Questionnaire development in ELSA-Brasil: challenges of a multidimensional instrument

This article describes the development of the Brazilian Longitudinal Study for Adult Health (ELSA-Brasil) questionnaire. We first address the selection of topics whose contents have to cover the knowledge available on the complex causal network of outcomes and allow comparability with similar studies. Then we deal with the "translation and adaptation of measurement instruments" including neighborhood environment rating scales, depression and anxiety disorder rating scale and a food frequency questionnaire and discuss criteria that guided "theme block sequencing". And finally we focus on the practical importance of "pretesting and pilot studies". The ELSA may provide an original contribution regarding factors that cause or aggravate the outcomes of interest in the Brazilian population, as well as protective factors.

Questionnaires; Data Collection; Socioeconomic Factors; Life Style; Food Habits; Risk Factors; Multicenter Studies as Topic, methods; Cohort Studies

The authors declare that there are no conflicts of interest.

This manuscript was submitted for publication and underwent a peer review process as any other manuscripts submitted to this publication, and anonymity was guaranteed for authors and reviewers. Editors and reviewers declare no conflicts of interest that may affect the peer-review process.

SCIMAGO INSTITUTIONS RANKINGS

Fundação Oswaldo Cruz, Escola Nacional de Saúde Pública, Departamento de Epidemiologia e Métodos Quantitativos, Rio de Janeiro, RJ, BrasilFundação Oswaldo CruzBrasilRio de Janeiro, RJ, BrasilFundação Oswaldo Cruz, Escola Nacional de Saúde Pública, Departamento de Epidemiologia e Métodos Quantitativos, Rio de Janeiro, RJ, Brasil

Universidade Federal Fluminense, Instituto de Saúde da Comunidade, Departamento de Planejamento em Saúde, Rio de Janeiro, RJ, BrasilUniversidade Federal FluminenseBrasilRio de Janeiro, RJ, BrasilUniversidade Federal Fluminense, Instituto de Saúde da Comunidade, Departamento de Planejamento em Saúde, Rio de Janeiro, RJ, Brasil

Universidade Federal de Ouro Preto, Escola de Nutrição, Departamento de Nutrição Clínica e Social, Ouro Preto, MG, BrasilUniversidade Federal de Ouro PretoBrasilOuro Preto, MG, BrasilUniversidade Federal de Ouro Preto, Escola de Nutrição, Departamento de Nutrição Clínica e Social, Ouro Preto, MG, Brasil

Universidade Federal do Espírito Santo, Departamento de Enfermagem, Programa de Pós-Graduação em Saúde Coletiva, Vitória, ES, BrasilUniversidade Federal do Espírito SantoBrasilVitória, ES, BrasilUniversidade Federal do Espírito Santo, Departamento de Enfermagem, Programa de Pós-Graduação em Saúde Coletiva, Vitória, ES, Brasil

Universidade Federal do Rio Grande do Sul, Faculdade de Medicina, Programa de Pós-Graduação em Epidemiologia, Porto Alegre, RS, BrasilUniversidade Federal do Rio Grande do SulBrasilPorto Alegre, RS, BrasilUniversidade Federal do Rio Grande do Sul, Faculdade de Medicina, Programa de Pós-Graduação em Epidemiologia, Porto Alegre, RS, Brasil

Universidade Federal do Espírito Santo, Departamento de Educação Integrada em Saúde, Programa de Pós-Graduação em Saúde Coletiva, Vitória, ES, BrasilUniversidade Federal do Espírito SantoBrasilVitória, ES, BrasilUniversidade Federal do Espírito Santo, Departamento de Educação Integrada em Saúde, Programa de Pós-Graduação em Saúde Coletiva, Vitória, ES, Brasil

Universidade de São Paulo, Hospital Universitário, Centro de Pesquisa Clínica e Epidemiológica, São Paulo, SP, BrasilUniversidade de São PauloBrasilSão Paulo, SP, BrasilUniversidade de São Paulo, Hospital Universitário, Centro de Pesquisa Clínica e Epidemiológica, São Paulo, SP, Brasil

Universidade Federal da Bahia, Instituto de Saúde Coletiva, MUSA Programa Integrado em Gênero e Saúde, Salvador, BA, BrasilUniversidade Federal da BahiaBrasilSalvador, BA, BrasilUniversidade Federal da Bahia, Instituto de Saúde Coletiva, MUSA Programa Integrado em Gênero e Saúde, Salvador, BA, Brasil

Figures | Tables

Figures (1)
Tables (2)

Thumbnail

Figure
Standard-size tools for the Food Frequency Questionnaire used in all ELSA-Brasil study sites.

Thumbnail

Table 1
Topics included in the Longitudinal Study for Adult Health questionnaire.

Thumbnail

Table 2
Quality control form on the topic job stress.

Figure Standard-size tools for the Food Frequency Questionnaire used in all ELSA-Brasil study sites.

Table 1 Topics included in the Longitudinal Study for Adult Health questionnaire.

Sociodemographic characteristics ^a,b	Age, gender, skin color/race/ethnicity
	Migration history, place of residence, length of living in this area
	Educational and occupational history (respondent and spouse)
	Family income and asset ownership
	Housing conditions and household composition (current and past)
	Family life and spouse/partner characteristics
	Head of household
	Religion (current and past)
	Parental education and occupation
	Living conditions during childhood
Past medical history ^a	Self-rated health, past medical history of cardiovascular diseases, diabetes, kidney disease, cancer and other chronic diseases; medical procedures of interest
	Angina ⁵, intermittent claudication and heart failure questionnaire
	Headache questionnaire
Occupational history ^a,b	Job stress
	Job characteristics (level of autonomy, budget management, authority) Retirement
	Family-work imbalance
Family history ^a	Family history of specific conditions such as cardiovascular diseases, diabetes and sudden death
Reproductive history ^b	Menarche, menstrual cycles and menopause
	Contraceptive use
	Reproductive history
	Hormone therapy
	Infertility
Access to health care ^b	Access to screening programs, health insurance and utilization of health services
Psychosocial factors ^a,b	Environment/neighborhood characteristics (leisure, sports, food shopping)
	Social network
	Experience of discrimination
	Social capital Vital events
	Self-assessment of social status
History of body weight and image ^b	Birth weight and weight at age 20
History of body weight and image ^b	Body image (current and desired)
Food consumption ^b	Food frequency questionnaire
Smoking ^a	Former and current smoking, lifetime tobacco use, attempts to quit smoking with drug therapy, use of tobacco products with lower nicotine content, passive smoking
Alcohol consumption ^b	Type of alcoholic beverages; alcohol use frequency; pattern of alcohol consumption
Physical activity ^b	Current physical activity including occupational, home, leisure-time activities and sports
Medication use ^b	Use of prescription and non-prescription medication, vitamins, dietary supplements, and other drugs used in the past month Patients were asked to bring to the study visit all medications and prescription drugs they are on
Cognitive function ^b	Three tests for assessing immediate and delayed memory of the Consortium to Establish a Registry for Alzheimer’s disease validated for the Brazilian population.
	Trail test to assess executive cognitive function related to attention, concentration and psychomotor speed.
	Verbal fluency tests: animals and words beginning with the letter F
Mental health ^b	Clinical Interview Schedule-Revised 16 validated for Brazilian Portuguese including all 14 sections: somatic symptoms, fatigue, concentration, depression, irritability, sleep, concerns with physical symptoms, depressive thoughts, worry, anxiety, phobia, panic, compulsions and obsessions

^a Stage 1: Workplace interview^b Stage 2: Interview + assessments at ELSA study site

Table 2 Quality control form on the topic job stress.

Description / Rationale
	The purpose of the inclusion job stress scale(s) in the ELSA questionnaire is to explore the associations of this construct with main study outcomes (cardiovascular diseases and diabetes). There are virtually no studies on job stress conducted in Brazil. Two scales have been used in other countries to assess job stress. The oldest one, proposed by Robert Karasek in the 1970s, is based on the demand-control model. In 1982, Johannes Siegrist proposed a new instrument based on two other concepts: effort and reward. The dimensions evaluated in each scale are different, but both were used in the Whitehall Study (Levi et al, 2000).
	Cross-cultural adaptations for both scales are available in Brazil, and it was suggested to include both at baseline. There is evidence showing that both scales should be concomitantly administered because they assess different dimensions of job stress and may help better understanding potential associations with health outcomes (Kivimaki et al, 2006; Chandola et al, 2005; Calnan et al, 2004; Ostry et al, 2003; Kivimaki et al, 2002, Peter et al, 2002) and cardiovascular events.
Association between job stress and ELSA outcomes
	A recent meta-analysis on the association between job stress and coronary heart disease showed considerable heterogeneity of effects across studies, and a scarcity of studies with an adequate sample of women. This meta-analysis included both measurement scales and found a 50% higher risk of coronary heart disease among workers classified as having job stress (Kivimaki et al, 2006).
	Similarly, there were no consistent results regarding the association with hypertension. The results of most studies suggest a positive association with systolic and/or diastolic blood pressure among men but not among women. (Ohlin et al 2007; Belkic et al, 2000 and Brisson, 2000).
	It is interesting to note different results for diabetes according to the scale applied: no association was found using the demand-control scale (Agardh et al, 2003; Kroenke, 2006); the Whitehall II Study applied the effort-reward scale and reported a positive association (Kumari et al, 2004). However, the number of studies using these scales, either together or individually, was low for this outcome. As for metabolic syndrome, a positive association was reported in the Whitehall Study (Chandola, 2006) using the demand-control scale.
Instrument characteristics
	The demand-control scale consists of 17 questions with four answer choices that assess three dimensions: psychological demands (5 questions), control of work process (6 questions), and social support at work (6 issues). The first two dimensions are scored on a 1-4 Likert scale ("often" to "hardly ever or never"); the social support scale is scored on a scale indicating extent of agreement/disagreement ("strongly agree" to "strongly disagree”). Each dimension generates a score that allows to assessing exposure as a continuous ordinal (demand-control ratio) or categorical variable (obtained by combining the demand and control dimensions).
	The effort-reward scale consists of 23 questions with four answer choices that assess three dimensions: effort (extrinsic, 6 questions), reward (11 questions), and overcommitment (intrinsic effort, 6 questions). The dimensions are scored on a 1-5 Likert scale indicating the extent of agreement/disagreement and intensity of discomfort. Each dimension generates a score that allows to assessing exposure as a continuous ordinal (effort-reward ratio) or categorical variable (ratio is divided into tertiles).
Validity/Accuracy and Reliability
	In a Brazilian cohort study of workers (Pró-Saúde Study) (Alves et al, 2004) the internal consistency of the instrument was estimated at 0.72, 0.63 and 0.86 (Cronbach's alpha coefficient)for demand, control, and social support dimensions, respectively. The agreement estimated by intraclass correlation coefficient was 0.88, 0.87 and 0.86 for the same dimensions, respectively. A preliminary assessment of the scale factor structure and its three dimensions —demand, control and social support— was consistent with the theoretical model.
	For effort-reward scale reliability (intraclass correlation coefficient) of its three dimensions (effort, reward and overcommitment) was 0.76, 0.86, and 0.78, respectively. Estimates of internal consistency (Cronbach's alpha) for these same dimensions were 0.68, 0.78, and 0.78, respectively. The factor structure was fairly consistent with the theoretical model.

How to cite

copy

Faculdade de Saúde Pública da Universidade de São Paulo Avenida Dr. Arnaldo, 715, 01246-904 São Paulo SP Brazil, Tel./Fax: +55 11 3061-7985 - São Paulo - SP - Brazil
E-mail: revsp@usp.br

Acompanhe os números deste periódico no seu leitor de RSS

Versão para download de PDF

PDF

Português
Inglês

Versões e tradução automática

Versão original do texto

English
Português

Tradução automática

Google Translator
Microsoft Translator

Como citar

RIS

BIBTEX

Outros formatos de citação e exportação:

SciELO - Scientific Electronic Library Online
Rua Dr. Diogo de Faria, 1087 – 9º andar – Vila Clementino 04037-003 São Paulo/SP - Brasil
E-mail: scielo@scielo.org

Leia a Declaração de Acesso Aberto

Métricas

Questionnaire development in ELSA-Brasil: challenges of a multidimensional instrument

PlumX

Mensagem

[1] The authors declare that there are no conflicts of interest.