Acre , Brasil : estudo de base populacional Association of handgrip strength with self-reported diseases in adults in Rio Branco , Acre State , Brazil : a population-based study

This study aimed to analyze the association of handgrip strength with self-reported diseases and multimorbidity among adults in Rio Branco, Acre State, Brazil, through a population based survey involving 1,395 adults of both sexes. Associations by sex were estimated by logistic regression analysis. The mean handgrip strength in men (44.8kg) is higher than in women (29kg) and decrease with age. The mean handgrip strength difference between those classified as strong and weak was 21kg and 15.5kg for men and woman, respectively. Controlling for age group, body mass index and physical activity when it was relevant, men with low handgrip strength were more likely to have hypertension [OR = 2.21 91.35; 3.61)], diabetes [OR = 4.18 (1.35; 12.95)], musculoskeletal disorders [OR = 1.67 (1.07; 2.61)] and multimorbidity [OR = 1.99 (1.27; 3.12)]. Among woman, associations between handgrip strength and cardiovascular disease, dyslipidemia, musculoskeletal disorders and multimorbidity were not sustained in the multivariate models. This study endorses the use of handgrip strength as a health biomarker. Handgrip Strength; Morbidity; Health Surveys ARTIGO ARTICLE

and decrease with age.The mean handgrip strength difference between those classified as strong and weak was 21kg and 15.5kg for men and woman, respectively.Controlling for age group, body mass index and physical activity when it was relevant, men with low handgrip strength were more likely to have hypertension [OR = 2.21 91.35; 3.61)], diabetes [OR = 4.18 (1.35; 12.95)], musculoskeletal disorders [OR = 1.67  (1.07; 2.61)] and multimorbidity [OR = 1.99  (1.27; 3.12)].Among woman, associations between handgrip strength and cardiovascular disease, dyslipidemia, musculoskeletal disorders and multimorbidity were not sustained in the multivariate models.This study endorses the use of handgrip strength as a health biomarker.

Introduction
Handgrip strength is recognized as an estimator of overall strength and has been presented as a biomarker for important health outcomes 1 .Studies with predominantly middle-aged and elderly individuals show that low handgrip strength is associated with sarcopenia 2 , functional limitations and disabilities 3 , falls 4 , decreased bone mineral density, and increased fracture risks 5 , and is considered a useful marker for frailty in the elderly 6 .Among men 40 to 68 years of age followed for 25 years, low handgrip strength was predictive of functional limitations and disabilities, while higher handgrip strength apparently protected against these conditions in old age, indicating that handgrip strength can be used for early screening of individuals at increased risk of physical disability in old age 4 .Mean decline in handgrip strength during the follow-up period was 8-9kg and was inversely associated with age and blood glucose, but directly associated with cognitive function, body mass index (BMI), and hemoglobin level 7 .
In addition to disorders inherent to the musculoskeletal system, low handgrip strength has also been associated with changes in nutritional status 8 , post-surgical clinical complications 9 , length of hospital stay 10 , various chronic diseases 11,12 , and mortality 13 , although the mechanisms of these associations are not well understood.
Low handgrip strength has been associated with increased odds of anxiety, stroke, chronic kidney disease (CKD), chronic obstructive pulmonary disease (COPD), and hyperthyroidism in men and anemia, falls, and kyphosis in women 10 .Among men and women 30 to 72 years of age followed for 22 years, handgrip strength decline was associated with incidence of chronic diseases such as cardiovascular events, diabetes mellitus, chronic bronchitis, chronic back pain, hypertension, and asthma as well as important weight loss, physical inactivity, and persistent smoking 12 .According to a Brazilian study, women with metabolic syndrome showed lower mean handgrip strength than healthy ones 14 .A study in men in the United States found a protective effect of muscle strength against metabolic diseases, regardless of cardiopulmonary fitness and overweight 15 .Several studies have found an association between diabetes mellitus and decreased handgrip strength 16,17,18 .
The 37-year follow-up of a cohort of one million men starting with their enlistment in the Swedish Army (mean age 18.2 years) identified an inverse relationship between handgrip strength and risk of heart disease and stroke 19 .Meanwhile, the seven-year monitoring of older elderly showed less variation in handgrip strength among elders with higher initial handgrip strength, regardless of gender, with handgrip strength being an important predictor of mortality 13 .
Potential prediction of morbidity based on measurement of handgrip strength suggests the variable's use as a biomarker in the assessment of health conditions in the population.This emphasizes the importance of accumulating knowledge from studies in different contexts to determine cut-off points for different diseases, not currently available in the literature.
Despite evidence from international studies, we are unaware of any epidemiological study on the topic in Brazil and that includes a wide age spectrum.The current study thus aimed to begin filling this gap by analyzing the association between handgrip strength and specific diseases and multimorbidity among adults in Rio Branco, Acre State, Brazil.

Methods
A cross-sectional study was conducted with adults in Rio Branco, under the research project Health and Nutrition of Children and Adults in Rio Branco, Acre, from November 2007 to October 2008.
A two-stage probabilistic cluster sample was used, with 35 census tracts as the primary units, 31 in the urban area and four in the rural area.Twenty-five households were randomly selected in each census tract and represented the secondary units, increased by 15% to compensate for losses or refusals, totaling 977 households, where all residents 18 years or older and able to answer the questions were asked to participate in the study.
The selected sample consisted of 1,516 adults from 18 to 96 years of age (the procedures have been presented elsewhere 20 ).Pregnant women and participants who did not perform the handgrip strength test were excluded from the survey, leading to a loss of 121 subjects (7.8%), with no statistically significant difference according to socio-demographic profile.The final sample included 1,395 participants, considering the demographic characteristics (age and gender), leisuretime physical activity, and self-reported diseases, as well as the biometric variables height, weight, and handgrip strength.
The independent handgrip strength variable was obtained with a hydraulic hand-held dynamometer (SAEHAN SH5001, Saehan Corp., Dangjin, South Korea) with kgf resolution.The assessment adopted the sitting position with the elbow in 90° flexion, following procedures used by the American Society of Hand Therapists 21 .The handgrip strength score was the higher value of two measurements of the dominant hand.Individuals were classified into tertiles according to their handgrip strength score -strong (highest tertile), moderate strength (middle tertile), and weak (lowest tertile).
The dependent variables for self-reported diseases were identified by the individual's account of diagnosis by a health professional for the following conditions: hypertension, diabetes mellitus, cardiovascular events (myocardial infarction, stroke), dyslipidemia (high cholesterol or triglycerides), depression, chronic kidney disease, and musculoskeletal disorders (tendonitis, repetitive strain injury, spine or back disease, arthritis, non-infectious rheumatism, gout, and osteoporosis).The multimorbidity variable was built adopting the definition of the simultaneous occurrence of two or more chronic diseases in the same individual.A value of 1 for "yes" and 2 for "no" was assigned to each variable indicating occurrence of the disease.
The covariates were age, leisure-time physical activity, and BMI.Age was categorized in two groups, 18-39 years and 40 years and older.Leisure-time physical activity was defined according to weekly frequency and duration.According to recommendations by the World Health Organization (WHO) 22 , individuals totaling 150 minutes of moderate physical activity or 75 minutes of vigorous activity were classified as active, and those who did not achieve these levels were considered sedentary.BMI was defined as weight divided by height-squared, using WHO 23 cutoff points: underweight (BMI < 18.5); normal weight (BMI = 18.5 to 24.9); overweight (BMI 25 to 29.9); and obese (BMI ≥ 30).
Data were double-entered and validated using Epi Info 6.04 (Centers for Disease Control and Prevention, Atlanta, USA).
In the descriptive analysis, we verified the absolute and relative frequencies of all the variables analyzed by gender and the estimated differences in frequencies between men and women using Pearson's chi-square test, with significance set at α = 0.05.We also obtained handgrip strength measures of central tendency and dispersion according to gender and age group.
For men and women, logistic regression models estimated the magnitude of association as odds ratio (OR) between the dependent variables indicating diseases and handgrip strength in tertiles, with the highest handgrip strength tertile (strong) as the reference.Three models were estimated for each dependent variable: the first model focused on crude association between disease and handgrip strength; the second model on age group-adjusted association; and the third model on association adjusted for age group, BMI, and when relevant, leisure-time physical activity.Age group-handgrip strength interaction was tested.Significance was set at α = 0.05.
All the analyses took into account the effect of the sample design and weights of observations, using SAS version 9.3 (SAS Inst., Cary, USA) surveyfreq, surveymeans, and surveylogistic procedures.
The research project that collected the data used in this study was approved by the Ethics Research Committee of the Federal University of Acre under Protocol n. 2307.001150/2007-22 and obtained informed consent from each participant.

Results
With the sample expansion using the sampling weights, the 1,395 observations corresponded to 248,479 individuals.Estimates point to a predominantly female population (54.6%) aged up to 39 years (59.3%).There were statistically significant differences (p < 0.05) between genders in the distributions of type of physical activity, BMI, hypertension, dyslipidemia, depression, musculoskeletal disorders, and multimorbidity (Table 1).
Overall mean handgrip strength was 36kg (44.8kg in men and 29kg in women).Regardless of gender, handgrip strength was also higher in the 18-39-year age group than in 40 years and over.In the handgrip strength analysis per tertile, strong and weak men had a mean handgrip strength of 55.3kg and 34.1kg, respectively, while strong and weak women had a mean handgrip strength of 36.1kg and 20.6kg, respectively (Table 2).
Table 3 shows the results for the three logistic regression models used in the analysis of associations between different diseases and handgrip strength for men.After adjusting for age group, the odds of hypertension were statistically higher among individuals classified as moderately strong or weak, as well as for diabetes mellitus, musculoskeletal disorders, and multimorbidity among weak individuals when as compared to the reference group of individuals classified as strong.Considering the models with adjustment by age group, BMI, and (when relevant) by leisure-time physical activity, despite some variation in the magnitude of associations by adjusting only for age, the results remained consistent.The odds of all the target diseases were higher in the older age group, while increased BMI was significant for the occurrence of hypertension, diabetes, dyslipidemia, and multimorbidity.Lei- sure-time physical activity was positively associated with multimorbidity.Table 4 presents the results corresponding to the previous table but for women, showing as-sociations between classification as weak (versus strong) and cardiovascular events, dyslipidemia, musculoskeletal disorders, and multimorbidity, only in the model adjusted for other variables.Interaction terms between age and handgrip strength were tested in the multivariate models in Table 3 and 4, but were not statistically significant (p < 0.05).

Discussion
The results showed associations between handgrip strength and self-reported hypertension, diabetes, musculoskeletal disorders and multimorbidity in men only.Reduced muscle strength, a condition known as dynapenia 2 , was observed in the 40-and-over age group.
The study found significant differences in the magnitude and gradient of muscle strength between men and women, corroborating previous studies 11,24 , which can be explained by hormonal differences inherent to gender.
There was no statistically significant association between low handgrip strength and the cardiovascular events.However, metabolic syndrome, an important risk factor for cardiovascular disease that results precisely from the combination of dyslipidemia, hyperglycemia, and hypertension 25 , was shown to be individually associated with handgrip strength in men.The components of metabolic syndrome are thus associated with chronic systemic inflammation and increased interleukin-1 and -6 (IL-1 and IL-6) and tumor necrosis factor-alpha (TNF-α) 26 .High levels of inflammatory markers such as IL-6 and C-reactive protein (CRP) increase the risk of muscle strength loss in older men and women 27 , who thus tend to decline in physical function and in-crease in functional disability, dependence in activities of daily living, and mortality 28,29 .Evidence have been reported on the progressive reduction of handgrip strength in the presence of catabolic biomarkers (CRP, IL-6, IL-1RA, TNF-α) 30 , which increase oxidative stress, reducing muscle mass and causing consequent loss of strength in the elderly 2,31,32 .
The present results corroborate other studies showing that diabetic men have lower handgrip strength than their non-diabetic peers, but the same was not evident among women 11,17 .
Prospective studies indicate that type-2 diabetes reduces muscle strength and mass 18 , and that higher strength protects against the development of diabetes 33 .In vitro and in vivo clinical evidence attests that hyperglycemia affects contractile function and muscle strength 34 .
This study also agrees with others identifying the association between low strength and hypertension in men, but not in women 35,36 .It has been reported that resistance training appears to prevent metabolic disorders such as dyslipidemia, impaired fasting glucose, pre-hypertension, and increased waist circumference, but not hypertension 37 , while recognizing that increased strength may improve vascular health and reduce complications 38 and mortality among hypertensive individuals 39 .The statistical association between handgrip strength and hypertension may mean that muscle strength expresses overall individual fitness 40 more than a direct relationship with hypertension.
The association between low handgrip strength and musculoskeletal disorders in men Cad.Saúde Pública, Rio de Janeiro, 31(6):1-12, jun, 2015 from Rio Branco appears to echo the relationship between handgrip strength and overall strength, which itself reflects operation of the musculoskeletal system 1 .Low strength has already been associated with history of falls in both sexes and with kyphosis in women 11 .This study's findings thus highlight the importance of using handgrip strength as a health biomarker, realizing that reduced levels of muscle strength may lead to disability and functional limitations, particularly among older individuals 2,3,41 .The assessment of handgrip strength during middle age could allow early identification of risks of future disability 3 , dependence in activities of daily living, and cognitive decline in older age 4 .It can also play a role in predicting fracture risks 5 and tracking sarcopenia 2 .
Part of the modeling process in this study focused on differentiating between individuals 60 years and older and 40-59 years of age, confirming the role of aging in the occurrence of diseases and providing relatively consistent results.However, the small number of strong individuals in the 60-and-older group, especially among men, led to loss of power in the inferences.
Future studies should contribute to understanding the effect of handgrip strength in older individuals, as well as differences observed here in the role of handgrip strength as a predictor of morbidity between men and women.
As far as we know, this is the first study to test interaction between handgrip strength and age when assessing the association between handgrip strength and diseases.The findings do not indicate potentiation or attenuation of the effect of low handgrip strength with older age in the occurrence of diseases.
The study has some limitations, such as the inability to make causal inferences.The statistical associations should only be considered as such, with due caution concerning assumptions about which variables precede which.Another limitation is the lack of clinical parameters for diseases, although self-reported chronic diseases express an approximate measure of the information obtained by clinical examination 42 .
We highlight that this research was unprecedented in Brazil, as the first Brazilian populationbased study with adults that assesses handgrip strength and diseases.The models used here also considered the effect of handgrip strength adjusted by the main variables published in the literature: age, BMI, and leisure-time physical activity.

Conclusion
The findings confirm the association between low handgrip strength and chronic diseases, musculoskeletal disorders, and multimorbidity among men, supporting the measurement of muscle strength with a hand-held dynamometer as a useful, relatively lowcost, and easy-to-apply marker for clinical evaluation and monitoring of individual health conditions, especially in primary care.
The study also indicates the need for further epidemiological research to improve our understanding of the findings based on clinical parameters for diseases and focused on specific age groups, explaining the differences between men and women and contributing to proposals of reference values and cutoff points for health risks.

Table 2
Distribution of handgrip strength in kg by age group and handgrip strength tertile by gender in adults in Rio Branco, Acre State, Brazil, 2007-2008.

Table 3 Logistic
regression analysis of handgrip strength tertiles and self-reported diseases in men in Rio Branco, Acre State, Brazil, 2007-2008.

Table 4
Logistic regression analysis of handgrip strength tertiles and reported diseases in women in Rio Branco, AcreState, Brazil,   2007-2008.