guidelines. All-cause mortality may be reduced through strength training interventions.
AC CE P
NU
Highlights
SC
Hershey Medical Center, 3Department of Public Health Sciences, Penn State College of Medicine,
22
S0091-7435(16)30016-0 doi: 10.1016/j.ypmed.2016.02.038 YPMED 4561
To appear in:
Preventive Medicine
Received date: Revised date: Accepted date:
6 November 2015 10 February 2016 23 February 2016
Please cite this article as: Kraschnewski Jennifer L., Sciamanna Christopher N., Poger Jennifer M., Rovniak Liza S., Lehman Erik B., Cooper Amanda B., Ballentine Noel H., Ciccolo Joseph T., Is strength training associated with mortality benefits? A 15 year cohort study of US older adults, Preventive Medicine (2016), doi: 10.1016/j.ypmed.2016.02.038
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Title Page Is strength training associated with mortality benefits? A 15 year cohort study of US older adults.
PT
Jennifer L. Kraschnewski, M.D., MPH1,2,3, Christopher N. Sciamanna, M.D.,1,2,3, MPH, Jennifer M. Poger, M.Ed.1, Liza S. Rovniak, Ph.D., MPH1,3, Erik B. Lehman, MS3, Amanda B. Cooper, M.D.2, Noel H. Ballentine,
RI
M.D., F.A.C.P.2, & Joseph T. Ciccolo, Ph.D.4 1
Department of Medicine, Penn State College of Medicine, 2Department of Medicine, Penn State
4
NU
Department of Applied Physiology, Columbia University
SC
Hershey Medical Center, 3Department of Public Health Sciences, Penn State College of Medicine,
Corresponding Author:
MA
Jennifer L. Kraschnewski, MD, MPH
Assistant Professor of Medicine and Public Health Sciences
D
Associate Director, MD-MPH Program
Department of Medicine
TE
Medical Director of Research, Penn State Hershey PRO Wellness Center
Division of General Internal Medicine
AC CE P
Penn State Hershey Medical Center 500 University Drive, HO34 Hershey, PA 17033 (717) 531-8161
FAX: (717) 531-7726
[email protected]
Abstract Word Count: 246 Article Word Count: 2,950 Number of References: 55 Number of Tables: 4 Number of Figures: 0
1
ACCEPTED MANUSCRIPT Abstract Background:
PT
The relationship between strength training (ST) behavior and mortality remains understudied in large, national samples, although smaller studies have observed that greater amounts of muscle strength are
SC
guidelines and future mortality in an older US adult population.
RI
associated with lower risks of death. We aimed to understand the association between meeting ST
Methods:
NU
Data were analyzed from the 1997-2001 National Health Interview Survey (NHIS) linked to death certificate data in the National Death Index. The main independent variable was guideline-concordant
MA
ST (i.e. twice each week) and dependent variable was all-cause mortality. Covariates identified in the literature and included in our analysis were demographics, past medical history, and other health behaviors (including other physical activity). Given our aim to understand outcomes in older adults,
AC CE P
Results:
TE
multiple logistic regression analysis.
D
analyses were limited to adults age 65 years and older. Multivariate analysis was conducted using
During the study period, 9.6% of NHIS adults age 65 and older (N=30,162) reported doing guidelineconcordant ST and 31.6% died. Older adults who reported guideline-concordant ST had 46% lower odds of all-cause mortality than those who did not (adjusted odds ratio: 0.64; 95% CI: 0.57, 0.70; p<0.001). The association between ST and death remained after adjustment for past medical history and health behaviors.
Conclusions: Although a minority of older US adults met ST recommendations, guideline-concordant ST is significantly associated with decreased overall mortality. All-cause mortality may be significantly reduced through the identification of and engagement in guideline-concordant ST interventions by older adults.
Keywords: strength training; mortality; resistance training; older adults
2
ACCEPTED MANUSCRIPT Introduction
Engaging in regular physical activity reaps a multitude of health benefits for adults of all ages; it is
PT
perhaps of greatest value in preventing premature mortality associated with all-cause, cardiovascular disease, diabetes, and even some cancers.1-11 Significant quality of life improvements and decreased risk
RI
of mortality have also been observed in physically active, older adults.4,9,12-16 Unlike aerobic exercise’s well-established effects, strength training has only recently garnered attention for its benefits in
SC
regaining muscle mass and strength often depleted with age and disability.17-20 While guidelineconcordant physical activity has consistent and powerful relationships with longer life expectancy,
NU
meeting strength training guidelines may also play an important role in decreasing premature mortality
MA
and warrants further investigation.
Although physical activity guidelines have been widely available for decades from the American College
D
of Sports Medicine (ACSM),21 specific recommendations regarding strength training have only been developed more recently; in 2007, the American Heart Association (AHA) together with the ACSM
TE
established recommendations encouraging all adults to participate in strength training activities at least twice each week,19 with a similar emphasis on muscle-strengthening activity for older adults.22 In
AC CE P
addition, the Behavioral Risk Factor Surveillance System (BRFSS), which began assessing aerobic activity with the first survey in 1984, only began assessing strength training in 2011.23 Over this time, studies have demonstrated robust effects of strength training on strength, muscle mass and physical function as well as improvement in a range of chronic conditions, including diabetes, osteoporosis, low back pain, and obesity.7,18,24-26 Recently, Krist and colleagues observed considerable improvements in muscle strength and mobility among elderly nursing home residents with impaired mobility at baseline after they engaged in a resistance exercise program twice weekly over a two month period.17 Similarly, Mayer and colleagues found that progressive strength training in the elderly helped retain motor function and reduce sarcopenia.20 Older adults with functional limitations can achieve significant improvements in physical function and muscle strength by engaging in strength training activity.27
In addition to improvements in physical function, several clinical studies have demonstrated that higher levels of muscular strength are associated with lower all-cause, cardiovascular, and cancer mortality risk, among both men and the elderly.28-30 Findings from 15 epidemiological and 8 clinical studies summarized in the 2015 systematic review by Volaklis et al. provides evidence that muscular strength is 3
ACCEPTED MANUSCRIPT inversely and independently associated with all-cause and cardiovascular mortality.31 Although there is increasing evidence illuminating the substantive benefits of muscle strength on decreased mortality risk,31 no studies to our knowledge describe the relationship between strength training behavior and
PT
mortality in a large, nationally representative sample over an extended time period, particularly in older
RI
adults.
SC
The aim of this investigation was to understand the association between meeting strength training guidelines and future mortality in older adults. We hypothesized that guideline-concordant strength
NU
training is significantly associated with decreased overall mortality in older US adults.
MA
Methods
In the present cohort study, longitudinal data from the 1997-2001 National Health Interview Survey (NHIS), linked to death certificate data from the National Center for Health Statistics National Death
D
Index,32 were analyzed. Survey sampling methods and data collection procedures have been described
TE
extensively elsewhere.33 Briefly, the NHIS employs a multi-stage stratified sampling method to collect overall health, disease, and disability data of the US population from a nationally representative
AC CE P
sampling of all 50 states and the District of Columbia .33 Personal household interviews conducted by Census interviewers collect basic demographic and family relationship information about all household members. Additionally, one randomly selected adult in each household is interviewed in detail regarding their health and health-related behaviors through the Sample Adult questionnaire, which was analyzed in this study. Our analysis was limited to those individuals age 65 years and older, given our interest in assessing strength training activity in the Medicare-eligible population. Three years of the study were used (1997, 1998, 1999) to provide enough data, including deaths (through 2011), to allow for significant relationships to emerge after adjusting for potential confounders. This study was exempt from Institutional Review Board approval.
The main independent variable, guideline-concordant strength training, was assessed using the following question: “How often do you do leisure-time physical activities specifically designed to strengthen your muscles, such as lifting weight or doing calisthenics?” Participant responses included both the number of times strength training was performed and the unit of time (i.e.“per week,” “per month”). These responses were categorized to signify whether the individual performed these activities 4
ACCEPTED MANUSCRIPT at least twice each week, consistent with the ACSM/AHA Guidelines.34 Additionally, in order to assess how often moderate and vigorous physical activity was performed, participants were asked the following questions: How often do you do light or moderate activities for at least 10 minutes that cause
PT
only light sweating or a slight to moderate increase in breathing or heart rate?”; “About how long do you do these light or moderate activities each time?”; “How often do you do vigorous activities for at
RI
least 10 minutes that cause heavy sweating or large increases in breathing or heart rate?”; “About how
SC
long do you do these vigorous activities each time?” Physical activity was categorized according to the 2008 Physical Activity Guidelines for Americans: highly active (≥300 min per week of light- to moderateintensity aerobic activity, ≥150 min per week of vigorous-intensity aerobic activity or in combination);
NU
sufficiently active (150–300 min per week of light- to moderate-intensity aerobic activity, 75–150 min per week of vigorous-intensity aerobic activity or in combination); insufficiently active (some activity but
MA
not enough to meet the more active definitions); and inactive (no light-to moderate- or vigorousintensity activity of ≥10 min per session).35-37 Minutes of moderate and vigorous intensity activity were
D
re-categorized, based on whether the individual met the ACSM/AHA Guidelines for moderate intensity aerobic activity (≥ 150 minutes per week) and vigorous intensity aerobic activity (≥ 75 minutes per
TE
week). The main outcome variable was mortality (all-cause, cancer-related, and cardiac-related) as identified from the National Death Index.32 The cause of death classification was based on the
AC CE P
International Classification of Disease codes (ICD-10). The following codes were included to define each cause: Cancer (19-44); Cardiac (53-75).
The following demographic variables were included as covariates in the analysis: age, gender, race (Caucasian, African-American, Other), ethnicity (Hispanic vs. non-Hispanic), education level (less than high school, high school or GED, some college/ associate's degree, bachelor degree or higher), and marital status (married/living with partner, widowed, divorced, other). Additionally, we included health behaviors (body mass index (BMI), smoking, alcohol intake, physical activity) and comorbid conditions (hypertension, diabetes, heart disease, non-skin cancer), consistent with variables shown to have associations with the outcome in other studies.35,38 Based on cut-points from the National Institutes of Health, BMI was re-categorized into normal, overweight and obese.39
A weighted analysis employing procedures in SAS 9.4 (SAS Institute, Cary, NC) which take into consideration the complex sampling stratification and clustering incorporated by the NHIS study was used for all analyses. Using weighted logistic regression, bivariate comparisons were made between all5
ACCEPTED MANUSCRIPT cause mortality, cancer-related mortality, or cardiac-related mortality and strength training as well as demographic variables, health behaviors, and comorbid conditions. Weighted multivariable logistic regression models were then utilized to further investigate the association between strength training
PT
and the types of mortality while adjusting for the three groups of covariates: demographic variables only, both demographic variables and health behaviors, and finally demographic variables, health
RI
behaviors, and comorbid conditions. Odds ratios were calculated for all logistic regression analyses to
SC
quantify the direction and magnitude of the associations.
A confirmatory analysis was then performed using propensity scores.40,41 Three sets of propensity
NU
scores, one set for each group of covariates from the primary analysis, were calculated based on the predicted probabilities from a weighted logistic regression incorporating strength training as the
MA
outcome and the covariates as the predictors. The three sets of propensity scores where then applied in two ways. First, they were added as the only covariate for adjustment as a substitute for the set of
D
predictors used in their construction in a weighted multivariable logistic regression model with strength training as the predictor and all-cause mortality, cancer-related mortality, or cardiac-related mortality as
TE
the outcomes. Second, they were used in a 1-to-1 method of matching subjects with strength training <2 times/week to subjects with strength training ≥2 times/week. A weighted logistic regression was
AC CE P
then performed using this matched sample including strength training as the only predictor in the model, given that the sample was already adjusted for the covariates through the matching process.
Results
Overall, 9.6% of NHIS adults 65 years of age and older (N=30,162) reported doing strength training at least twice weekly, consistent with guidelines from the ASCM/AHA,19 and 31.6% of the study population died during the 15 year follow-up period. Table 1 displays demographic, health behavior, and comorbid condition characteristics associated with strength training frequency. Older adults who met guidelines for strength training were, on average, slightly younger and were more likely to be married white males with higher levels of education. Individuals who had a normal body weight, engaged in aerobic exercise, and abstained from alcohol and tobacco use were also more likely to meet strength training guidelines. Finally, older adults who had guideline-concordant strength training were less likely to have diabetes and hypertension, but did not differ in terms of coronary heart disease or cancer status from their lessactive counterparts. 6
ACCEPTED MANUSCRIPT
The bivariate relationships between mortality and demographic (Group 1), health behavior (Group 2) and comorbid condition (Group 3) variables are presented in Table 2. The bivariate relationships
PT
between meeting strength training guidelines and the other covariates were also examined. Older adults who performed strength training at least twice a week had 45% lower odds of all-cause mortality, 19%
SC
RI
lower odds of death from cancer, and 41% lower odds of cardiac death.
As shown in Table 3, after adjusting for demographic covariates (Group 1), those who reported guideline-concordant strength training had 36% lower odds of all-cause mortality than those who did
NU
not (adjusted odds ratio: 0.64; 95% CI: 0.57, 0.70; p<0.001). There remained a statistically significant decrease in odds of dying from cancer and cardiac etiologies as well; 17% (adjusted odds ratio: 0.83;
MA
95% CI: 0.70, 0.98; p<0.05) and 28% (adjusted odds ratio: 0.72; 95% CI: 0.62, 0.83; p<0.001) respectively. After adjusting for demographic variables, health behaviors, and comorbid conditions (Groups 1, 2, & 3),
D
all-cause mortality remained significant, with a 19% decreased odds of dying (see Table 2). However, the point estimate of the relationship between meeting strength training guidelines and cancer and cardiac
AC CE P
remained in the same direction.
TE
mortality after controlling for these covariates was no longer statistically significant, although it
Analyses adjusted for the propensity scores in Table 4 show similar results. There was 37% lower odds of all-cause mortality after adjusting for demographic covariates (Group 1) (adjusted odds ratio: 0.63; 95% CI: 0.57, 0.70; p<0.001). Similarly, there remained a statistically significant decrease in all-cancer and allcardiac deaths, although cancer odds were not as strongly associated, as also evident in Table 3 analyses; 19% (adjusted odds ratio: 0.81; 95% CI: 0.68, 0.96; p<0.05) and 31% (adjusted odds ratio: 0.69; 95% CI: 0.59, 0.80; p<0.001) respectively. A similar pattern remained when adjusting for demographic, health behavior, and comorbid condition variables (Groups 1, 2, & 3) with the adjusted propensity scores; there was 18% lower odds of all-cause mortality (adjusted odds ratio: 0.82; 95% CI: 0.73, 0.92; p<0.001), while all-cancer and all-cardiac mortalities showed a similar trend but were no longer statistically significant. When matched by propensity scores, there was 31% lower odds of all-cause mortality and 24% lower odds of cardiac death when adjusting for demographic variables (Group 1). Additionally, there was a statistically significant decrease in all-cause mortality when controlling for demographic, health behavior, and comorbid condition variables (Groups 1, 2, & 3) using propensity
7
ACCEPTED MANUSCRIPT score matching (adjusted odds ratio: 0.84; 95% CI: 0.72, 0.96; p<0.05). All-cancer deaths were no longer statistically significant when controlling for any of the covariates in the matching analysis.
PT
Discussion
RI
The purpose of the present study was to understand the relationship between meeting strength training guidelines and future mortality. Overall, the results showed a consistent and inverse association in a
SC
large, national sample;42 those who performed guideline-concordant strength training had significantly lower odds of all cause, cancer, and cardiac death than their less-active counterparts. This association
NU
remained significant for all-cause mortality after adjusting for each of the covariate groups (demographic, health behavior, and comorbid condition variables, respectively). In addition, these
MA
patterns remained significant, albeit less strongly, after adjusting specifically for demographic variables in cancer and cardiac mortalities. Therefore, older adults who perform strength training activities not
D
only improve their physical functioning as previously demonstrated, 17,20,43 but their survival rate as well.
TE
Although not statistically significant in cancer and cardiac deaths, point estimates suggest that guidelineconcordant strength training behaviors are associated with decreased risk of cause-specific mortality.
AC CE P
This weaker association may be the result of sample size, as a lesser percentage of individuals died from cancer (N=2,192; 7.2%) or cardiac (N=4,275; 13.4%) etiologies than from all-causes.
Results from the propensity score analyses confirmed our findings from the weighted logistic regression models. After matching subjects who strength trained <2 times/week to subjects who strength trained ≥2 times/week in the propensity score matching, there remained a significantly lower odds of all-cause mortality when controlling for all variables (Groups 1, 2, & 3). Although sample size was greatly reduced in the propensity score matching, as those who met strength training guidelines was a much smaller group than those who were not guideline concordant (N=5,408), this reduction in power is balanced by the efficiency in matching.44
Although few studies have examined the impact of strength training on future mortality, existing outcomes are similar to the results of the present study.45,46 For example, Singh and colleagues performed a randomized controlled trial involving 12 months of high-intensity weight-lifting exercise as part of a multidisciplinary intervention for patients recovering from hip fractures and found an 81% 8
ACCEPTED MANUSCRIPT reduction in mortality (N=4) compared to the usual care control condition (N=8) (adjusted odds ratio: 0.19; 95% CI; 0.04, 0.91; p < .04).45 Similarly, Hardee and colleagues conducted a prospective epidemiological investigation of self-reported physical activity and resistance training among cancer
PT
survivors and observed a 33% lower risk of all-cause mortality among those who participated in resistance training ≥ 1 day per week (N=39) (95% CI, 0.45-0.99) after adjusting for potential
RI
confounders, including physical activity. No correlation was found with those who engaged in physical activity only.46 While these studies are limited by sample size, the present study benefits from using a
SC
large, nationally representative sample of a range of health outcomes for the US population.42
NU
Consistent with prior findings, meeting strength training guidelines was strongly associated with younger age, among other demographic and health behavior variables, including meeting physical
MA
activity guidelines and not smoking.34,35,47 Ciccolo and colleagues (2010) in a separate survey of 9,651 US adults observed that only 21.7% of older adults met strength training guidelines versus 37.5% of adults
D
aged 35–54.38 This was consistent with our small percentage of older adults performing guidelineconcordant strength training activities (N=2,725; 9.6%). After controlling for physical activity level, which
TE
was strongly associated with reduced mortality, specifically reporting strength exercises appeared to
AC CE P
confer additional benefit beyond reporting physical activity alone.
This study has several important limitations to consider. First, the study design is that of a cohort study, limiting ability to determine cause and effect. In addition, participation in strength training was selfreported without information regarding session quality or length. Self-report measures remain widely used for physical activity assessment, however, typically overestimate activity level when compared with objective measurement.48 Unmeasured confounders, including diet and self-report of health status could have obscured the data. Unfortunately, dietary habits and objective physical and biological data are difficult to assess and beyond the scope of this secondary data analysis. Although not ideal, selfreport remains a reliable and valid indicator of health.49 Another limitation is that respondents may have misinterpreted the survey questions or may not have fully understood the definitions of strength training and calisthenics, which may have weakened the accuracy of strength training prevelance.34 Additionally, given that calisthenics isn’t always categorized as progressive training, the survey question used from the NHIS combining strength training with calisthenics may have undesirably included some aerobic activity as well. Since respondents were not asked to provide details regarding the type of strength training activities performed, some activities such as stair climbing could have been missed.34 9
ACCEPTED MANUSCRIPT Although it is possible that a small percentage of older adults were physically unable to perform strength training activities, potentially confounding the data, the literature suggests otherwise; Drey and colleagues (2012), for example, randomized older adults (>65 years) into either a strength training (ST),
PT
power training (PT) or control group and found the highest fidelity in the ST group, with 87% attendance rates over 12 weeks.50 A final limitation is this study relied on death certificate data, which has been
RI
previously questioned for its reliability and accuracy in properly identifying cause of death.51,52 However, when used for observational studies the National Death Index has been shown to be accurate in
SC
ascertaining deaths, with high sensitivity (96%) and specificity (100%).53
NU
Conclusion
MA
Despite guidelines, only 21.7% of older adults currently meet recommendations of strength training at least twice each week, substantially lower than the 2010 national objective of 30%.54 This underscores the need for additional programs to increase strength training among older adults to help engage
D
patients in the “Exercise is Medicine” campaign.34 This study further demonstrates the importance of
TE
encouraging doctors to recommend physical activity to patients as they would any other effective treatment, particularly since exercise has been shown to be as effective as some medications.55
AC CE P
Identifying interventions to successfully engage older adults in guideline-concordant strength training has the potential to significantly reduce all-cause mortality in this population.
Conflict of Interest Statement
The authors declare that there are no conflicts of interest.
10
ACCEPTED MANUSCRIPT References 1.
Moore SC, Patel AV, Matthews CE, et al. Leisure time physical activity of moderate to vigorous intensity and mortality: a large pooled cohort analysis. PLoS Med. 2012;9(11):e1001335. Autenrieth CS, Baumert J, Baumeister SE, et al. Association between domains of physical activity
PT
2.
and all-cause, cardiovascular and cancer mortality. Eur J Epidemiol. 2011;26(2):91-99. Samitz G, Egger M, Zwahlen M. Domains of physical activity and all-cause mortality: systematic
RI
3.
SC
review and dose-response meta-analysis of cohort studies. Int J Epidemiol. 2011;40(5):13821400. 4.
Wen CP, Wai JP, Tsai MK, et al. Minimum amount of physical activity for reduced mortality and
5.
NU
extended life expectancy: a prospective cohort study. Lancet. 2011;378(9798):1244-1253. Woodcock J, Franco OH, Orsini N, Roberts I. Non-vigorous physical activity and all-cause
MA
mortality: systematic review and meta-analysis of cohort studies. Int J Epidemiol. 2011;40(1):121-138.
Lollgen H, Bockenhoff A, Knapp G. Physical activity and all-cause mortality: an updated meta-
D
6.
analysis with different intensity categories. Int J Sports Med. 2009;30(3):213-224. Davidson LE, Hudson R, Kilpatrick K, et al. Effects of exercise modality on insulin resistance and
TE
7.
functional limitation in older adults: a randomized controlled trial. Arch Intern Med.
8.
AC CE P
2009;169(2):122-131.
Nocon M, Hiemann T, Muller-Riemenschneider F, Thalau F, Roll S, Willich SN. Association of physical activity with all-cause and cardiovascular mortality: a systematic review and metaanalysis. Eur J Cardiovasc Prev Rehabil. 2008;15(3):239-246.
9.
Wannamethee SG, Shaper AG. Physical activity in the prevention of cardiovascular disease: an epidemiological perspective. Sports Med. 2001;31(2):101-114.
10.
Davey Smith G, Shipley MJ, Batty GD, Morris JN, Marmot M. Physical activity and cause-specific mortality in the Whitehall study. Public Health. 2000;114(5):308-315.
11.
Haapanen N, Miilunpalo S, Vuori I, Oja P, Pasanen M. Characteristics of leisure time physical activity associated with decreased risk of premature all-cause and cardiovascular disease mortality in middle-aged men. Am J Epidemiol. 1996;143(9):870-880.
12.
Brown WJ, McLaughlin D, Leung J, et al. Physical activity and all-cause mortality in older women and men. Br J Sports Med. 2012;46(9):664-668.
13.
Ueshima K, Ishikawa-Takata K, Yorifuji T, et al. Physical activity and mortality risk in the Japanese elderly: a cohort study. Am J Prev Med. 2010;38(4):410-418. 11
ACCEPTED MANUSCRIPT 14.
Leitzmann MF, Park Y, Blair A, et al. Physical activity recommendations and decreased risk of mortality. Arch Intern Med. 2007;167(22):2453-2460.
15.
Sundquist K, Qvist J, Sundquist J, Johansson SE. Frequent and occasional physical activity in the
16.
PT
elderly: a 12-year follow-up study of mortality. Am J Prev Med. 2004;27(1):22-27. Wannamethee SG, Shaper AG, Walker M. Physical activity and mortality in older men with
Krist L, Dimeo F, Keil T. Can progressive resistance training twice a week improve mobility,
SC
17.
RI
diagnosed coronary heart disease. Circulation. 2000;102(12):1358-1363.
muscle strength, and quality of life in very elderly nursing-home residents with impaired mobility? A pilot study. Clin Interv Aging. 2013;8:443-448.
Gennuso KP, Zalewski K, Cashin SE, Strath SJ. Resistance training congruent with minimal
NU
18.
guidelines improves function in older adults: a pilot study. J Phys Act Health. 2013;10(6):769-
19.
MA
776.
Haskell WL, Lee IM, Pate RR, et al. Physical activity and public health: updated recommendation
D
for adults from the American College of Sports Medicine and the American Heart Association. Circulation. 2007;116(9):1081-1093.
Mayer F, Scharhag-Rosenberger F, Carlsohn A, Cassel M, Muller S, Scharhag J. The intensity and
TE
20.
effects of strength training in the elderly. Dtsch Arztebl Int. 2011;108(21):359-364. American College of Sports Medicine position statement on the recommended quantity and
AC CE P
21.
quality of exercise for developing and maintaining fitness in healthy adults. Med Sci Sports. 1978;10(3):vii-x. 22.
Nelson ME, Rejeski WJ, Blair SN, et al. Physical activity and public health in older adults: recommendation from the American College of Sports Medicine and the American Heart Association. Circulation. 2007;116(9):1094-1105.
23.
Centers for Disease Control and Prevention. Adult Participation in Aerobic and MuscleStrengthening Physical Activities — United States, 2011. MMWR. 2013;62(17):326-330.
24.
Nelson ME, Fiatarone MA, Morganti CM, Trice I, Greenberg RA, Evans WJ. Effects of highintensity strength training on multiple risk factors for osteoporotic fractures. A randomized controlled trial. JAMA. 1994;272(24):1909-1914.
25.
Fiatarone MA, O'Neill EF, Ryan ND, et al. Exercise training and nutritional supplementation for physical frailty in very elderly people. N Engl J Med. 1994;330(25):1769-1775.
12
ACCEPTED MANUSCRIPT 26.
Pollock MV, KR. Resistance training for health. The President's Council on Physical Fitness and Sports. https://www.presidentschallenge.org/informed/digest/docs/199612digest.pdf. Accessed August 27, 2015. Kraschnewski JL, Sciamanna CN, Ciccolo JT, et al. Is exercise used as medicine? Association of
PT
27.
meeting strength training guidelines and functional limitations among older US adults. Prev
Ortega FB, Silventoinen K, Tynelius P, Rasmussen F. Muscular strength in male adolescents and
SC
28.
RI
Med. 2014;66:1-5.
premature death: cohort study of one million participants. BMJ. 2012;345:e7279. 29.
Ruiz JR, Sui X, Lobelo F, et al. Muscular strength and adiposity as predictors of adulthood cancer
30.
NU
mortality in men. Cancer Epidemiol Biomarkers Prev. 2009;18(5):1468-1476. Ruiz JR, Sui X, Lobelo F, et al. Association between muscular strength and mortality in men:
31.
MA
prospective cohort study. BMJ. 2008;337:a439.
Volaklis KA, Halle M, Meisinger C. Muscular strength as a strong predictor of mortality: A
32.
D
narrative review. Eur J Intern Med. 2015;26(5):303-310. Centers for Disease Control and Prevention. National Death Index.
33.
TE
http://www.cdc.gov/nchs/ndi.htm. Accessed August 27, 2015. Centers for Disease Control and Prevention and U.S. Department of Health and Human Services.
2012.
AC CE P
2011 National Health Interview Survey (NHIS) Public Use Data Release NHIS Survey Description.
ftp://ftp.cdc.gov/pub/Health_statistics/NCHs/Dataset_Documentation/NHIS/2011/srvydesc.pdf. Accessed July 6, 2015. 34.
Centers for Disease Control and Prevention. Trends in Strength Training --- United States, 1998-2004. MMWR. 2006;55(28):769-772.
35.
Carlson SA, Fulton JE, Schoenborn CA, Loustalot F. Trend and prevalence estimates based on the 2008 Physical Activity Guidelines for Americans. Am J Prev Med. 2010;39(4):305-313.
36.
Schoenborn CA, Adams PE. Health behaviors of adults: United States, 2005-2007. Vital Health Stat 10. 2010(245):1-132.
37.
U. S. Department of Health and Human Services. Physical Activity Guidelines for Americans. 2008. http://www.health.gov/paguidelines/guidelines/default.aspx. Accessed July 31, 2015.
13
ACCEPTED MANUSCRIPT 38.
Ciccolo JT, Pettee Gabriel KK, Macera C, Ainsworth BE. Association between self-reported resistance training and self-rated health in a national sample of U.S. men and women. J Phys Act Health. 2010;7(3):289-298. National Institutes of Health. Clinical Guidelines on the Identification, Evaluation, and Treatment
PT
39.
of Overweight and Obesity in Adults--The Evidence Report. Obes Res. 1998;6(2):51S-209S. Austin PC. An Introduction to Propensity Score Methods for Reducing the Effects of Confounding
RI
40.
41.
Rosenbaum P, Rubin, DB. The Central Role of the Propensity Score in Observational Studies for Causal Effects. Biometrika. 1983;70(1):41-55.
Miller D, Fernandez CA, Lee DJ. National Health Interview Survey. In: Gellman MD, Turner JR,
NU
42.
SC
in Observational Studies. Multivariate Behav Res. 2011;46(3):399-424.
eds. Encyclopedia of Behavioral Medicine. New York, NY: Springer New York; 2013:1286-1288. Kennis E, Verschueren SM, Bogaerts A, Van Roie E, Boonen S, Delecluse C. Long-term impact of
MA
43.
strength training on muscle strength characteristics in older adults. Arch Phys Med Rehabil.
44.
D
2013;94(11):2054-2060.
Polkinghorne KR, McDonald SP, Atkins RC, Kerr PG. Vascular access and all-cause mortality: a
45.
TE
propensity score analysis. J Am Soc Nephrol. 2004;15(2):477-486. Singh NA, Quine S, Clemson LM, et al. Effects of high-intensity progressive resistance training
AC CE P
and targeted multidisciplinary treatment of frailty on mortality and nursing home admissions after hip fracture: a randomized controlled trial. J Am Med Dir Assoc. 2012;13(1):24-30. 46.
Hardee JP, Porter RR, Sui X, et al. The effect of resistance exercise on all-cause mortality in cancer survivors. Mayo Clin Proc. 2014;89(8):1108-1115.
47.
Chevan J. Demographic determinants of participation in strength training activities among U.S. adults. J Strength Cond Res. 2008;22(2):553-558.
48.
Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, McDowell M. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc. 2008;40(1):181-188.
49.
Dalen JD, Huijts T, Krokstad S, Eikemo TA. Are there educational differences in the association between self-rated health and mortality in Norway? The HUNT Study. Scand J Public Health. 2012;40(7):641-647.
50.
Drey M, Zech A, Freiberger E, et al. Effects of strength training versus power training on physical performance in prefrail community-dwelling older adults. Gerontology. 2012;58(3):197-204.
14
ACCEPTED MANUSCRIPT 51.
Trepka MJ, Maddox LM, Lieb S, Niyonsenga T. Utility of the National Death Index in ascertaining mortality in acquired immunodeficiency syndrome surveillance. Am J Epidemiol. Jul 1 2011;174(1):90-98. Smith Sehdev AE, Hutchins GM. Problems with proper completion and accuracy of the cause-ofdeath statement. Arch Intern Med. 2001;161(2):277-284.
Stampfer MJ, Willett WC, Speizer FE, et al. Test of the National Death Index. Am J Epidemiol.
RI
53.
SC
1984;119(5):837-839. 54.
PT
52.
Centers for Disease Control and Prevention. Adult participation in aerobic and musclestrengthening physical activities — United States, 2011. MMWR. 2013;62(17):326-330.
NU
Naci H, Ioannidis JP. Comparative effectiveness of exercise and drug interventions on mortality
TE
D
MA
outcomes: metaepidemiological study. BMJ. 2013;347:f5577.
AC CE P
55.
15
ACCEPTED MANUSCRIPT Table 1. Demographic, health behavior, and comorbid condition characteristics associated with strength training frequency.a All
Strength Training Frequency
Characteristic 2+ times/week
< 2 times/week
P-value
74.2
73.0 (72.7, 73.2)
74.3 (74.2, 74.4)
<0.001
42.4 57.6
52.3 47.7
41.3 58.7
<0.001
88.7 8.2 3.1
91 5.2 3.8
88.5 8.6 3
<0.001
4 96
5.8 94.2
<0.001
18 26.8 25.5 29.8
34.9 33.7 17.9 13.5
<0.001
56.1 32.7 6.5 4.7
63.4 25.7 6.9 4
55.3 33.4 6.5 4.8
<0.001
42.4 37 20.6
46.9 39.7 13.4
41.9 36.7 21.4
14.7 12.1 18 55.2
35.8 25.9 22.2 16
12.5 10.7 17.5 59.3
19.4 77.6 3
23.6 72.6 3.8
18.9 78.1 2.9
PT
(N=29,975)b
Marital Status
33.3 33.1 18.6 15.1
Married Widowed Divorced Other Group 2: Health Behaviors Body Mass Index <25 (Normal) 25-30 (Overweight) 30+ (Obese) Physical Activitye Highly Active Sufficiently Active Insufficiently Active Inactive Alcohol use Never Moderate regular Unhealthy regular
SC
NU MA D
5.6 94.4
AC CE P
Gender Male Female Race Caucasian African American Other Ethnicity Hispanic Non-Hispanic Educational Status Less than HSc HS or GEDd Some college College degree
TE
Age (5 year increase)
RI
Group 1: Demographics
<0.001
<0.001
<0.001 16
ACCEPTED MANUSCRIPT Table 1. Demographic, health behavior, and comorbid condition characteristics associated with strength training frequency (continued).
AC CE P
TE
D
MA
NU
SC
RI
PT
Smoking status Current 10.7 6.1 11.2 <0.001 Former 39.6 48.3 38.7 Never 49.6 45.6 50 Group 3: Comorbid Conditions Diabetes Yes 14.1 9.9 14.5 <0.001 No 85.9 90.1 85.5 Hypertension Yes 47.7 41.2 48.4 <0.001 No 52.3 58.8 51.6 Coronary heart disease Yes 12.7 13.3 12.6 0.378 No 87.3 86.7 87.4 Non-skin cancer Yes 15.9 16.8 15.8 0.238 No 84.1 83.2 84.2 a Numbers reported are percentages. b There are only 29,975 subjects out of 30,162 subjects overall who have non-missing data for the strength training variable from the 1997-2001 National Health Interview Survey. c HS = high school. d GED = general educational development. e 2008 Physical Activity Guidelines for Americans: highly active (≥300 min per week of light- to moderateintensity aerobic activity, ≥150 min per week of vigorous-intensity aerobic activity or in combination); sufficiently active (150–300 min per week of light- to moderate-intensity aerobic activity, 75–150 min per week of vigorous-intensity aerobic activity or in combination); insufficiently active (some activity but not enough to meet the more active definitions); and inactive (no light-to moderate- or vigorousintensity activity of ≥10 min per session).
17
ACCEPTED MANUSCRIPT Table 2. Strength training and characteristics associated with mortality in US older adults (N=30,162).a
27,250 (90.4) 2,725 (9.6)
Reference 0.55 (0.51, 0.61)
Reference 0.81 (0.69, 0.96)
74.2 (74.1, 74.3)
1.80 (1.75, 1.85)
11,412 (42.4) 18,750 (57.6) 25,842 (88.7) 3,396 (8.2) 924 (3.1)
RI
PT
All Cause OR (95% CI)
All Cardiac OR (95% CI) Reference 0.59 (0.51, 0.69)
1.86 (1.80, 1.93)
1.36 (1.28, 1.45) Reference
1.72 (1.57, 1.90) Reference
1.17 (1.09, 1.27) Reference
Reference 1.17 (1.08, 1.28) 0.57 (0.47, 0.69)
Reference 1.11 (0.94, 1.30) 0.78 (0.55, 1.08)
Reference 1.22 (1.08, 1.36) 0.61 (0.45, 0.81)
0.73 (0.65, 0.80) Reference
0.78 (0.64, 0.95) Reference
0.87 (0.74, 1.02) Reference
Reference 0.63 (0.59, 0.68) 0.60 (0.55, 0.66) 0.51 (0.46, 0.56)
Reference 0.88 (0.78, 0.99) 0.82 (0.71, 0.94) 0.77 (0.65, 0.90)
Reference 0.66 (0.60, 0.72) 0.63 (0.56, 0.70) 0.55 (0.48, 0.63)
12,410 (56.1) 12,949 (32.7) 2,794 (6.5) 1,962 (4.7)
Reference 1.75 (1.65, 1.86) 1.21 (1.09, 1.34) 1.40 (1.26, 1.58)
Reference 0.97 (0.87, 1.07) 0.99 (0.84, 1.16) 0.92 (0.74, 1.15)
Reference 1.82 (1.68, 1.97) 1.14 (0.98, 1.32) 1.55 (1.35, 1.79)
12,883 (42.3) 10,880 (36.8) 6,399 (20.9)
1.46 (1.35, 1.57) 0.93 (0.87, 1.01) Reference
1.35 (1.19, 1.55) 1.15 (0.99, 1.34) Reference
1.34 (1.20, 1.48) 0.92 (0.84, 1.02) Reference
3,541 (14.7) 3,032 (12.1) 4,653 (18.0) 14,952 (55.2)
0.32 (0.28, 0.35) 0.39 (0.35, 0.44) 0.59 (0.54, 0.64) Reference
0.73 (0.61, 0.87) 0.76 (0.64, 0.91) 0.80 (0.69, 0.94) Reference
0.33 (0.28, 0.39) 0.40 (0.34, 0.47) 0.66 (0.58, 0.74) Reference
5,717 (19.3) 23,579 (77.7) 866 (3.0)
Reference 1.46 (1.36, 1.58) 1.35 (1.13, 1.62)
Reference 1.08 (0.95, 1.23) 1.70 (1.32, 2.18)
Reference 1.51 (1.36, 1.67) 0.91 (0.70, 1.20)
3,436 (10.7) 11,236 (39.6)
1.79 (1.63, 1.95) 1.39 (1.31, 1.48)
2.94 (2.56, 3.42) 1.86 (1.67, 2.08)
1.05 (0.92, 1.19) 1.01 (0.93, 1.11)
NU
MA
D
2,631 (5.6) 27,522 (94.4)
SC
1.12 (1.08, 1.16)
TE
10,890 (33.3) 9,436 (33.0) 5,466 (18.6) 4,063 (15.0)
AC CE P
Characteristic Strength Trainingb < 2 times/week ≥ 2 times/week Group 1: Demographics Age (5-year increase) Gender Male Female Race Caucasian African American Other Ethnicity Hispanic Non-Hispanic Educational Status Less than HSc HS or GEDd Some College College Degree Marital Status Married Widowed Divorced Other Group 2: Health Behaviors Body Mass Index <25 (Normal) 25-30 (Overweight) 30+ (Obese) Physical Activitye Highly Active Sufficiently Active Insufficiently Active Inactive Alcohol Use Never Moderate Regular Unhealthy Regular Smoking Status Current Former
N (%) or Mean (95% CI)
Mortality All Cancer OR (95% CI)
18
ACCEPTED MANUSCRIPT Table 2. Strength training and characteristics associated with mortality in US older adults (N=30,162) (continued).
AC CE P
TE
D
MA
NU
SC
RI
PT
Never 15,490 (49.7) Reference Reference Reference Group 3: Comorbid Conditions Diabetes Yes 4,283 (14.1) 1.72 (1.60, 1.84) 0.92 (0.79, 1.06) 1.63 (1.49, 1.79) No 25,839 (85.9) Reference Reference Reference Hypertension Yes 14,585 (47.7) 1.31 (1.23, 1.38) 0.95 (0.87, 1.05) 1.48 (1.38, 1.60) No 15,577 (52.3) Reference Reference Reference Coronary Heart Disease Yes 3,577 (12.7) 1.92 (1.77, 2.08) 1.13 (0.98, 1.31) 2.32 (2.10, 2.53) No 26,434 (87.3) Reference Reference Reference Non-Skin Cancer Yes 4,586 (16.0) 1.60 (1.49, 1.70) 3.10 (2.80, 3.46) 1.01 (0.91, 1.12) No 25,496 (84.0) Reference Reference Reference a All odds ratios (OR) and confidence intervals (CI) are from a logistic regression analysis which was weighted based on the complex sampling stratification and clustering of the 1997-2001 National Health Interview Survey data. All percentages, means and confidence limits are also weighted based on the complex sampling stratification and clustering of the NHIS data. b Responses were categorized to signify whether the individual performed strength training activities at least twice each week, consistent with the ACSM/AHA Guidelines. c HS = high school. d GED = general educational development. e 2008 Physical Activity Guidelines for Americans: highly active (≥300 min per week of light- to moderateintensity aerobic activity, ≥150 min per week of vigorous-intensity aerobic activity or in combination); sufficiently active (150–300 min per week of light- to moderate-intensity aerobic activity, 75–150 min per week of vigorous-intensity aerobic activity or in combination); insufficiently active (some activity but not enough to meet the more active definitions); and inactive (no light-to moderate- or vigorousintensity activity of ≥10 min per session).
19
ACCEPTED MANUSCRIPT Table 3: Strength training associated with mortality adjusted for demographics, health behaviors, and comorbid conditions in US older adults.a
All-Cause
Total %
ST 2+ times/ week %
ST <2 times/ week %
31.6
21.3
32.7
Group 1 (N=29,637) OR (95% CI)
Groups 1 & 2 (N=25,848) OR (95% CI)
PT
Mortality
Groups 1, 2, & 3 (N=25,663) OR (95% CI)
AC CE P
TE
D
MA
NU
SC
RI
0.64 0.82 0.81 (0.57, 0.70) (0.73, 0.92) (0.71, 0.92) All Cancer 13.4 8.8 13.9 0.83 0.87 0.84 (0.70, 0.98) (0.70, 1.06) (0.68, 1.03) All Cardiac 7.2 6.1 7.4 0.72 0.88 0.88 (0.62, 0.83) (0.74,1.04) (0.73,1.04) a All percentages are weighted based on the complex sampling stratification and clustering of the 1997-2001 National Health Interview Survey data. All odds ratios (OR) and confidence intervals (CI) are from a logistic regression analysis which was weighted based on the complex sampling stratification and clustering of the NHIS data and was adjusted for the covariates corresponding the group as listed in Table 1.
20
ACCEPTED MANUSCRIPT Table 4: Strength training associated with mortality adjusted or matched via a propensity score based on demographics, health behaviors, and comorbid conditions in US older adults.a
AC CE P
TE
D
MA
NU
SC
RI
PT
Adjusted for Propensity Scoreb Matched by Propensity Scorec Group 1 Groups 1 & 2 Groups 1, 2, & 3 Group 1 Groups 1 & 2 Groups 1, 2, & 3 Mortality (N=29,637) (N=25,848) (N=25,663) (N=5,408) (N=4,622) (N=4,600) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) All-Cause 0.63 0.82 0.82 0.69 0.90 0.84 (0.57, 0.70) (0.73, 0.91) (0.73, 0.92) (0.60, 0.79) (0.78, 1.04) (0.72, 0.96) All Cancer 0.81 0.84 0.83 0.85 1.03 0.84 (0.68, 0.96) (0.69, 1.03) (0.67, 1.01) (0.68, 1.06) (0.78, 1.36) (0.66, 1.08) All Cardiac 0.69 0.88 0.89 0.76 0.94 0.84 (0.59, 0.80) (0.74,1.05) (0.75,1.06) (0.63, 0.93) (0.75, 1.19) (0.66, 1.04) a All odds ratios (OR) and confidence intervals (CI) for strength training are from a logistic regression analysis which was weighted based on the complex sampling stratification and clustering of the 1997-2001 National Health Interview Survey data and was either adjusted for the propensity score derived from the covariates corresponding the groups as listed in Table 1 or matched strength training groups 1-to-1 based on the same propensity score. b All data from the sample adjusted for propensity score without any covariates included since propensity score was generated from those covariates. c Reduced sample matched by propensity score without any covariates included.
21
ACCEPTED MANUSCRIPT Is strength training associated with mortality benefits? A 15 year cohort study of US older adults. Jennifer L. Kraschnewski, M.D., MPH1,2,3, Christopher N. Sciamanna, M.D.,1,2,3, MPH, Jennifer M. Poger,
PT
M.Ed.1, Liza S. Rovniak, Ph.D., MPH1,3, Erik B. Lehman, MS3, Amanda B. Cooper, M.D.2, Noel H. Ballentine, M.D., F.A.C.P.2, & Joseph T. Ciccolo, Ph.D.4
Department of Medicine, Penn State College of Medicine, 2Department of Medicine, Penn State
RI
1
4
Department of Applied Physiology, Columbia University
TE
D
MA
Guideline-concordant strength training reduces all cause, cancer, and cardiac death. Only a minority of older adults meet current strength training guidelines. All-cause mortality may be reduced through strength training interventions.
AC CE P
NU
Highlights
SC
Hershey Medical Center, 3Department of Public Health Sciences, Penn State College of Medicine,
22