Eur J Appl Physiol (2007) 100: 535–542 DOI 10.1007/s00421-006-0195-8

ORIGINAL ARTICLE

Paolo Capodaglio · Maria Capodaglio Edda Marco Facioli · Francesco Saibene

Long-term strength training for community-dwelling people over 75: impact on muscle function, functional ability and life style

Accepted: 22 March 2006 / Published online: 25 April 2006© Springer-Verlag 2006

Abstract The objective was to determine the impact ofa 1-year mixed strength-training programme on musclefunction, functional ability, physical activity and lifestyle. Twice-a-week hospital-based exercise classes and aonce-a-week home session were conducted. Nineteenhealthy community-dwelling training (T) men (76.6 §3.1 years), 19 women (77.5 § 4.0 years) and 20 matchedcontrols (C) participated in this study. Training wasgiven with a two multi-gym machines for the lower limbs(Sitting calf and Leg press, TECHNOGYM, Italy) at60% of the repetition maximum (1 RM) and at home itwas with elastic bands. The following were the measure-ments made: muscle function—maximum isometricstrength of the knee extensors (KE) and ankle plantarXexors (PF) measured with a Cybex Norm dynamome-ter, leg extensor power (LEP) with the NottinghamPower Rig; functional abilities—functional reach, chairrise, bed rise, 6-min walking test, stair climbing, get upand go, one-leg standing; physical activity—aerobicactivities over 3 MET intensity (AA3), intensity classes;life-style—mean daily energy expenditure (MDEE). Sig-niWcant gains in muscle function and functional abilitiesin both training females and males were observed, butfemales improved signiWcantly more than males. Males(T + C) showed higher AA3 times than females (T + C)(P = 0.02), with females signiWcantly more involved inlight-intensity activities. We observed a 60% increase(t = 2.45) in AA3 time in T, but no increase in C.Trained males increased Class 2 physical activity time by

146% (t = 2.82) and trained females by 16% (t = 2.23).MDEE increased by 10% (t=2.62) in trained males. Ourlong-term mixed programme can improve muscle func-tion and functional abilities in elderly females and func-tional abilities in males. It can positively aVect theamount of habitual physical activity and the life-style ofmales and females over 75.

Keywords Strength training · Functional ability · Life-style · Elderly

Introduction

Mean daily energy expenditure (MDEE, Kcal kg¡1 day¡1)is strongly inXuenced by the habitual level of physicalactivity (PA) performed and is known to decrease withage (Shephard 1991). A decline in the PA contributes tothe age-related decline of aerobic power (Jackson et al.1995) and thus the capacity to perform moderate-to-vig-orous tasks of daily life, such as lifting a shopping bag,walking up a moderate slope, lifting the body mass froma chair (Shephard 1991). The increase of body fat massand the decrease of mechanical eYciency tend to increasethe energy expenditure of an older person during PA atany given intensity. Environmental barriers and socio-cultural stereotypes might contribute to a popular beliefthat PA for elderly people is somehow unsafe. On thecontrary, regular PA has been shown to maintain biome-chanical eYciency and aerobic capacity in the healthyadult even at a moderate intensity (i.e. 1-h exercise at50% of maximal oxygen intake) (ACSM 1998). PA boutsat moderate intensity three times a week or 30 min everyday, corresponding to 400–700 Kcal week¡1, are recom-mended to maintain a healthy status and counteract theeVects of ageing.

The decrease in PA is typically age-related and issecondary to the deterioration of biological functions(i.e. balance, poor eyesight, breathlessness, etc.) and pain(i.e. arthrosis).

This study was supported by EC Contract QLK6-CT-2001-00323(“Better Ageing”).

P. Capodaglio (&) · M. Capodaglio Edda · M. FacioliUO Neuroriabilitazione, Fondazione Maugeri IRCCS, via Boezio 28, 27100 Pavia, ItalyE-mail: pcapodaglio@fsm.itFax: +39-0382-593081

F. SaibeneIstituto Tecnologie Biomediche Avanzate, Consiglio Nazionale delle Ricerche, Milano-Segrate, Italy

536

Another reason for this is, the 3.5% decline per annumof leg extensor power (LEP) and 1.5% decline of themuscle strength for knee extensors (Skelton et al. 1994).LEP is needed for many basic activities in daily life, suchas walking, negotiating stairs. Women, in particular, mayreach levels below the threshold for tasks important foran independent life (Skelton et al. 1994).

Maintaining LEP in old age is important in order to beable to continue to complete the daily tasks successfullyand within the reasonable safety margins. Impairments inLEP are important factors limiting mobility in both nurs-ing home residents (Bassey et al. 1992) and community-dwelling elders Suzuki et al. 2001; Bean et al. 2002).

There is now greater attention on the need to increasemuscle power in older populations (Skelton et al. 1994;Bassey et al. 1992; Suzuki et al. 2001; Bean et al. 2002;Malbut et al. 1993; Bassey et al. 1993; Foldavari et al.1999; Jozsi et al. 1999; Earles et al.2001; Macaluso et al.2003; Latham et al. 2004) and to maintain it in the longterm. Long-term programmes, 6–12 months, have beenshown to increase muscle strength (Cress et al. 1991;Pyka et al. 1994), aerobic capacity (Carmeli et al. 2003)and function (Shephard 1991). But whether these pro-grammes can induce signiWcant changes in an individ-ual’s life-style is still under debate (van der Bij et al.2002). A shift towards “chronic” physical activity shouldbe the goal of the interventions targeted to improve andmaintain health eVects in the long term.

The majority of training studies consist of supervisedtraining. This modality can reach a limited number ofpeople and is unlikely to be maintained in the long term.Partially supervised, home-based or mixed-training pro-grammes could potentially have a larger impact on theelderly population (McMurdo et al. 1995).

The aim of this study was to determine if a long-termmixed strength-training programme could impact onmuscle function, functional ability and ultimately on life-style in healthy community-dwelling men and womenover 75 years of age.

Methods

Subjects

We recruited 38-healthy elderly aged 70–83 years, 19males (M) aged 76.6 § 3.1 and 19 females (F) aged77.5 § 4.0, to participate in a 1-year training programme.Exclusion criteria were: neoplasticity, previous stroke,Alzheimer, major lung pathology, motor neuron disease,Parkinson, major musculoskeletal disorders, majorendocrine disorders, severe diabetes, history of hyperten-sion or myocardial infarction within the previous twoyears, use of beta-blockers or not in sinus rhythm, acutefebrile or systemic disease within the previous two years.

Subjects were non-randomly allocated to training (T)and control group (C) according to quasi-experimentaldesign criteria. T included 11 M (MT) and 12 F (FT); C

included 8 M (MC) and 7 F (FC), their physical charac-teristics are presented in Table 4.

Intervention

Control group was not engaged in the PA programme,but regular contact was maintained with C and theyunderwent the same measurements as in T.

Training group underwent the following physicaltraining: supervised exercise classes were held twice-a-week in the hospital gym; the subjects started each exer-cise class with a 15-min warm-up including light rhyth-mical, low-intensity aerobic activities, Xexibilityexercises, supported stretching exercises. Tai-Chi Yangstyle exercises involving slow controlled movements indiVerent planes of the upper and lower limbs wereincluded.

Strength training was carried out with multi-gymmachines at 60% of the repetition maximum (1 RM).Typically, the 70–80% of the 1 RM is used, but we decideto investigate if lower training loads can be eVective.Training was carried out on two variable-resistancemachines for the lower limbs (Sitting calf and Leg press,TECHNOGYM, Italy). Subjects performed one set of 12repetitions on each machine at an intensity initially set at40% 1 RM value and gradually increased to 60% of1 RM within a month. A 2-min rest between warm-upand training sets and between each set of repetitions wasprovided. The 1 RM was measured on each machineevery 2 weeks and the training load readjusted to keepthe training stimulus constant at 60% 1 RM. Subjectswere asked to execute the concentric phase of each exer-cise over 2 s followed by an eccentric phase over 3 s. Therange of motion was 50° on the sitting calf and 90° onthe leg press. At the end of the session, the subjects per-formed a 10-min cool-down including stretching of thetrained muscles. The total duration of the session wasapproximately 60 min.

Home programme

Once a week the training session was held at home usingelastic bands (Theraband) reproducing the movementstrained on the two variable-resistance machines. Subjectswere familiarised with the elastic bands and were givenclear instructions on how to reach the same intensity usedwith the machines. The degree of resistance of the elasticbands (grey Theraband) had previously been tested in ourlaboratory by calculating the correlation between the elon-gation (cm) and resistance (kg) of the elastic bands whenWxed loads were applied (Capodaglio 2002). It was there-fore possible to provide clear instructions on the magni-tude of the recommended exercise intensity (expressed incentimeters of elongation of the elastic band).

Each home session consisted of 20 repetitions for eachmovement. As for the aerobic work, subjects were encour-aged to perform some outdoor 30-min aerobic exerciseonce a week. Monthly logs were completed to documenttheir home activities and adherence to protocol.

537

Measurements

Muscle function

We measured maximum isometric strength of the kneeextensors (KE) and ankle plantar Xexors (PF) with aCybex Norm dynamometer.

Knee extensors strength was measured with the sub-jects in a sitting position with hip Xexed at 90° at Wveangles in the range 20°–90° (full knee extension 0°).

Plantar Xexors isometric torque was measured withthe participants lying prone with both hips and kneesextended (0°) at Wve angles in the range ¡ 20° (dorsiXex-ion) + 20° (plantarXexion).

Leg extensor power was measured with the Notting-ham Power Rig equipment. The equipment measures thepower (W) delivered by a seated subject in accelerating aXywheel from rest by pressing a footplate until the leg isextended (Bassey et al. 1990).

Functional abilities

We used the following tests at baseline (t1) and endtraining (t3): functional reach (FR), chair rise 1 (CR1)and 10 times (CR10), bed rise (BR), 6-min walking test(6MWT), stair climbing (SC), get up and go (GU&G),one-leg standing (1LS).

Functional reach (FR)

The subjects were asked to stand and reach forwardbeyond arm’s length as far as possible without takingtheir heels oV the Xoor (Duncan et al. 1990).The reachdistance was measured in centimetres from the initial tothe Wnal knuckles position. The longest out of threereaches was recorded.

Chair rise 1 (CR1) and 10 times (CR10)

The subjects were asked to rise as fast as they can withtheir arms folded from a seat at 0.42 m from the Xoor(Skelton et al. 1994). The fastest out of two rises wasrecorded with a 30-s stopwatch. Then the subjects wereasked to rise ten times consecutively and time of perfor-mance was recorded (Csuka et al. 1991).

Bed rise (BR)

The subjects were asked to climb down from a standard0.72 m-high hospital bed and to rise to an upright posi-tion as fast as they can. The fastest out of two rises wasrecorded with a 30-s stopwatch.

6-min walking test (6MWT)

The 6MWT (Butland 1982) measures the distance inmeters walked at a preferred pace in 6 min along a mea-sured walkway. Subjects were instructed to walk as far as

possible during the 6 min and were given encouragementthroughout the test.

Stair climbing (SC)

Subjects were asked to climb up a staircase (2 Xights of12 steps each) as rapidly as possible without stoppingand without using the handrail as support, to turnaround on the top platform and walk down. Performancewas recorded with a stopwatch (Csuka et al. 1991).

Get up and go (GU&G)

Subjects were asked to rise from a standard 42 cm-highchair, walk to a wall 3 m away, turn, and return to the chairto a seated position (Mathias et al. 1986; Podsiadlo et al.1991). The performance time was recorded with a stopwatch.

One-leg standing (1LS)

Participants were asked to stand on one leg as long aspossible with the contralateral knee Xexed at 90°. Thetest was over when the subjects were not able to maintainbalance and the suspended leg touched the Xoor. Theperformance time was recorded with a stopwatch (Vellaset al. 1997).

Life-style (PA and MDEE)

Both groups were assessed at baseline (t1; July 2002), at 6(t2; January 2003) and 12 months (t3; June 2003). Wemeasured energy expenditure in habitual PA (profes-sional, leisure, domestic, basic daily activities, activetransport) with the Paqap© questionnaire. This is aninterview-based, 7-day recall computerised questionnairethat elicits responses about PA habitually performed bythe individual during a typical week. Its validity hasalready been tested on elderly subjects (Bonnefoy 1996).SpeciWcally, it has been shown to provide a very reliablecalculation of MDEE and high correlations not onlywith AA3 but also with less intense activities (daily,household, shopping, sitting). PA is quantiWed by thesubject, who describes duration (min, hours), frequency(per day, week, month) and intensity of the activities. Atarget total time of 168 h is expected to be described,with a tolerance interval of § 15%. For each type ofactivity, the reference energetic intensity (McArdle et al.1982; Astrand 1986) is multiplied by the duration. Theresults are expressed as mean daily energy expenditure(MDEE), aerobic activities greater than 3 MET (AA3)and PA intensity classes. Intensity classes are distin-guished as follow: Class 1 = 1–3.9 MET; Class 2 = 4–5.9 MET; Class 3 = 6–7.9 MET; Class 4 = 8–9.9 MET;Class 5 > 10 MET; where MET = 3.5 ml O2 kg¡1min¡1.

Statistical analysis

We used a two factor ANOVA (gender and training) toanalyse diVerences in strength values at baseline between

538

T and C in both M and F. Results are expressed as meanvalues § standard deviation.

A two factor ANOVA (gender and training) forrepeated measures were used to analyse pre-post trainingdiVerences in T and C and diVerences following trainingaccording to gender.

The relationship between the time to complete the FAtests and the strength and power variables was deter-mined using Pearson product moment correlations.Level of signiWcance was set at P < 0.05.

Physical activity data for T and C and for M and Fwere compared using the ANOVA–MANOVA and thepost-hoc comparison test (F or Rao’s R values arereported). The intra-group diVerences between the testswere assessed with the t test for dependent samples. Cor-relation between the variables was assessed with thePearson correlation coeYcient (r). The level of signiW-cance was stated at P < 0.05. The program used was Sta-tistics for Windows, ver. 5.0.

Results

The overall adherence rate to the programme (number ofsession actually performed divided by the total numberof home and hospital-based sessions) was 65% (range:12–89% in M and 17–89% in F). At baseline, T and C didnot diVer in age, weight, height, body mass index (BMI)and AA3 time (Table 4).

Muscle function

At baseline F showed signiWcantly lower mean KE val-ues (¡47%) and mean PF values (¡59%) than M.

SigniWcant diVerences in PF were also evidentbetween T and C at baseline (Table 1).

Mean values and standard deviations of all theparameters are shown in Table 2. We observed signiW-cant pre-post training diVerences between T and C in allof the parameters (Table 3). The eVect of trainingshowed signiWcant gender diVerences, in particular, FTshowed signiWcantly higher gains than MT (Table 3).

Functional abilities

Mean values and standard deviations of all the parametersare shown in Table 2. The eVect of training on functionalabilities did not show gender diVerences and signiWcantgains were observed both in FT and MT (Table 3).

Leg extensor power showed higher correlations thanKE–PF with GU&G, SC, CR10, 6MWT.

The 1LS and CR1 tests showed higher correlationswith KE and PF than LEP.

PA and MDEE

There were signiWcant diVerences between M and F inweight, height, BMI and AA3 time but not in MDEEand age (Table 4).

Table 1 Baseline KE and PF data (mean § standard deviations):non-signiWcant baseline diVerences were found for KE data betweentraining and control group

SigniWcant diVerences in PF values (P < 0.05) are evident betweentraining and control group in the female (*) and male (**) subgroup

KE right (Nm)

KE left (Nm)

PF left (Nm)

PF right (Nm)

F T 95.3 § 21.8 93.6 § 16.4 57.2 § 11.1* 67.2 § 15.8*C 68.1 § 22.4 66.0 § 166 42.6 § 17.0* 41.3 § 16.8*

M T 148.9 § 23.5 143.8 § 21.7 104.4 § 22.6 111.6 § 14.0**C 136.1 § 51.3 150.4 § 28.6 86.1 § 24.4 84.9 § 21.6**

Table 2 Meanvalues § standard deviations ofpre-post training in muscle func-tion (PF, KE, LEP), functionalabilities (FR, 1LS, CR1, CR10,BR, GU&G, SC, 6MWT) in FT,FC, MT and MC

FT FC MT MC

PF¡20° Pre 52.9 § 10.4 46.2 § 19.3 99.7 § 26.9 83.8 § 19.6Post 68.0 § 21.1 42.3 § 15.9 119.6 § 30.4 80.1 § 14.6

KE 90° Pre 79.3 § 9.3 64.3 § 10.4 124.2 § 21.2 108.3 § 48.7Post 89.8 § 1.8 61.5 § 7.5 136.9 § 25.4 99.8 § 33.4

LEP Pre 62.7 § 12.4 41.7 § 2.9 118.4 § 30.9 113.0 § 26.6Post 74.5 § 15.6 36.5 § 3.1 126.0 § 27.2 114.6 § 21.7

FR Pre 10.6 § 2.2 12.5 § 2.9 16.3 § 4.0 12.6 § 4.2Post 20.5 § 5.6 12.1 § 3.1 24.6 § 4.7 16.0 § 6.3

1LS Pre 12.0 § 7.5 15.3 § 16.2 23.0 § 40.5 14.6 § 12.4Post 15.7 § 9.5 13.7 § 14.2 30.6 § 36.2 12.6 § 10.4

CR1 Pre 1.1 § 0.3 1.5 § 0.4 1.1 § 0.3 1.5 § 0.5Post 0.8 § 0.3 1.7 § 0.4 0.9 § 0.3 1.5 § 0.6

CR10 Pre 28.5 § 4.6 33.0 § 6.1 36.3 § 16.9 36.9 § 31.0Post 20.7 § 4.5 35.8 § 4.8 22.2 § 2.1 38.7 § 31.9

BR Pre 3.4 § 0.7 4.8 § 2.7 2.6 § 0.6 3.3 § 2.5Post 2.3 § 0.5 4.9 § 2.3 2.3 § 0.4 4.0 § 2.4

GU&G Pre 7.4 § 0.7 8.3 § 0.5 6.7 § 0.5 6.8 § 0.9 Post 5.9 § 0.8 8.4 § 0.7 5.4 § 0.7 7.1 § 0.6

SC Pre 36.2 § 5.7 38.8 § 11.5 34.2 § 4.8 32.7 § 10.1Post 31.9 § 6.0 41.3 § 11.2 30.1 § 5.8 33.0 § 8.0

6MWT Pre 461.1 § 50.1 347.1 § 60.3 468.0 § 108.7 404.0 § 132.9Post 481.4 § 55.7 337.1 § 63.8 491.6 § 115.6 407.5 § 123.2

539

The overall MDEE in our subjects was30.7 Kcal kg¡1 day¡1 (2,036 Kcal kg¡1 day¡1).

In all the subjects, we observed a high inter-subjectvariability in PA distribution, relative to light- (walking,recreational), medium- (cycling, gardening) and high-(sports) intensity activities. The variability was especiallyhigh between genders. In all three tests, M (T + C)devoted an average higher amount of time in AA3 thanF (T + C): hiking, gardening (17 h a week vs. 7 h a week,Rao’s R = 2.7) with higher related EE (3,900 vs.875 Kcal week¡1, Rao’s R = 8.91); active transferring(cycling, walking) (9.8 vs. 6 h week¡1, Rao’s R = 4.3) andrelated EE (1,943 vs. 718 Kcal week¡1, Rao’s R = 8.39).

Female (T + C) was more involved in lower intensityhousehold activities than M (T + C) (24 vs. 10 h week¡1,Rao’s R = 7.02), with a diVerent relative EE (2,810 vs.1,470 Kcal week¡1, Rao’s R = 5.21).

Consequently, AA3 time is signiWcantly diVerent in alltests between M and F, both in T and C (Table 4).MDEE increased by 10% only in MT (t=2.62) (Table 4).

At t3, AA3 time increased 60% in T (13.4 vs. 8.4 hweek¡1, t = 2.45): 51% in MT (t = 2.41), and 41% inFT (2.14) (Table 4). No increases in AA3 were observedin C.

At t3, Class 2 PA time signiWcantly increased both inMT (+ 146%) and FT (+ 18%) (Fig. 1).

The changes in life-style are described in Fig. 1 andTable 5.

Table 3 Percent increases in muscle function (KE, PF, LEP), functional abilities (FR, 1LS, CR1, CR10, BR, GU&G, SC, 6MWT) in FT,MT and FC, MC

The signiWcance of pre-post training diVerences between T and C (pre-post T–C) are reported. The signiWcance of the diVerent eVect of train-ing according to gender is reported in the last column and, in brackets, the signiWcances in F and M

FT FC MT MC Pre-post T–C Gender

PF¡20° + 21.5% ¡5% + 8% ¡4% P < 0.001 P = 0.023 (F P = 0.001, M P = 0.17)KE 90° + 12% ¡2% + 4% ¡5% P < 0.001 P = 0.006 (F P = 0.019, M P = 0.04)LEP + 22.5% ¡9% + 4% ¡8% P < 0.001 P = 0.006 (F P = 0.001, M P = 0.04)FR + 85% ¡1.7% + 60% + 1% P < 0.001 NS (F P < 0.001, M P = 0.003)1LS + 29.1% ¡11% + 25% ¡4.8% P < 0.001 NS (F P = 0.007, M P = 0.017)CR1 + 29% ¡11% + 17% ¡2% P < 0.001 NS (F P = 0.002, M P = 0.052)CR10 + 28% ¡8% + 20% ¡7% P < 0.001 NS (F P < 0.001, M P = 0.018)BR + 29% ¡4% + 8% ¡3% P < 0.001 NS (F P < 0.001, M P = 0.02)GU&G + 20.5% ¡0.6% + 18.6% ¡2% P < 0.001 NS (F P < 0.001, M P < 0.001)SC + 12% ¡3.3% + 12% ¡2% P < 0.001 NS (F P < 0.001, M P = 0.034)6MWT + 4.7 ¡2.7% + 4.5% ¡2% P < 0.001 NS (F P = 0.016, M P = 0.026)

Table 4 Characteristics of the38 subjects (19 M, 19 F), andtime devoted to AA3 as esti-mated by Paqap in the threetests

T C

M+F M F M+F M F

N 23 11 12 15 8 7Age 76.6 § 3.8 76.6 § 3.0 76.6 § 4.6 77.7 § 3.1 76.6 § 3.5 79.0 § 2.1Weight (kg) 66.2 § 12.8 75.2 § 9.4* 58.0 § 9.8* 70.4 § 10.8 77.5 § 5.1* 62.3 § 10.0*Height (cm) 166.4 § 7.8 172.4 § 5.2* 160.9 § 5.2* 167.6 § 9.9 175.6 § 4.8* 158.5 § 4.8*BMI 23.7 § .0 25.2 § 2.2* 22.3 § 3.1* 24.9 § 2.4 25.1 § 1.4 24.7 § 3.3AA3 Time

(h week¡1)t1 8.4 § 6.8^ 9.4 § 5.2* ^ 1.7 § 0.9* 8.7 § 6.8 9.7 § 7.4* 1.8 § 1.3*t2 11.0 § 8.3 11.7 § 6.4* 1.9 § 1.3* 7.1 § 5.4 8.2 § 5.6* 1.2 § 1.0*t3 13.4 § 7.3^ 14.2 § 5.5*^ 2.4 § 0.4* 7.5 § 6.3 8.8 § 6.4* 1.4 § 0.6*MDEE

(Kcal kg¡1 day¡1)t1 31.9 § 6.2 31.5 § 6.9 28.1 § 5.2 28.0 § 4.8 28.7 § 5.4 27.8 § 5.6t2 30.7 § 6.1 29.0 § 4.9^ 28.2 § 7.1 29.0 § 5.9 30.1 § 6.6 27.2( 5.6t3 32.8 § 5.5 34.1 § 4.6^ 29.4 § 4.2 30.1 § 4.3 31.3 § 4.2 28.5 § 4.4

SigniWcant diVerences (P < 0.05)between M and F (*), and be-tween t1 and t3 (^) are described

Fig. 1 Changes in life-style after the 1-year training. At t3, MTshowed a 146%increase (t = 2.82) in Class 2 (4–5.9 MET, white col-umn) time, and a 20%increase in Class 3 (6–7.9 MET, black column)time. FT increased Class 2 PAtime by 16% (t = 2.23).

t10

2

4

6

8

10

12

14

16

MT FT MC FC

ho

urs

/ w

eek

t2 t3 t1 t2 t3 t1 t2 t3 t1 t2 t3

*

*

540

Discussion

In the developed countries, 30–80% of the elderly popu-lation is physically inactive (Sports Council and HealthEducation Authority 1992). Our data showed a slightdeterioration in the control group values over the year.This emphasises the importance of maintaining exerciseprograms since inactive elderly are at high risk of experi-encing a decline in function that can lead to loss of inde-pendence.

The majority of the elderly in our study were active:they spent nearly 7% of the weekly time and more than14% of MDEE in light-to-moderate activities. They vol-unteered to the programme and were selected accordingto the health criteria.

In line with the previous reports (Krems et al. 2004;Berthouze et al. 1995; Liu et al. 2001), the preferredactivities were walking, cycling, gardening and sports formen and house-holding for females.

At baseline, F were 47–59% weaker then M in KE andPF, respectively, which is in line with the previous data(Aniansson 1983). Anthropometric characteristics andstrength values of our subjects were comparable to theage-matched data of Skelton et al. (1994). Our trainingintensity (60% 1 RM) improved muscle function andfunctional abilities both in MT and FT, but FT beneWtedmore from the programme. The adherence rate was com-parable in FT and MT; attendance values were scatteredbut only very few M and F subjects had very low atten-dance values (12–17%).

Initial level of frailty seems therefore to aVect theimpact of strengthening.

Another possible reason for lower gains in musclefunction in MT could be that their age values were morescattered around the mean value then in FT. It is knownin fact that both strength and power in men are notice-ably age-related: KE declines by 4.5 Nm per annumwhile LEP decreases by 5.4 W per annum (Skeltonet al.1994; Häkkinen et al. 1984).

Other studies have reported gender diVerences inresponse to training (Malbut et al. 2002; Aniansson et al.1983; Roth et al. 2001). Roth (2001) has shown signiW-cant muscle cross-sectional area increases in females over70 but not in men after a 6-month training. Vincent et al.(2002) have reported signiWcant functional changeseither after training at 80 and 60% of the 1 RM in 60- to83-year-old males and females.

In our study, functional abilities signiWcantlyimproved by 12 and 20% in MT and FT, respectively.This can be explained by the higher muscle functionincreases in FT, and in particular by task-independentLEP increases, together with enhanced movement coor-dination, motor control and higher velocity of execution.LEP appears to be a crucial determinant of performancefor females during daily functional activities.

The potential eVect of age on increase of performancefollowing training was not speciWcally addressed in ourstudy due to the small age range of the subjects.

The signiWcant improvement in functional abilities,PA time and specially medium-intensity activities reXectsthe eYcacy of the long-term programme in changing lifestyle in T (M + F).

Both AA3 and Class 2 PA time incremented at t3 inMT (+ 51 and + 146%, respectively) and in FT (+ 41 and+ 18%, respectively), due to the active involvement in thegym-session and to other life activities. Although posi-tive, these results indicate a more diYcult involvement ofFT in medium-high intensity PA (Table 5), probably dueto a variety of social and family reasons.

At t3, 46% of C group performed only Class 1 activi-ties.

Male trained performed a higher amount of PA time,reXected by a 10% MDEE increase at t3 (t = 2.62),whereas FT did not show this eVect.

The overall MDEE in our subjects is similar to the2,135 Kcal day¡1 value reported by Bonnefoy (1996) foractive subjects aged 64–84 years. MDEE in elderly wassigniWcantly lower (P < 0.001), than that reported foryounger very active subjects (3,699 Kcal day ¡1) previ-ously tested with Paqap© (Capodaglio et al. 2001),whereas PA time was similar (14.4 vs. 13.4 h week¡1).Other authors (Krems et al. 2004; Liu et al. 2001) havepreviously reported that PA levels (time) are similar inelderly and younger subjects, although with diVerent pat-terns.

Our data, in particular the high inter-individual vari-ability and the relevant diVerence between gender inAA3 time (Table 4), show that seasonal inXuencesappear to have inXuenced the aerobic and open-air activ-ities performed at t1, t2, t3.

Environmental factors (temperature and humidity),type of mobility (public transport, bicycle, walking) andavailability of facilities might also have accounted forsome variations in quantity and quality of PA reportedin the three tests.

We found the Paqap© questionnaire to be a usefultool for assessing PA and MDEE in the elderly popula-tion because of its detailed approach to measurement of

Table 5 Number of the subjects involved in Class 1, 2 and 3

At t3, 7 FT vs. 1 FC were involved in Class 2 activities; at t3, 12 T(52%) were engaged in Class 2, and 7 T (30%) in Class 3 activities,while most of the C subjects were engaged in Class 1 activities (norelevant PA). However, at t3, ¡25% FT were involved in Class 3 andClass 2, while + 18% MT were involved in Class 3

MT FT

t1 t3 t1 t3

Class 1 3 1 0 3Class 2 5 5 9 7Class 3 3 5 3 2

MC FCt1 t3 t1 t3

Class 1 4 3 1 4Class 2 2 1 5 1Class 3 2 1 1 0

541

daily activities including low-medium intensity ones,which are prevalent in elderly.

Consistent with prior investigations evaluatingstrength and mobility tasks (Ferrucci et al. 1997; Keysoret al. 2003), we found that the relationship between MFand FA were best characterised by the equationy = a · log x + b. The curvilinear trend of the functionsuggests that other factors than muscle functionimprovements contribute to the results of the functionaltests. In the 6MWT, the individual maximal aerobicpower should be considered among the determinants,and in other tests involving the rapid execution of com-plex movements various neuromuscular factors thatwere not under investigation in our study.

Leg extensor power appears to be highly correlatedwith GU&G, SC, CR10 and 6MWT, describing more ofthe variance than does isometric strength (KE and PF) inthese tasks. On the other hand, short-term performancemeasures, such as 1LS, BR and CR1, showed higher cor-relations with KE-PF rather than LEP. This is probablydue to the fact that strength per se rather than the rate atwhich force is generated is important in these short tasks.Short-term performances also showed higher inter-sub-ject variability than the other functional tests.

These results are in line with the conclusions of the InCHIANTI study (Bean et al. 2003), which used the sameequipment to measure LEP. Validity and reliability ofthe method have been shown and signiWcant correlationshave been found with other measures of explosive power(Bean et al. 2002). The Power Rig appears to be a safermethod for the elderly and provides a relevant functionalassessment, since the angles of the hip, knee and ankleduring the push are somewhat similar to those occurringduring daily tasks. We used average LEP values fromboth legs, since they are used simultaneously in chair ris-ing or walking and did not analyse right-left asymmetry.Since both legs are used to move the body mass againstthe force of gravity, it seemed most appropriate to relatethe performance measures to the ratio of LEP (W) tobody mass (kg).

Female trained and MT showed a signiWcant improve-ment in 6MWT, which is a measure of endurance not spe-ciWcally treated by our programme. In this test, subjectsdo not reach the maximal capacity: the intensity is self-selected and they are allowed to stop and rest during thetest. Therefore, the 6MWT appears an adequate means toreXect the demands of daily tasks. In our study, the base-line 6MWT values for M were higher than the referencevalues (Enright et al. 1998), whereas F was 20% lower.

The biases in our study are represented by the non-randomised allocation of the subjects into training andcontrol and by the baseline signiWcant diVerences in PFstrength between T and C. In general, the recruitment ofelderly groups matched in terms of PA and Wtnessappears quite diYcult to achieve. In our study a sponta-neous enrollment of the subjects in the T and C groupoccurred, depending on the personal motivation, diYcul-ties reported in reaching the training facility on a regularbasis, level of Wtness or subjective perception of physical

status. Therefore, it could not be excluded that the moreWt subjects formed the T group.

Conclusions

This study reinforces the idea that long-term group inter-ventions are eVective in promoting healthy PA levels andmaintaining the eVects by favouring a more active life-style in elderly subjects over 70 years of age. In particularF, who may reach levels below the muscle power thresh-old for daily tasks and are at higher risk of fracture, ben-eWt from the intervention program. Late-life PA, and inparticular walking, increases muscle strength and aerobiccapacity and reduces functional limitations. Therefore,promoting a more active life-style in the elderly couldminimise disability reducing the risk of chronic diseases(Keysor et al. 2003). Participation in low-to-moderateintensity activities may be beneWcial for many people,even if such participation does not produce major gainsin cardio-respiratory Wtness. Participation to exerciseclasses, as proposed in our intervention program, proba-bly enhances the beneWts of PA, facilitating social inter-action and prompting the acquisition of an active life-style. Large muscle rhythmic aerobic forms of exercise(walking, running, swimming, cycling), which were anintegral part of the early years of most adults’ lives couldbe assumed as regularly performed physical activities.Programmes for promoting an active life-style andimproving health condition in old age may well derivesome beneWts from this and similar future studies.

Changes in life-style after the 1-year training. At t3,MT showed a 146% increase (t = 2.82) in Class 2 (4–5.9MET, white column) time, and a 20% increase in Class 3(6–7.9 MET, black column) time. FT increased Class 2PA time by 16% (t = 2.23).

Acknowledgments The support given by Technogym Italy isacknowledged. The authors are indebted to the participants in thisstudy for the commitment given to this project.

References

ACSM (1998) The recommended quantity and quality of exercisefor developing and mantaining cardiorespiratory and muscularWtness, and Xexibility in healthy adults. Med Sci Sports Exerc30(6):975–991

Aniansson A, Sperling L, Rundgren A, Lehnberg E (1983) Musclefunction in 75-year-old men and women. A longitudinal study.Scand J Rehabil Med suppl 9:92–102

Åstrand PO, Rodahl K (1986) Textbook of work physiology.McGraw-Hill, New York

Bassey EJ, Ramsdale SJ (1993) Leg extensor power improves inwomen with feasible exercise programmes. J Physiol 467:121P

Bassey EJ, Short AH (1990) A new method for measuring poweroutput in a single leg extension: feasibility, reliability and valid-ity. Eur J Appl Physiol 60:385–390

Bassey EJ, Fiatarone MA, O’Neill EF, Kelly M, Evans WJ, LipsitzLA (1992) Leg extensor power and functional performance invery old men and women. Clin Sci 82:321–327

542

Bean J, Kiely DK, Herman S et al (2002) The relationship betweenleg power and physical performance in mobility-limited elders. JAm Geriatr Soc 50:461–467

Bean J, Leveille SG, Kiely DK, Bandinelli S, Guralnik JM, FerrucciL (2003) A comparison of leg power and leg strength within theIn CHIANTI study: which inXuences mobility more? J Gerontol58A(8):728–739

Berthouze SE et al (1995) Relationship between mean habitual dailyenergy expenditure and maximal oxygen uptake. Med Sci SportsExerc 27(8):1170–1179

Bonnefoy M, Kostka T, Berthouze SE, Lacour JR (1996) Validationof a physical activity questionnaire in the elderly. J Appl Physiol74:528–533

Butland RJA, Pang J, Gross ER, Woodcock AA, Geddes DM (1982)Two-, six, and 12-minute walking tests in respiratory disease.BMJ 284:1607–1608

Capodaglio EM, Aranda S, Lanfranconi F (2001) Primi risultati diuno studio per la validazione in lingua italiana di un questionar-io informatizzato di attivita’ Wsica (Paqap). In: Atti del Con-gresso Nazionale Specialisti in Medicina dello Sport; June 24–27; Chieti, Italy

Capodaglio P, Facioli M, Burroni E, Giordano A, Ferri A, ScaglioniG (2002) EVectiveness of a home-based strengthening programfor elderly males in Italy. A preliminary study. Aging Clin ExpRes 14:28–34

Carmeli E, Orbach P, Lowenthal DT, Merrick J, Coleman R (2003)Long-term eVects of activity status in the elderly on cardiorespi-ratory capacity, blood pressure, blood lipids, and body composi-tion: a Wve-year follow-up study. Sci World J 3:751–67

Cress ME, Thomas DP; Johnson J, Kasch FW, Cassens RG, SmithEL, Agre JC (1991) EVect of training on VO2 max, thighstrength, and muscle morphology in septuagenarian women.Med Sci Sports Exerc 23:752–78

Csuka M, McCarthy DJ (1991) Simple method for measurement oflower extremity muscle strength. Am J Med 78:77–81

Duncan PW, Weiner DK, Chandler J et al (1990) Functional reach:a new clinical measure of balance. J Gerontol (Med Sci)45:M192–M197

Earles DR, Judge JO, Gunnarsson OT (2001) Velocity training in-duces power-speciWc adaptations in highly functioning olderadults. Arch Phys Med Rehabil 82(7):872–878

Enright PL, Sherrill DL (1998) Reference equations for the 6-minutewalk in healthy adults. Am J Respir Crit Care Med 158:1384–1387

Ferrucci L, Guralnik JM, Buchner D et al (1997) Departures from lin-earity in the relationship between measures of muscular strengthand physical performance of the lower extremities: the women’shealth and ageing study. J Gerontol Med Sci 52:M275–M285

Foldavari M, Clark M, Laviolette MJA et al (1999) Association ofmuscle power with functional status in community dwelling eld-erly women. Med Sci Sports Exerc 31:S378

Häkkinen K, Kallinen M, Izquierdo M et al (1984) Changes in ago-nist-antagonist EMG, muscle CSA, and force during strengthtraining in middle-aged and older people. J Appl Physiol1198:1341–1349

Jackson AS, Beard EF, Wier LT, Ross RM, Stuteville JE, Blair SN(1995) Changes in aerobic power of men ages 25–70 year. MedSci Sports Exerc 27:113–120

Jozsi AC, Campbell WW, Joseph L, Davey SL, Evans WJ (1999)Changes in power output with resistance training in older andyounger men and women. J Gerontol A Biol Sci Med Sci54:M591–M596

Keysor JJ (2003) Does late-life physical activity or exercise preventsor minimize disablement? Am J Prev Med 25:129–136

Krems C., Lührmann P. M., Neuhäuser-Berthold M (2004) Physicalactivity in young and elderly subjects. J Sport Med Phys Fitness44(1):71–78

Latham NK, Bennet DA, Stretton CM, Anderson CS (2004) Sys-tematic review of progressive resistance strength training in olderadults. J Gerontol Med Sci 59:48–61

Macaluso A, Young A, Gibb KS, Rowe DA, De Vito G (2003) Cy-cling as a novel approach to resistance training increases musclestrength, power and selected functional abilities in healthy olderwomen. J Appl Physiol 95(6):2544–2553

Malbut KF, Greig CA, Pippet B et al (1993) Training exercise in-creases strength and power in very elderly women. Clin Sci81(29):16P

Malbut KE, Dinan S, Young A (2002) Aerobic training in the ‘oldestold’: the eVect of 24 weeks of training. Age Ageing 31:255–260

Mathias S, Nayak USL, Isaacs B (1986) Balance in elderly patients:the ‘Get up and go’ test. Arch Phys Med Rehabil 67:387–389

McArdle WD, Katch FI, Katch VL (1982) Exercise physiology: en-ergy, nutrition, and human performance. Lea and Febiger, Phil-adelphia

McMurdo MET, Johnstone R (1995) A randomized controlled trialof a home exercise programme for elderly people with poormobility. Age Ageing 24:425–428

Liu B, Woo J, Tang N, Ng K, Ip R, Yu A (2001) Assessment of totalenergy expenditure in a Chinese population by a physical activityquestionnaire: examination of validity. Int J Food Sci Nutr52(3):269–282

Podsiadlo D, Richardson S (1991) The timed ‘Up & Go’: a test of ba-sic functional mobility for frail elderly persons. J Am Geriatr Soc39:142–148

Pyka G, Lindenberger E, Charette S, Marcus R (1994) Musclestrength and Wber adaptations to a year-long resistance trainingprogram in elderly men and women. J Gerontol 49:M22–M27

Roth SM, Ivey FM. Martel GF et al (2001) Muscle size responses instrength training in young and older men and women. J Am Ge-riatr Soc 49:1428–1433

Shephard RJ (1991) Fitness and aging. In: Blais C (eds) Aging in thetwenty-Wrst century. Captus University, Downsview

Skelton DA, Grieg CA, Davies JM et al (1994) Strength, power andrelated functional ability of healthy people aged 65–89 years.Age Ageing 23:371–377

Sports Council and Health Education Authority (1992) Allied Du-bar National Fitness Survey: a report on activity patterns and Wt-ness levels. Sports Council and Health Education Authority,London

Suzuki T, Bean J, Fielding R (2001) Muscle strength and power ofthe ankle plantar and dorsi Xexors predict functional perfor-mance in community dwelling older women. J Am Geriatr Soc49:1161–1167

van der Bij A, Laurant GHM, Wensing M (2002) EVectiveness ofphysical activity interventions for older adults. Am J Prev Med22(2):120–133

Vellas BJ, Rubenstein LZ, Ousset PJ, Faisant C, Kostek V, Nourhas-hemi F, Allard M, Albarede JL (1997) One-leg standing balanceand functional status in a population of 512 community-livingelderly persons. Aging Clin Exp Res 9:95–98

Vincent KR, Braith RW, Feldman RA, Magyari PM, Cutler RB,Persin SA, Lennon SL, Gabr AH, Lowenthal DT (2002) Resis-tance exercise and physical performance in adults aged 60 to 83.J Am Geriatr Soc 50:1081–1100