Handgrip strength values of Portuguese older adults: a population based study

Handgrip strength (HGS) is a marker of overall body muscle strength and is used to identify sarcopenia and frailty phenotypes [1].

At hospital admission, lower HGS was associated with a decreased probability of discharge alive over time [2], moreover in community-dwelling older adults, HGS was a potential predictor of mortality (multivariate-adjusted relative risk of all causes of death for the lowest HGS quintile: 1.38, 95% CI 1.01–1.88, in men and 1.54, 95% CI 1.20–1.98, in women) [3]. Muscle strength is an indicator of muscle quality has shown to be more significant than muscle mass in estimating mortality risk [4]. One possible explanation is that loss of motor neurons with age leads to a rise in size of remaining motor units, but with higher preservation of type 1 fibers, resulting in conservation of mass with relatively less type 2 fibers, and thus, lower strength [5].

According to recent demographic projections, the world population is rapidly getting older [6, 7]. The European Union population aged ≥65 years is estimated to rise by over 25% by 2035 [6, 7]. In this context, it has been reported that HGS reference values for older adults are required [8], but HGS data have not yet been described for the Portuguese population. The use of other populations as reference may be unsuitable, because there are variations in the skeletal muscle mass and function from individuals from different ethnicities [9].

The aim of this study was to describe HGS values of older population aged ≥65 years in Portugal, considering the possible influence of anthropometric parameters.

Characteristics of the sample are summarized in Table 1. Age ranged from 65 to 100 years, with a mean of 75 ± 7.1 years. Women represented a slightly higher proportion of the sample (58%). Most of the participants were from the Centre of Portugal (51.6%) and approximately half of them were married (44.5%).

Mini Mental State Examination ranged from 10 to 30 points in total sample. According to cut-offs of the Portuguese version of MMSE, scores are presented by mean ± standard deviation; median (25th percentile-75th percentile) as follow: [12 ± 1.7; 13 (11–14) in “cognitive impairment” and 22 ± 0.3; 22 (19–25) in “no cognitive impairment”, if participants had no education]; [18 ± 3.2; 19 (16–21) in “cognitive impairment” and 28 ± 2; 28 (26–29) in “no cognitive impairment”, if participants had 1 to 11 years of school completed]; [26 ± 1.2; 26 (25–26) in “cognitive impairment” and 29 ± 0.8; 29 (29–30) in “no cognitive impairment”, if participants had >11 years of school completed].

In total sample, chronic diseases were presented in the following proportions of cases: depression (19.9%), chronic neck problems or neck pain (50.1%), chronic lumbar problems or lumbar pain (61.9%), arthrosis (59.7%), asthma and allergies (8.6%), chronic bronchitis or emphysema or chronic obstructive pulmonary disease (15.8%), myocardial infarction or its chronic consequences (3.7%), angina pectoris or coronary heart disease (17.1%), hypertension (63.9%), stroke or its chronic consequences (7.1%), diabetes (27.5%), hepatic cirrhosis (3.9%), bladder control problems or urinary incontinence (12.1%), chronic renal disease (9.1%).

Height ranged from 131.8 to 170 cm in women, with a mean of 151.2 ± 6.2 cm, and from 140.3 to 185 cm in men, with a mean of 164.8 ± 6.8 cm. Participants showed an average of BMI of 29.9 ± 5 kg/m² in women and of 28.5 ± 4 kg/m² in men. According to the MNA-SF, 1.3% of women and 1.3% of men were classified as undernourished, and 17.4% of women, as well as 11.1% of men were classified at nutritional risk.

Handgrip strength was higher among men than among women (30.3 ± 9.2 Kgf vs 18 ± 5.4 Kgf, p < 0.001). Handgrip strength values of Portuguese older adults, stratified by age and by sex-specific tertiles of height were described in Table 2.

Age (r = − 0.44, in women; r = − 0.42, in men), cognitive status (r = 0.32, in women; r = 0.37, in men), self-reported sitting time (r = − 0.38, in women; r = − 0.34, in men), practice of physical activity (r = 0.37, in women; r = 0.31, in men) and MNA-SF (r = 0.19, in women; r = 0.16, in men) were significantly correlated (p < 0.001) with HGS values (Table 3). Height was the anthropometric parameter most correlated with HGS (r = 0.34, p < 0.001, in women and r = 0.40, p < 0.001, in men) (Table 3 and Fig. 1).

Information about HGS mean values of older adults from different countries are presented in Table 4. In general, HGS values of Portuguese older adults were lower than other older populations. However, present sample presented higher HGS mean values when compared to Spanish and Chilean women, as well as Spanish and Mexican men (Fig. 2). Despite United Kingdom older adults presenting higher HGS values than the Portuguese sample, they were closer when the results are presented stratified by height (Fig. 3).

Several HGS values, resulting from populations with different nationalities and ethnicities, have been published in the literature cited as reference values [20‐34], however for the Portuguese population HGS values had never been described. In addition, HGS of older adult populations, despite their particular characteristics, are not always analyzed separately from adult populations. The grouping of strength values together for all older adults was also referred as a limitation in Bohannon et al. meta-analysis, failing to take into consideration the continued decline in muscle strength that occurs as individuals age [35]. This problem was overcome by the report of data for strata of age after 75 years, but restricted to samples from confined geographic regions of United States [36].

Overall, Portuguese older adults had lower HGS values than other older populations. A wide range of factors, including individual characteristics and those related to different nationalities and ethnicities may have influenced these results.

The present study included individuals classified with cognitive impairment according to the Portuguese version of the MMSE and a decreased grip strength is strongly associated with developing mild cognitive impairment [37]. Mild cognitive impairment is a transitional state that goes beyond typical age-related cognitive changes, but is marked by less severe impairment than dementia [38]. The association between HGS and mild cognitive impairment could be due to several mechanisms. Both motor and cognitive performance rely upon the nervous system to execute activity, thus, a compromised nervous system (e.g., as a result of inflammation) may lead to general deficits in both areas [39]. On the other hand, some degree of cognitive impairment is common in the ageing process and the non-exclusion of those participants decreased the risk of a selection bias [1, 40]. In addition, although 9.5% of the participating older adults presented cognitive impairment, they were included only if they were able to understand the instructions and to undertake the HGS technique correctly, in order to avoid inclusion of cases with true underlying neurodegenerative processes. Nevertheless, additional clinical diagnostic tools would be necessary to confirm the stage of the cognitive impairment and also to infer about these results.

Another point is that our sample presented a very high proportion of participants who reported chronic diseases, in particular arthrosis, cervical and lumbar pain. However, according to Onder et al., daily pain is highly prevalent among community-dwelling older adults, despite being associated with impaired muscle strength [41]. A long sitting time period, as well few hours of practice of physical activity per day were also factors that can justify low HGS values in our study, since these parameters are potential independent health indicators among older adults [42].

Nutritional impairment has been closely associated with low HGS values [43] and in this sample, 14.8% of participants were at nutritional risk and 1.3% undernourished. However, this proportion of older adults at nutritional risk was similar to those previously reported, which does not explain the low HGS values [44].

Simultaneously, overweight identified through the World Health Organization criteria, was highly frequent among participants of both sexes [45]. Although overweight and obesity have also been associated with reduced strength [46], this association was not significant in the present work. Body mass index reflect the ratio of weight to height, and there was also no significant association between weight and HGS. It is important to emphasize that total weight use as a surrogate indicator of adiposity can be prone to errors, because fat mass and fat free mass are not distinguished [47]. In addition, according to some epidemiological evidence, older populations display higher optimal BMI intervals than younger people [47]. These facts can induce confounding in understanding association between BMI and HGS.

On the other hand, height was directly and significantly associated with HGS in women and men. Accordingly, in Han et al. study, skeletal muscle mass adjusted by height correlated better with muscular functions and HGS, than that adjusted for body weight [48]. Height reflects part of the bone structure, and bone mass has also demonstrated implications for muscular and strength performances [26, 48]. The International Working Group on Sarcopenia recommends the inclusion of height value for determining relative muscle mass, in a context of functional disability [49]. The effect of skeletal muscle mass on HGS values should not be interpreted alone, but together with other structural and anatomic measurements, including the hand anatomy [50].

In addition, although in overall HGS of United Kingdom sample were higher than HGS of Portuguese sample, Portuguese HGS values were more similar to United Kingdom values when stratified by height. This fact reinforces the importance of height adjustment in the interpretation of HGS results.

Indeed, different HGS mean values were observed according to different countries, but this fact has also been reported by other studies [20] and the nature of these differences has not yet been resolved [20]. However, heterogeneous designs may explain some of the differences among studies results. For example, in data from Danish older adults, HGS decline with ageing was presented in the end of the follow-up [27]. On the contrary, in the other analyzed studies, including the present report, HGS data was obtained from cross-sectional designs [22, 23, 26, 28, 29, 31‐33] or from cross-sectional analysis of cohorts [20, 21, 24, 25, 30, 34].

As expected, HGS values of populations from Chile, Mexico and Spain were closer to present data than HGS values of those from United States, China or Australia. Nevertheless, Chilean and Spanish studies were based on convenience samples [28, 31] and the lack of representation can justify some deviation in the mean of HGS comparing to our study. In addition, in Chilean, Spanish and Brazilian studies [28, 31, 32], as well as in studies from United States and China [29, 33], potential comorbidities of participants were not described. Otherwise, in studies from Sweden, United Kingdom, Poland and Australia [20, 23, 25, 34], participants with chronic diseases, such as osteoarthritis and inflammatory disease, were excluded and this fact can explain, in part, their higher HGS values in comparison to present study.

Handgrip strength of Italian population was also slightly higher than HGS of Portuguese sample, but is important to note that Veronese et al. presented HGS results for the mean value of both hands [21] while results for the non-dominant hand presented here. Non-dominant hand HGS measurement were used because values of the dominant hand are more affected by occupational load than values of the non-dominant [19]. Consequences of occupational load can induce localized muscular development or an excessive articular contact stress and so, the non-dominant hand can express more independently the overall muscle strength [19].

It is also uncertain which of the two factors, genetic or environment, are more decisive to the HGS results [20]. Besides ethnic differences in height and in skeletal muscle mass and function [9], there are well-recognized differences in dietary protein intake according to different countries, and the variation in dietary patterns may also explain differences in muscle strength [51]. Education and socio-economic status are other factors that can explain differences in HGS ranges among countries [20, 52, 53].

In present analysis, individuals from the United States showed the highest HGS values and Spanish, followed by Chilean and Mexican individuals, showed the lowest HGS values. This is in agreement with results from a recent study in which grip strength data from regions in South America were clearly low compared to developed regions of the United States [54]. In a systematic review about associations between HGS and sociodemographic factors, older adults with lower HGS values had lower educational levels [52]. Accordingly, in our sample 14% of participants had never attended school, as well as 77% of them had only between one and four completed school years. Countries with contrasting income may therefore be expected to be associated with wider discrepancies in HGS values [20, 52, 53].

The main strength of this work was to describe, for the first time, HGS values of the Portuguese population aged ≥65 years in a nationwide representative sample. Handgrip strength reference values of at least 85% of the mean of the population are used to describe normality [28]. However, the values of our sample were frequently below the cut-off of 85% of the HGS mean due to changes related to the aging process and were similar to those associated with weakness and mobility limitation (<16 Kgf in women and <26 Kgf in men), according to the Foundation for the National Institutes of Health Sarcopenic Project [49]. This Portuguese older population presented very low HGS values. Considering this fact together the high frequency of overweight, obesity and chronic disabilities, these first results are a matter of concern indicating the need of public health interventions. In addition, as a wide range of variables may influence HGS, reflection is required on how normative HGS values should be constructed for this age group and which variables should be considered in their definition.

The comparison between HGS data published in the literature is hampered by the heterogeneous study designs and by heterogeneous characteristics of the samples and of the methodology used. Regarding to HGS measurements, some authors reported only “left” or “right” hand, without referring to the hand dominance [22, 23, 25, 26, 29, 31, 32, 34], and in other cases only the average of both hands was referred [20, 21, 24, 27, 33], introducing some bias in data analysis. The importance of a consensus on which hand HGS measurements should be performed, is herein reinforced, to improve data comparison in future works. In addition, in some studies, the lowest age range included subjects aged 60 years and over, while in others, the lowest age was 65 years. The maximum age of participants was also not reported in five of the 15 analyzed studies [29, 30, 32‐34]. Therefore, a bias related to the differences in age groups may has been introduced in the comparison of HGS values.

This study described, for the first time, HGS values of Portuguese population aged ≥65 years, in accordance with age and with sex-specific tertiles of height. The definition of HGS reference values in this age group merits further reflection.