Construct Validity and Reliability of the ‘Lifestyle (PAHO) in the Adult Population’ Questionnaire

Abstract

Having validated instruments for the evaluation of lifestyles becomes a necessity for professional practice in terms of promoting the adoption of health-promoting behaviors in various aspects of daily life, such as social interactions, nutrition, physical activity, oral health, environmental health, mobility, and psychoactive substance use.

Objective

To evaluate the construct validity and reliability of the ‘Lifestyle Questionnaire in the Colombian Population” over 18 years of age from the northeastern of Colombia.

Methodology

A psychometric study was conducted with the participation of 551 adults using the instrument proposed by the Pan American Health Organization (PAHO). A confirmatory analysis using the structural equation modeling (SEM) technique was carried out, in addition to an internal consistency analysis.

Results

The confirmatory factor analysis reports the same structure, with adequate goodness-of-fit indicators (X²/df = 2.10, CFI = .88, TLI = .87, RSMEA = .05, SRMR = .04). The reliability estimation of the test and its dimensions, with internal consistency coefficients Cronbach’s α and McDonald’s ω, reported optimal levels in the total test (ω = .93).

Conclusions

The instrument developed by PAHO showed adequate construct validity and reliability indicators for evaluating lifestyles in northeastern Colombian people over 18 years old.

Keywords

lifestyles adult construct validity reliability

Introduction

Lifestyles refer to a set of behaviors and conduct of individuals, considered an abstract and complex concept that has been studied for many years.¹ Various disciplines, including sociology, anthropology, and epidemiology have investigated this concept. Epidemiology has widely used the concept of lifestyles about health, though with a more restricted interpretation; it mainly associates it with consciously adopted behaviors that may pose risks to individuals’ well-being.²

Consequently, the adoption of lifestyles develops throughout an individual’s life, indicating that it is a learning process that can occur through the assimilation or imitation of family models, as well as informal or formal groups.³

During the transition to adulthood, individual behaviors, lifestyle, and well-being can offer valuable insights into how to promote health throughout the life cycle. During this period, it is common for individuals to adopt a variety of behaviors, some beneficial and others less healthy.⁴ Consequently, it is confirmed that the adoption of unhealthy lifestyles in life is associated with the presence of various metabolic, cardiovascular, endocrine, and chronic disabling diseases.^5-7

Moreover, research on successful aging emphasizes the need to recognize and adopt healthy lifestyles from an early age.⁸ This is even more relevant given the growing percentage of older adults, projected to increase by 16% by the year 2050.⁹

The Pan American Health Organization (PAHO) and the World Health Organization (WHO) highlight the importance of addressing various determinants of health, including a balanced diet, regular physical activity, stress management, adequate rest, and the elimination of harmful practices such as tobacco and alcohol consumption. Both institutions emphasize promoting healthy habits in all aspects of daily life to contribute to well-being and disease prevention in the population.^7,10

In this context, the role of health professionals is fundamental in promoting healthy habits throughout an individual’s life. Thus, having properly validated research tools represents an opportunity to identify relevant variables that influence people’s health, particularly regarding lifestyles and their components, such as physical activity, nutrition, rest, relationships with others, oral health, sexuality, mobility, harmful substance use, life purpose, and the environment.^2-4

The literature includes various instruments that evaluate lifestyles, such as the Health-Promoting Lifestyle Profile (HPLP), (six dimensions),¹¹ the Healthy Lifestyles Questionnaire (EVS), (seven dimensions),¹² the FANTASTIC instrument, (seven dimensions),¹³ and the one proposed by the Pan American Health Organization (PAHO) on lifestyles, (ten dimensions).¹⁴

The PAHO Lifestyle Assessment Instrument offers significant advantages over other tests, as it was specifically designed for the Americas, enhancing its applicability, regional validation, and utility in health policy formulation. In addition to enabling precise analysis of individual risk factors, it facilitates the integration of data into public health information systems, optimizing planning and decision-making. In this regard, the authors aim to verify its psychometric properties previously identified in Spanish. However, given the significant cultural and linguistic variability of the northeastern Colombian population, influenced by its border location, it is essential to assess its reliability and validity in this specific context.

In Colombia, some institutions, particularly in the education, health, and social services sectors have utilized the original instrument PAHO lifestyle assessment instrument to evaluate lifestyles (physical activity, nutrition, rest, social relationships, oral health, sexuality, mobility, substance use, sense of life, and environment)^14,15; however, the psychometric properties of this questionnaire have not been rigorously evaluated with the population of northeastern Colombia, who have particular accents and verbal expressions, different from other populations in the same country.

Furthermore, a limited number of Colombian instruments exist for assessing healthy habits in the general adult population,¹⁶ while one such instrument exists, its psychometric properties have not been reported. Available research primarily focuses on the psychometric properties of the adult¹³ and adolescent¹⁷ versions of the FANTASTIC instrument, as well as instruments adapted by Varela et al¹⁸ and Niño-Bautista et al¹⁹ for assessing healthy habits in children.

Therefore, this study aims to evaluate the construct validity and reliability of the lifestyle questionnaire in the northeastern Colombia population over 18 years old.

Methods

Type of Research

A study was conducted to validate the psychometric properties of a lifestyle instrument developed by the Pan American Health Organization (PAHO) in adults from Norte de Santander, Colombia. First, a description of the sample was provided, followed by the results regarding the lifestyles of the respondents. Finally, the psychometric analyses related to the test were presented.

Instrument

The survey titled “What is My Lifestyle Like?”, proposed by PAHO, translated for Colombia in the context of the “Healthy University” program,²⁰ comprises 10 domains (45 questions in total): (1) social interactions; (2) physical exercise; (3) rest; (4) diet; (5) oral hygiene; (6) sexuality; (7) mobility; (8) substance use; (9) purpose in life; and 10) environmental surroundings (Table 1). Each question is assessed on an ordinal scale, where 0 means “Never, or rarely”; 1 means “Sometimes”; and 2 means “Always, or almost always”. To identify healthy lifestyle habits, the score of all responses is added for a total of 90. A total score of less than 40 points could indicate areas that need attention and suggests that interventions may be necessary to promote healthier habits.^14,20

Table 1.

Item and Factors of the “what is My Lifestyle Like?” Survey.

Factor	Items
F1. Social interactions	1, 2, 3, 4, 5, 6, 7
F2. Physical activity	8, 9, 10
F3. Rest	11, 12, 13, 14
F4. Nutrition	15, 16, 17, 18
F5. Oral health	19, 20, 21, 22
F6. Sexuality	23, 24, 25, 26, 27
F7. Mobility	28, 29, 30, 31
F8. Substance use	32, 33, 34, 35, 36
F9. Purpose of life	37, 38, 39, 40, 41
F10. Environment surrounding	42, 43, 44, 45

Procedure

A convenience sample of 551 adult participants from the Norte de Santander (Colombia) region was recruited. This sample not only meets but exceeds the calculated size for an observational study (380), as well as the recommendation of authors such as Snook and Gorsuch,²¹ who establish a minimum of 300 observations as sufficient to ensure reliable measurements. Additionally, it surpasses the suggested number based on factor loadings proposed by Hair et al,²² which, in our case, was set at 350.

The inclusion criteria included being of legal age, possessing reading and writing skills, and not having a cognitive disability. All participants provided written informed consent. Table 2 presents a detailed description of the sample characteristics.

Table 2.

Characteristics of the Sample.

Sex	f	%
Male	197	35.8
Female	354	64.2
Educational level
Primary or less	150	27.2
Completed secondary	189	34.3
Incomplete higher education	90	16.3
Completed higher education	122	22.1
Employment status
Unemployed	106	19.2
Informal worker	156	28.3
Employed with a contract	205	37.2
Other	84	15.2
Marital status
Single	289	52.5
Married	138	25.0
Cohabiting	85	15.4
Divorced	19	3.4
Widowed	20	3.6
Health system affiliation
Contributive	362	65.7
Subsidized	160	29.0
Other	29	5.3
State benefits
Yes	125	22.7
No	364	66.1
Not applicable	62	11.3

Source: Own elaboration.

A research team received a one-week training program on the application of the research instrument. Participants were provided with a link to a Google Forms questionnaire, which included informed consent, the Spanish Version of the Lifestyle (PAHO) questionnaire for the adult population, and basic sociodemographic variables (gender, age, marital status, educational level, and occupation).

Participants completed the self-administered questionnaire voluntarily and autonomously at the University of Pamplona. Researchers were available to address any questions and ensure unbiased responses by minimizing potential distractions or misunderstandings. The average questionnaire completion time was twenty (20) minutes. Upon data collection, the database was downloaded in Excel format for subsequent analysis.

Ethics

The data obtained in this study were handled with strict confidentiality and anonymity. Participation was voluntary and supported by informed consent. The guidelines of the Declaration of Helsinki of 1975 and Article 11 of Resolution #8430 of 1993 of the “Ministerio de Salud” of Colombia were followed.²³

Statistical Analysis

A descriptive analysis of frequencies and percentages was conducted to describe the sample, along with a descriptive analysis of continuous variables to establish possible significant differences according to sociodemographic criteria. A Kolmogorov-Smirnov analysis was performed to determine the distribution of the data. Those dimensions with a P-value >.05 were considered parametric.²⁴

For confirmatory analysis, the structure was evaluated considering acceptable fit indicators such as Tucker-Lewis Index (TLI) and Comparative Fit Index (CFI) > .90; Additionally, criteria such as RMSEA less than .010, SRMR close to zero, and X²/df close to 3.0 were used as defined by Brown.²⁵

A Spearman correlation was developed to determine how the factors of the evaluated scale are associated, and statistical significance <.01 was established. Finally, Cronbach’s α and McDonald’s ω statistics were applied to assess the internal consistency of the obtained factors as well as the total test, assuming values >.70 for the total test as adequate.²⁶

For the development of the referred calculations, the software tools SPSS version 26® and the graphical interface of the R Project - JAMOVI 2.3® application were used.

Results

A sample of 551 participants was analyzed (64.2% women), with a mean age of 34.9 years (SD 15.0 years). They completely responded to the instruments provided, leaving no items left unanswered. The remaining sociodemographic data can be seen in Table 1.

On the other hand, Kolmogorov-Smirnov analysis revealed that the data distribution for all instrument dimensions deviated from normality.

For the confirmatory analysis, a model was obtained with adequate goodness-of-fit indicators for X²/df, RMSEA, and SRMR, while for CFI and TLI, the values are close to the desired threshold (Table 3).

Table 3.

Confirmatory Factor Analysis of the ‘Lifestyle (PAHO) in the Adult Population’ Questionnaire.

Model	X²	df	X²/df	CFI	TLI	SRMR	RMSEA (95% CI)
Obtained	1892	900	2.10	.88	.87	.051	.044 (.041-.047)
Desired	---	---	<3.00	>.90	>.90	Close to 0	<.100

The correlation matrix shows positive and significant correlations between all factors of the scale; it should be noted that none of them exceeds ,70, which rules out possible collinearity between the factors of the scale (Table 4).

Table 4.

Correlations Between the Factors of the Healthy Lifestyles Scale Proposed by the Pan American Health Organization.

Spearman’s rho	F1	F2	F3	F4	F5	F6	F7	F8	F9	F10
F1		.274^a	.542^a	.449^a	.394^a	.314^a	.440^a	.366^a	.556^a	.466^a
F2	.274^a		.383^a	.455^a	.261^a	.123^a	.187^a	.208^a	.280^a	.276^a
F3	.542^a	.383^a		.484^a	.287^a	.268^a	.295^a	.235^a	.539^a	.369^a
F4	.449^a	.455^a	.484^a		.404^a	.292^a	.393^a	.417^a	.456^a	.453^a
F5	.394^a	.261^a	.287^a	.404^a		.317^a	.333^a	.383^a	.388^a	.470^a
F6	.314^a	.123^a	.268^a	.292^a	.317^a		.370^a	.424^a	.312^a	.311^a
F7	.440^a	.187^a	.295^a	.393^a	.333^a	.370^a		.507^a	.414^a	.471^a
F8	.366^a	.208^a	.235^a	.417^a	.383^a	.424^a	.507^a		.335^a	.494^a
F9	.556^a	.280^a	.539^a	.456^a	.388^a	.312^a	.414^a	.335^a		.493^a
F10	.466^a	.276^a	.369^a	.453^a	.470^a	.311^a	.471^a	.494^a	.493^a

^aThe correlation is significant at the 0.01 level (two-tailed).

Lastly, the reliability estimation of the test and dimensions with internal consistency coefficients, Cronbach’s α, and McDonald’s ω reported optimal levels for the total test (ω = .93). Therefore, it can be concluded that the measures obtained with it are highly reliable (Table 5).

Table 5.

Internal Consistency of the ‘Lifestyle (PAHO) in the Adult Population’ Questionnaire.

Factor	Cronbach’s α	McDonald’s ω
F1	.772	.777
F2	.822	.830
F3	.700	.707
F4	.635	.648
F5	.647	.666
F6	.752	.769
F7	.712	.731
F8	.786	.793
F9	.790	.798
F10	.672	.688
Total	.929	.933

Source: Own elaboration.

Discussion

In response to the call to action of the “2030 Agenda for Sustainable Development”, which seeks to mobilize efforts to protect and improve lives worldwide, one thematic area gaining significant attraction focuses on individuals’ lifestyles, given their profound impact on health and quality of life. Consequently, governmental organizations and the Pan American Health Organization (PAHO) provide resources to guide and evaluate health behaviors, including instruments for assessing individual lifestyles.²⁷ This study demonstrates that the PAHO Lifestyle Questionnaire (Spanish Version) exhibits adequate construct validity and reliability for evaluating lifestyles in adults from the Colombian northeastern population aged 18 and older.

This finding underscores the scale’s utility in assessing healthy behaviors within the target population. Furthermore, optimal fit indices (χ²/df) suggest that the data align with the evaluated theoretical model.²⁸ While CFI and TLI values approach the acceptance limit (Brown), they may indicate a correspondence of the data with a neutral and independent structure.²⁴ Acceptable RMSEA and SRMR values suggest low approximation error between the proposed model and the evaluated population, along with good agreement between expected and estimated covariances.^25,28

Therefore, the obtained model demonstrates reasonable fit, with potential for improvement. High and positive correlations between factors align with the instrument’s theoretical model, with no evidence of conceptual overlap between dimensions (correlations not exceeding 0.70). Adequate item-total correlations further support the instrument’s overall reliability.²⁶

Regarding internal consistency, the overall scale exhibits acceptable values, indicating coherent measurement of the construct. However, subscales F4, F5, and F6 exhibit values below 0.70, considered acceptable but not optimal. This may be attributed to the limited number of items (four) contributing to each dimension, potential heterogeneity among items within these subscales, or the items’ stronger association with the general construct of healthy habits than with the specific dimension.²⁷ Future research with larger, probabilistic samples will help determine if this trend persists or is incidental.²⁹

When comparing this study with other investigations using different lifestyle assessment instruments, both similarities and differences emerge. Psychometric evaluations of the FANTASTIC instrument (version 3) and the EVS instrument align with our findings, demonstrating reliable measurements and maintaining original dimensions.

However, regarding internal consistency within dimensions, the nutrition and oral hygiene dimensions of the EVS instrument exhibit Cronbach’s alpha values below the 0.70 criterion established for satisfactory internal consistency.^25,27 In contrast, the Spanish version of the Healthy Lifestyles Questionnaire demonstrates higher values across its four dimensions. Findings diverge from those of the Health Promoting Lifestyle Profile-II (HPLPII),³⁰ which exhibited low convergent and discriminant validity in its 52-item version, necessitating the removal of six items to stabilize values for five of the six dimensions. In this context, the PAHO Lifestyle Assessment Instrument and the HPLPII, with over 30 items and 10 dimensions, contrast with the current study, which confirms the relevance of the original instrument’s dimensions while retaining all items.

The findings of this research should be interpreted considering several limitations. Despite the large sample size, its convenience sampling limits the generalizability of results to other Colombian or Spanish-speaking populations. Furthermore, as this was a cross-sectional study, test-retest reliability analysis to assess measurement stability over time was not conducted. Additionally, social desirability bias or overestimation of healthy habits among participants may have influenced the reported results.³¹

Future research should address these limitations through a new study with a probabilistic sample and incorporate Item Response Theory (IRT) analysis to investigate differential item functioning (DIF) across subgroups.³² Additionally, factorial invariance analysis (configural, metric, scalar, and strict) can be implemented to determine if the instrument’s structure remains consistent across different population profiles (sex, age, etc.), ensuring that observed differences in population studies are not attributable to inherent instrument biases.³³

Nevertheless, this study provides empirical evidence supporting the use of the Spanish version of the Lifestyle (PAHO) in the adult population of northeastern Colombia. The instrument exhibits acceptable goodness-of-fit indicators, supporting the proposed 10-factor structure and enabling a reliable measurement of healthy behaviors, particularly in healthy adults within this context. Additionally, it is recommended that future confirmatory studies apply the test-retest method to assess the instrument’s temporal stability.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The author(s) received financial support from the University of Pamplona for the publication of this article.

ORCID iDs

Mayra Alejandra Barajas Lizarazo

César Andrés Gómez Acosta

Katty Dayana Escobar Velásquez

References

Nawas

. El estilo de vida. Rev Latinoam Psicologia. 1971;3(1):91-107. https://www.redalyc.org/pdf/805/80503105.pdf

Álvarez

. Lifestyles: from the individual to the context. Rev Fac Nac Salud Publica. 2012;30(1):95-101. https://www.redalyc.org/articulo.oa?id=12023071011

Sanabria-Ferrand

González

Urrego

. Estilos de vida saludable en profesionales de la salud colombianos. Estudio exploratorio. Rev Med. 2007;15(2):207-217. https://www.redalyc.org/articulo.oa?id=91015208

Lawrence

Mollborn

Hummer

. Health lifestyles across the transition to adulthood: implications for health. Soc Sci Med. 2017;193:23-32. doi:10.1016/j.socscimed.2017.09.041

El-Sherif

Taylor

Ayakannu

. Ovarian cancer: lifestyle, diet and nutrition. Nutr Cancer. 2020;73(7):1092-1107. doi:10.1080/01635581.2020.1792948

Arocha

. Sedentary lifestyle a disease from xxi century. Clín Invest Arterioscler. 2019;31(5):233-240. doi:10.1016/j.arteri.2019.04.004

Arvanitis

Lowenstein

. Dyslipidemia. Ann Intern Med. 2023;176(6):ITC81-ITC96. doi:10.7326/AITC202306200

Friedman

. Lifestyle (medicine) and healthy aging. Clin Geriatr Med. 2020;36(4):645-653. doi:10.1016/j.cger.2020.06.007

United Nations . Envejecimiento. UN. https://www.un.org/es/global-issues/ageing

10.

Pan American Health Organization . Hábitos y estilos de vida saludables: asesoramiento para los pacientes. Washington, D.C.: PAHO. https://iris.paho.org/bitstream/handle/10665.2/50805/OPSNMH19001_spa.pdf?sequence=1&isAllowed=y

11.

Espinoza

Vanegas

. Validación de un cuestionario para medir el estilo de vida de los estudiantes Universitarios en el marco de la teoría de Nola Pender (EVEU). Rev Torreón Univ. 2019;7(19):38-49. doi:10.5377/torreon.v7i19.7909

12.

Leyton

Lobato

Batista

Aspano

Jiménez

. Validación del cuestionario de estilo de vida saludable (EVS) en una población española. Rev Iberoam Psicol Ejerci Depor. 2018;13(1):23-31. https://www.redalyc.org/articulo.oa?id=311153534002

13.

Ramírez

Agredo

. Fiabilidad y validez del instrumento “Fantástico” para medir el estilo de vida en adultos colombianos. Rev Salud Pública. 2012;14(2):226-237. doi:10.1590/S0124-00642012000200004

14.

Mora

Múnera

. Evaluation of healthy lifestyles at the school of medicine of fundación universitaria de ciencias de La salud. Rev Repert Med Cir. 2015;24(4):267-274. doi:10.31260/RepertMedCir.v24.n4.2015.600

15.

Ballén

. Estilos de vida saludables en estudiantes de último año de la carrera de nutrición y dietética en la Pontificia de la Pontificia Javeriana de Bogotá. Bogotá: Pontificia Universidad Javeriana; 2022. https://repository.javeriana.edu.co/handle/10554/62530

16.

Arrivillaga-Quintero

Salazar-Torres

. Creencias relacionadas con el estilo de vida de jóvenes latinoamericanos. Psicol Conduct. 2005;13(1):19-36. https://www.behavioralpsycho.com/wp-content/uploads/2020/04/02.Arrivillaga_13-1oa-1.pdf

17.

Betancurth

Vélez-Alvárez

Jurado-Vargas

. Validación de contenido y adaptación del cuestionario FANTÁSTICO por técnica Delphi. Rev Cientif Salud Uninorte. 2015;31(2):214-227. doi:10.14482/sun.31.2.5583

18.

Varela-Arévalo

Ochoa-Muñoz

Tovar-Cuevas

. Medición de hábitos saludables y no saludables en niños: Síntesis de la información utilizando indicadores y conglomerados. Rev Mex Trastor Aliment. 2018;9(2):264-276. doi:10.22201/fesi.20071523e.2018.2.487

19.

Niño-Bautista

Hakspiel-Plata

Mantilla-Hernandez

Cardenas-Herrera

Guerrero-Parra

. Adaptación y validación de instrumento para evaluar habilidades psicosociales y hábitos saludables en escolares. Univ Salud. 2017;19(3):366-377. doi:10.22267/rus.171903.99

20.

Becerra-Bulla

Pinzón-Villate

Vargas-Zárate

. Hacia la creación del programa universidad promotora de la salud desde la alimentación y nutrición en la Universidad Nacional de Colombia, sede Bogotá. Rev Fac Med. 2011;59(1):S67-S76. https://repositorio.unal.edu.co/handle/unal/36429

21.

Snook

Gorsuch

. Component analysis versus common factor analysis: a Monte Carlo study. Psychol Bull. 1989;106(1):148-154. doi:10.1037/0033-2909.106.1.148

22.

Hair

Gabriel

MLDS

Silva

Braga

. Development and validation of attitudes measurement scales: fundamental and practical aspects. RAUSP Management J. 2019;54(4):490-507. https://www.redalyc.org/articulo.oa?id=553861448011

23.

Ministry of Health of Colombia . Resolución #8430 de 1993. Minsalud. https://minsalud.gov.co/sites/rid/Lists/BibliotecaDigital/RIDE/DE/DIJ/RESOLUCION-8430-DE-1993.pdf

24.

Luzuriaga-Jaramillo

Espinosa-Pinos

Haro-Sarango

Ortiz-Román

. Histograma y distribución normal: Shapiro-Wilk y Kolmogorov Smirnov aplicado en SPSS: Histogram and normal distribution: Shapiro-Wilk and Kolmogorov Smirnov applied in SPSS. LATAM. 2023;4(4):596-607. doi:10.56712/latam.v4i4.1242

25.

Brown

. Confirmatory Factor Analysis for Applied Research. 2nd ed. New York City: Guilford Press; 2015.

26.

Viladrich

Angulo-Brunet

Doval

. A journey around alpha and omega to estimate internal consistency reliability. Anal Psicol. 2017;33(3):755-782. doi:10.6018/analesps.33.3.268401

27.

Pan American Health Organization . Curso de vida saludable. Washington, D.C.: PAHO. https://www.paho.org/es/temas/curso-vida-saludable

28.

Field

. Discovering Statistics Using IBM SPSS Statistics. 7th ed. New York City: Sage Publications Ltd; 2018.

29.

McNeish

. Thanks coefficient alpha, we’ll take it from here. Psychol Methods. 2018;23(3):412-433. doi:10.1037/met0000144

30.

Reyna

Peche

González

, et al. Psychometric properties of the Spanish version of the Health-Promoting Lifestyle Profile-II instrument in Mexican university students. Enferm Glob. 2022;21(66):398-423. doi:10.6018/eglobal.490521

31.

Tracey

. A note on socially desirable responding. J Counsel Psychol. 2016;63(2):224-232. doi:10.1037/cou0000135

32.

Zumbo

. A Handbook on Differential Item Functioning. New York City: Guilford Press; 2007.

33.

Bauer

Belzak

Cole

. Simplifying the assessment of measurement invariance over multiple background variables: using regularized moderated nonlinear factor analysis to detect differential item functioning. Struct Equ Model. 2020;27(1):43-55. doi:10.1080/10705511.2019.1642754