Reproducibility and validity of the International Physical Activity Questionnaire (IPAQ) in elderly men

Abstracts

É necessário encontrar meios para quantificar de maneira eficiente e econômica o nível de atividade física da população. Os questionários são formas viáveis e econômicas, embora seja discutível a fidedignidade dessas medidas. Este estudo objetivou determinar a reprodutibilidade e a validade do Questionário Internacional de Atividade Física (IPAQ) em homens idosos. A amostra foi composta por 29 homens idosos acima de 60 anos, com média de idade de 66,6 anos (DP = 4,3), participantes do programa de extensão da Universidade Federal de Santa Catarina. Os instrumentos utilizados foram: a) IPAQ, forma longa; b) pedômetro; e c) Diário de Atividade Física de Bouchard (DAF). Para a reprodutibilidade foram realizadas duas aplicações do IPAQ, com intervalo de 21 dias (r s = 0,95). A análise estatística adotada foi a correlação de Spearman (r s), o percentual de concordância (%C), o índice kappa (k) e a plotagem de Bland e Altman. A amostra foi dividida, utilizando-se como critério a mediana. A reprodutibilidade apresentou correlação de r s = 0,95. A associação entre o IPAQ e o DAF foi de: r s = 0,38; %C = 69 e k = 0,04 e a associação entre o IPAQ e o pedômetro de: r s = 0,24; %C = 62 e k = 0,19. Concluiu-se que a validade variou de moderada a baixa, enquanto a reprodutibilidade foi adequada.

Medida; Prática geral; Vigilância; IPAQ

ORIGINAL ARTICLE

Reproducibility and validity of the International Physical Activity Questionnaire (IPAQ) in elderly men

Reproducibilidad y validez del Cuestionario Internacional de Actividad Física (IPAQ) en hombres ancianos

Tânia R. Bertoldo Benedetti^I; Priscilla de Cesaro Antunes^{I, III}; Ciro Romélio Rodriguez-Añez^II; Giovana Zarpellon Mazo^IV; Édio Luiz Petroski^I

^ICentro de Desportos, Universidade Federal de Santa Catarina, Florianópolis, SC, Brasil

^IIPontifícia Universidade Católica do Paraná, Curitiba, PR

^IIIBolsista do CNPq

^IVCentro de Educação Física, Fisioterapia e Desportos, Universidade do Estado de Santa Catarina, Florianópolis, SC, Brasil

^{Correspondence to}

ABSTRACT

It is necessary to efficiently and economically find means to quantify the level of physical activity of the population. Questionnaires are viable and economical; however, their trustworthiness is questionable. This study aimed to determine the reproducibility and validity of the International Physical Activity Questionnaire (IPAQ) in elderly men. The sampling was composed of 29 elderly men over the age of 60, average age of 66.6 years (SD = 4.3), who were members of the extension program at the Federal University of Santa Catarina. The instruments used were: a) IPAQ extended version; b) pedometer and c) Bouchard's Physical Activity Diary (PAD). For reproducibility, the IPAQ was applied twice with a 21 day interval (r_s = 0.95). The statistical analysis used was Spearman Correlation (r_s); Concordance Percentile (%C); Kappa Index (k); and Bland and Altman plot. The sample was divided using the median as criterion. The reproducibility showed a correlation of r_s 0.95. The combination between IPAQ and PAD was: r_s = 0.38; %C = 69 and k = 0.04. The combination between IPAQ and the Pedometer was: r_s = 0.24; %C = 62 and k = 0.19. It was concluded that the validity varied from moderate to low while the reproducibility was adequate.

Keywords: Measurement. General practice. Surveillance. IPAQ.

RESUMEN

Es necesario encontrar medios para cuantificar de manera eficiente y económica el nivel de actividad física de la población. Los cuestionarios son formas viables y económicas, aunque se discuta la fidedignidad de estas medidas. Este estudio ha tenido como objetivo determinar la reproducibilidad y la validez del Cuestionario Internacional de Actividade Física (IPAQ) en hombres ancianos. La muestra se compuso de 29 hombres ancianos con más de 60 años, con edad media de 66,6 años (DP = 4,3), participantes del programa de extensión de la Universidad Federal de Santa Catarina. Los instrumentos utilizados fueron: a) IPAQ, forma larga; b) pedómetro y c) Diario de Actividad Física de Bouchard (DAF). Para la reproducibilidad fueron realizadas dos aplicaciones de IPAQ, con intervalo de 21 días (r_s = 0,95). El análisis estadístico adoptada fue el de correlación de Spearman (r_s), el porcentaje de concordancia (%C), el índice kappa (k) y el "plotaje" de Bland y Altman. La muestra fue dividida, utilizando como criterio la mediana. La reproducibilidad presentó una correlación de r_s = 0,95. La asociación entre el IPAQ y el DAF fue de: r_s = 0,38; %C = 69 y k = 0,04 y la asociación entre el IPAQ y el pedómetro de: r_s = 0,24; %C = 62 y k = 0,19. Se concluye que la validez varió de moderado a bajo, mientras que la reproducibilidad fue adecuada.

Palabras-clave: Medida. Práctica general. Vigilancia. IPAQ.

INTRODUCTION

Regular physical activity reduces the risk of mortality and morbidity, regardless of other alterations in lifestyle^(1-4).

According to the Ministry of Health, in Brazil circulatory diseases are the main causes of death (32%)⁽⁵⁾. Physical inactivity is one of the important risk factors that lead to chronic-degenerative diseases. 83% of the Brazilian population does not practice any kind of physical activity. Besides that, above 54 years of age, 38,7% of women have circumference above 88 cm and 15,6% of men, above 102 cm, means which are considered risk factors for the development of cardiovascular diseases⁽⁶⁾. The disuse syndrome (physical inactivity) caused by the lack of movement and body exercises tremendously increases the search for medical assistance. Moreover, a dependence on other people relationship seems to be early developing in the elderly⁽⁷⁾.

A growing need of strategies and actions in order to promote collective health which includes physical activity and means to quantify it has been observed. Physical activity is an important component of a healthy lifestyle, especially by the evidence of its several benefits. The practice of physical activities is associated with the decrease of cardiovascular diseases and diabetes, among others⁽³⁾. Paffenbarger et al.⁽³⁾ observed a dose/response relationship in which the risk for these diseases progressively decreased when the total energetic cost in weekly physical activity increased; being considered optimum between 1500 and 3500 Kcal; although its benefits may be observed from 500 Kcal/week. Nonetheless, it is challenging to measure the total energetic cost of physical activities, which makes the implementation of suitable programs difficult.

Therefore, the World Health Organization (WHO); the American Center for Disease Control and Prevention (CDC) and the Karolinska Institute in Sweden, joined researchers with the purpose to develop and test an instrument which would obtain measurements of physical activities internationally comparable. It was proposed with this aim the International Physical Activity Questionnaire – IPAQ, validated in 12 countries and 14 research centers. The IPAQ is a questionnaire which allows the estimation of weekly time spent on physical activities of moderate and strong intensity, in different contexts of the daily life, such as: work; transportation; housework and leisure besides the time spent in passive activities performed on the seating position. The questionnaire was published in its short and long versions The IPAQ short version consists of seven open questions and its information allow to estimate the time spent per week in different dimensions of physical activity (walks and physical exertions of moderate and vigorous intensities) and of physical inactivity (seating position). The IPAQ long version presents 27 questions related with physical activities, performed in an ordinary week, with vigorous, moderate and mild intensities, with minimum duration of 10 continuous minutes, distributed in four dimensions of physical activity (work; transportation; housework and leisure) and of the time spent per week in the seating position. When the IPAQ short and long versions were compared, the results are different⁽⁸⁾. In the short version, the weekly time spent in moderate and vigorous physical activities has been shorter than in the long version. Such fact may be explained by the difference in the number of domains in each version and the number of questions, since in the long version each domain is more explored.

In Brazil, the IPAQ has been tested by several researchers^(9-13) concerning reproducibility (test/retest) as well as concomitant validity. Generally, the results derived from these studies show that the IPAQ (usual week, self-administered through individual interview) is an instrument with good stability of measurements and acceptable accuracy for use in epidemiological studies with young adults, middle aged- adults and older women.

However, little is known about the validity of utilization of this instrument with children and older men. Mazo et al.⁽¹⁴⁾ highlight the difficulty in obtaining measurements of the physical activity in older individuals, especially in studies which involve large number of subjects. In this situation the questionnaires seem to be the option with greatest viability. One of the instruments tested by the authors was the Questionnaire by Baecke, modified for older subjects, which presented good stability among measurements of reproducibility. However, the concomitant validity level was only modest (when correlated with the physical activities file (DAF) and the pedometer). Similar data were found in the application of the IPAQ Questionnaire applied for older women⁽¹²⁾.

The literature presents a gap concerning valid instruments for the measurement of physical activity in older men. Therefore, the present study had as aim to verify the reproducibility (test/retest) as well as the concomitant validity of the Physical Activity Questionnaire (IPAQ version 8, long version for usual/ordinary week) in the evaluation of the physical activities level of older men.

METHODS

This study is part of the project "The Elderly Profile of Florianópolis County" and was approved by the Human Ethics Committee of UFSC, in 30/07/2001 (project 051/2001). All volunteers signed a free and clarified consent form.

The sample consisted of 29 older men, members of the extension project "Physical Activities for the Third Age" from the Federal University of Santa Catarina (UFSC).

For this study, a long version, ordinary/usual week IPAQ questionnaire adapted by Benedetti et al. was designed⁽¹²⁾ in order to facilitate the understanding on older population as well as the register of the calculation of the number of hours, according to ^{appendix 1}. Thus, examples of activities that are common to Brazilian older subjects, specifically in Florianópolis county were included in the IPAQ questions in the answers field, instead of simply indicating the weekly frequency and the time of performance of the respective physical activities. A table in which the interviewer should register the time concerning each day of the week as well as in its different shifts (morning, afternoon and night) in each question was also included. The times were added in order to obtain the total weekly cost prior to the interview.

In order to test the IPAQ validity, its result was compared with the ones obtained through the physical activities file (DAF)⁽¹⁵⁾ and the pedometer.

The DAF allows the estimation of the physical activity level as well as the energetic cost concerning a given period of observation, based on detailed description and performed by the subjects throughout the day. Moreover, the day is divided in 96 periods of 15 minutes, categorized in 9 levels of intensity⁽¹⁵⁾.

The pedometer is a mechanical counter which registers movements performed in response to the body's vertical acceleration. It is attached to the waist, close to the navel, on one's belt or clothes and counts the steps as well as calculates the caloric cost according to the individual's body weight, storing the data to be noted by the subjects. In this study, the CITIZEN pedometer, model TW-30 was used.

The IPAQ was twice applied (test and retest), with interval of 21 days, in individual interviews in order to obtain information concerning stability of their measurements. The DAF and the pedometer were simultaneously applied during three days: two during the week and one on Sunday.

For the estimation of the energetic cost in physical activities, with the data collected through the IPAQ, the procedures described by Craig et al. were adopted⁽¹³⁾. For the DAF, the mean energetic demand of the three days of register was used; for the pedometer, the caloric cost according to the body weight registered in the equipment in the three days of research was verified. The comparison of the data obtained through the IPAQ (kcal/day), with the data derived from the DAF (kcal/day) and the pedometer (kcal/day) was determined by indicators of concomitant validity.

For the analysis of the test/retest reproducibility of the IPAQ, the Spearman Correlation (r_s) was performed, considering the final scores of the applications of the IPAQ obtained from the sum of each of the specific dimensions which constitutes the instrument (work, transportation, housework tasks and leisure).

For the analysis of the concomitant validity, the Concordance Percentile (%C); the Kappa index (k); the Spearman Correlation (r_s) between the estimates of energetic cost obtained by the application of the IPAQ and the measurements of the physical activities derived from the pedometer's data as well as the DAF were used.

Both for the analysis of the reproducibility and the concomitant validity the Bland and Altman plot in dispersion diagram was used⁽²⁰⁾. Such procedure allows the visualization of the different means and the extreme limits of concordance, in the case of two standard deviations of difference.

The data analysis was performed through the statistical package SPSS, version 10.0 for Windows, being considered the significance level of p < 0,05.

RESULTS

Twenty-nine retired men, aged at least 60 years, majority mean age of 66,6 (± 4,3) years, mean body weight of 73,3 (± 10,48) kg participated in the study.

Reproducibility

The value obtained in the Spearman Correlation was r_s = 0,95, which indicates good stability between measurements (test/retest), with statistically significant result (p < 0,01).

Concordance of the measurements of physical activities between IPAQ test and retest according to Bland-Altman procedure ⁽²⁰⁾ in elderly men may be observed in figure 1.

The mean values between variables are reported in the abscises, and in the ordinates, the individual differences between the results of the instrument are reported. Such analysis procedure was proposed by Bland and Altman⁽²⁰⁾ and enables the visualization of the differences between the means and the extreme limits of concordance (± 2 SD of the difference), presented by the different instruments applied and analyzed. The graphic sorting presents medium variation between the standard deviations. The differences between the means of the test and retest application are close to the means between the questionnaire application. However, a moderate variation in the caloric cost between applications is observed.

The reliability interval (± 2SD) between the test and retest IPAQ application was 7,29 to –14,0 (kcal/min); these results show an individual variability concerning the questionnaire application concordances.

Validity

Table 1 shows the concordance percentile; the Kappa index and the Spearman Correlation in the IPAQ concomitant validity evaluation and the measurements of physical activities obtained with the pedometer and DAF data.

The concordance percentile between IPAQ x pedometer (62%) and IPAQ x DAF (69%) was good.

The Spearman Correlation and the Kappa Index between IPAQ x pedometer, (r_{s =} 0,24; K = 0,03) and IPAQ X DAF (r_{s =} 0,38; K = 0,35) were low. Nevertheless, between IPAQ X DAF there was statistically significant difference (p < 0,05).

The sorting diagrams with the plot of mean values of the caloric cost of the pedometer and IPAQ and the DAF and IPAQ may be observed in figures 2 and 3.

The graphic sorting in figures 2 and 3 presents great variation between Standard deviations, being observed a great variation in the caloric cost, using one or the other instrument. The mean differences between the instruments should be the closest as possible to the means between the instruments. Although it has happened, the majority of the older individuals were within the two standard deviations set.

DISCUSSION

Reproducibility – According to the data presented in the partial report of the Executive Committee for the IPAQ evaluation⁽¹³⁾, several studies performed with samples derived from 12 countries presented results lower than the ones found in this study, which Spearman correlation was of r_s = 0,95, which shows good stability between the test and retest measurements. Concerning the estimate of the total physical activity level, Craig et al.⁽¹³⁾ verified that the Spearman correlation indices varied r_s = 0,46 in South Africa and r_s = 0,92 in Italy, while in Brazil such index reached r_s = 0,69⁽¹³⁾. The reproducibility through the Spearman correlation as well as the self-administered short version IPAQ questionnaire applied in adolescents of both sexes, varied between 0,49 and 0,83⁽²¹⁾.

In the present study, the reliability interval (± 2dp) between the IPAQ test and retest varied between 7,29 and –14,0 (kcal/min). These results demonstrate the moderate individual variability. The higher the breadth observed between the reliability intervals limits, the lower the reliability of the instrument.

Validity – the data found in table 1 show that the concordance coefficient (%) between the IPAQ results with the ones from the Pedometer and DAF are good.

The Kappa index (k) and Spearman correlation results are low both for IPAQ x Pedometer measurements and IPAQ x DAF. Therefore, one may observe that such results corroborate the evidence already presented in other studies^(8-10,21). They also agree with the indicators presented in the partial report of the Executive Committee of the IPAQ evaluation⁽¹³⁾.

When the IPAQ results applied in older men were compared with the ones in older women, the concordance coefficient with the pedometer and DAF were of the same magnitude (IC = 63%-68% respectively). The Kappa index (k) was also similar in women with the pedometer (k = 0,27) and with the DAF (women k = 0,37), not varying much in the data⁽¹²⁾.

It is relevant to mention that the physical activity level evaluation obtained with the utilization of the IPAQ considers the usual activities of the elderly, while the information obtained through the pedometer present limitation in its use, once it simply registers certain types of activities, which could not reflect the usual behavior of the elderly. The same limitation can be also found when comparing the IPAQ and the DAF data, since the data collection of the latter depends on the behavior register in a given observation period, which could mask the typical behavior as well. Despite that, due to the existence of alternatives accessible to these procedures, these instruments have been widely applied^(16-19).

The Spearman correlation of the indicators of concomitant validity between the measurements obtained through the IPAQ and the pedometer was of r_s = 0,24 (p = 0,20). Contrasting the IPAQ measurements with the measurements obtained through the DAF, the index was higher: r_s = 0,38 (p = 0,04). Despite being low, the index was statistically significant. Although the value was not the expected one, these results agree with the ones reported by the Executive Committee for IPAQ evaluation⁽¹³⁾ as well as the ones found in older women⁽¹²⁾. Other studies that used more accurate movement sensors (Computer Science & Applications – CSA) in order to obtain reference measurements for evaluation of the concomitant validity presented also fairly modest validity indicators⁽¹⁴⁾. It was also verified low correlation between the IPAQ (self-administered short) and the DAF in adolescents, being the correlation r_s = 0,39 and the Kappa coefficient presenting analogue results⁽²¹⁾.

Low correlation between the IPAQ and the pedometer was observed. The concordance percentile was of 62%; however, the Kappa index was low and not significant.

The mean differences between the instruments should be the closest as possible. However, a great variation in the estimate of the caloric cost is observed when using one or the other instrument.

A grouping in the points is observed in figures 2 and 3. In both figures the reliability intervals (± 2SD) between the IPAQ and the pedometer (514 to –767,1) and between the IPAQ and the DAF (154 to –854), demonstrate a high individual variability concerning the concordances of the questionnaire application and the standard instruments utilized. The higher the breadth observed between the reliability intervals limits, the lower the validity of the instrument evaluated. Thus, the results of the present study suggest that the IPAQ may be a reliable instrument, especially to discriminate two groups of activities (mean: more and less active).

Therefore, the IPAQ is an acceptable measurement instrument to measure the physical activity level in different places and languages; it is of easy application and low cost for large populations. Moreover, it has been tested in twelve different countries and can be used in order to measure the physical activity level in Brazilian older individuals.

CONCLUSION

The results obtained in this study show that the International Physical Activity Questionnaire (IPAQ); long version; usual/ordinary week; presents excellent test/retest reproducibility level when evaluated through the Spearman correlation and good reproducibility when evaluated through the Bland and Altman technique. The reliability was higher than the one found in validity and reproducibility studies conducted in Brazil with young adults, adolescents, middle-aged individuals and older women.

Concerning the concomitant validity, one may infer that the concordances between the IPAQ and the pedometer as well as the IPAQ and the DAF instruments are moderate. Satisfactory discrimination capacity in two groups (more active and less active) may be suggested.

REFERENCES

Received in 12/9/05.

Final version received in 27/7/06.

Approved in 2/8/06.

All the authors declared there is not any potential conflict of interests regarding this article.

• 1. Lee IM. Physical activity and cancer prevention: data from epidemiologic studies. Med Sci Sports Exerc. 2003;35(11):1823-7.
• 2. Lee IM, Hsieh C, Paffenbarger R. Exercise intensity and longevity in men. The Harvard Alumni Health Study. JAMA. 1995;273:1179-84.
• 3. Paffenbarger RS, Hyde R, Wing A, Hsieh C. Physical activity, all-cause mortality, and longevity of college alumni. N Engl J Med. 1986;314:605-13.
• 4. Weuve J, Kang JH, Manson JE, Breteler MMB, Ware JH, Grodstein F. Physical activity, including walking, and cognitive function in older people. JAMA. 2004; 292(12):1454-61.
• ⁵
  5. Ministério da Saúde. Anuário Estatístico de Saúde do Brasil 2001. http://portal. saude.gov.br/saude/aplicacoes/anuario2001/index.cfm. Acesso em 12 de abril de 2003.
• ⁶
  6. Atlas Corações do Brasil, 2005. Disponível em: http://prevencao.cardiol.br/campanhas/coracoesdobrasil/atlas/default.asp. Acesso em 10/12/2005.
• 7. Cassel C. The practical guide to aging: what everyone needs to know. 3rd ed. New York: University Press, 1999;352 p.
• 8. Hallal PC, Victora CG. Reliability and validity of the International Physical Activity Questionnaire (IPAQ). Med Sci Sports Exerc. 2004;36(3):556.
• 9. Matsudo S, Araújo T, Matsudo V, Andrade D, Andrade E, Oliveira C, et al. Questionário Internacional de Atividade Física (IPAQ): estudo de validade e reprodutibilidade no Brasil. Rev Bras Ativ Fís Saúde. 2001;6(2):5-12.
• 10. Pardini R, Matsudo S, Matsudo V, Araújo T, Andrade E, Braggion G, et al. Validation of International Physical Questionnaire (IPAQ): pilot study in Brazilian young adults. Med Sci Sports Exerc. 1997;29(6):S5-S9.
• 11. Barros M, Nahas M. Reprodutibilidade (teste/reteste) do Questionário Internacional de Atividade Física (QIAF-versão 6): um estudo piloto com adultos no Brasil. Rev Bras Ciên e Mov. 2000;8(1):23-6.
• 12. Benedetti TRB, Mazo GZ, Barros MV. Aplicação do Questionário Internacional de Atividade Física para avaliação do nível de atividades físicas de mulheres idosas: validade concorrente e reprodutibilidade teste/reteste. Rev Bras Ciên e Mov. 2004; 12(1):25-33.
• 13. Craig CL, Marshall AL, Sjostrom M, Bauman AE, Booth ML, Ainsworth BE, et al. International physical activity questionnaire: 12-country reliability and vality. Med Sci Sports Exerc. 2003;35(8):1381-95.
• 14. Mazo G, Benedetti T, Mota J, Barros M. Validade concorrente e reprodutibilidade teste/reteste do Questionário de Baecke modificado para idosos. Rev Bras Ativ Fis Saúde. 2001;6(1):5-11.
• 15. Bouchard C, Tremblay A, Leblanc C, Lortie G, Savard R, Theriault G. A method to assess energy expenditure in children and adults. Am J Clin Nutr. 1983;37(3): 461-7.
• 16. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods for clinical measurement. Lancet. 1986;8:307-10.
• 17. Guedes DP, Lopes CC, Guedes JERP. Reprodutibilidade e validade do Questionário Internacional de Atividade Física em adolescentes. Rev Bras Med Esporte. 2005;11(2):151-8.
• 18. Beverige S, Watson D, Scruggs P. Validation of the YAMAX pedometer for measurement of the intensity of physical activity of fifth grade children during physical education classes. In: Measurement of Physical Activity Conference Series. Program and Abstract Book. Dallas, Texas: Cooper Institute [Abstract], 1999; 22.
• 19. Eston RG, Rowlands AV, Ingledew DK. Validity of heart rate, pedometry and accelerometry for predicting the energy cost of children's activities. J Appl Physiol. 1998;84:362-71.
• 20. Mahar M, Hales D, Willians G, Dinallo J. Physical activity assessment by Tritrac-R3D accelerometer, digiwalker pedometer and self-report measures in children. In: Measurement of Physical Activity Conference Series. Program and Abstract Book. Dallas, Texas: Cooper Institute, 17 [Abstract], 1999.
• 21. Voorrips LE, Ravelli AC, Dongelmans PC, Deurenberg P, Van Staveren WA. A physical activity questionnaire for the elderly. Med Sci Sports Exerc. 1991;23(8):974-9.

appendix 1

Publication Dates

• Publication in this collection
  11 Sept 2007
• Date of issue
  Feb 2007

History

• Accepted
  02 Aug 2006
• Received
  12 Sept 2005
• Reviewed
  27 July 2006