Fisberg et al. BMC Public Health  (2016) 16:93 
DOI 10.1186/s12889-016-2765-y
STUDY PROTOCOL Open Access
Latin American Study of Nutrition and
Health (ELANS): rationale and study design
M. Fisberg1,2,15*, I. Kovalskys3,4, G. Gómez5, A. Rigotti6, L. Y. Cortés7, M. Herrera-Cuenca8, M. C. Yépez9, R. G. Pareja10,
V. Guajardo3, I. Z. Zimberg2, A. D. P. Chiavegatto Filho11, M. Pratt12, B. Koletzko13, K. L. Tucker14 and the ELANS
Study Group
Abstract
Background: Obesity is growing at an alarming rate in Latin America. Lifestyle behaviours such as physical activity and
dietary intake have been largely associated with obesity in many countries; however studies that combine nutrition and
physical activity assessment in representative samples of Latin American countries are lacking. The aim of this study is to
present the design rationale of the Latin American Study of Nutrition and Health/Estudio Latinoamericano de Nutrición y
Salud (ELANS) with a particular focus on its quality control procedures and recruitment processes.
Methods/Design: The ELANS is a multicenter cross-sectional nutrition and health surveillance study of a nationally
representative sample of urban populations from eight Latin American countries (Argentina, Brazil, Chile, Colombia,
Costa Rica, Ecuador, Perú and Venezuela). A standard study protocol was designed to evaluate the nutritional intakes,
physical activity levels, and anthropometric measurements of 9000 enrolled participants. The study was based on a
complex, multistage sample design and the sample was stratified by gender, age (15 to 65 years old) and socioeconomic
level. A small-scale pilot study was performed in each country to test the procedures and tools.
Discussion: This study will provide valuable information and a unique dataset regarding Latin America that will enable
cross-country comparisons of nutritional statuses that focus on energy and macro- and micronutrient intakes, food
patterns, and energy expenditure.
Trial Registration: Clinical Trials NCT02226627
Keywords: Nutrition, Physical activity, Latin America, Cross-sectional study
Background of Latin American are overweight or obese [3]. Time trends
According to the World Health Organization, more than suggest that these figures might rise further and by 2030 up
1.4 billion adults were overweight and more than half a to 81.9 % of the Latin American and the Caribbean adult
billion were obese worldwide in 2008, and the prevalence population could be either overweight or obese [4].
of obesity nearly doubled between 1980 and 2008 [1]. These rapid epidemiological changes in the majority of
According to World Health Organization, in the Americas developing countries over the last decades have occurred
including the United States and non-continental countries, in the framework of the so-called nutritional transition.
61 % of adults are overweight or obese in 2014 [2]. In Latin The concept of the nutritional transition includes a shift
America, nearly a quarter of the population is obese, and in dietary intake and energy expenditure both of which
the prevalence has increased to a greater magnitude in are influenced by ongoing interactions between eco-
Mexico, Argentina, and Chile. A recent review estimated nomic, demographic, environmental, psychosocial and
that 20–25 % of the children and adolescents (0–18 years) cultural factors and are occurring simultaneously in
society [5]. Latin American countries are experiencing
different stages of the nutrition transition, although the
* Correspondence: mauro.fisberg@gmail.com
1 prevalence of undernutrition is declining at differentInstituto Pensi, Fundação Jose Luiz Egydio Setubal, Hospital Infantil Sabara,
São Paulo 01239-040, Brazil rates, and the prevalence of overweight is dramatically
2Universidade Federal de São Paulo, São Paulo 04023-062, Brazil increasing [6].
Full list of author information is available at the end of the article
© 2016 Fisberg et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Fisberg et al. BMC Public Health  (2016) 16:93 Page 2 of 11
Eating patterns that are characterized by high energy by region, cultural background, socioeconomic status, age
density diets, increased intakes of processed foods and gender; (3) add new scientific-based evidence to
containing large amounts of refined sugars and satu- describe the interplay among energy intake, energy expend-
rated fats, and low intakes of fiber and micronutrients iture, and anthropometric measurements. Our overarching
is considered the major preventable behavioral risk hypothesis is that the relationships between dietary and
factors for obesity [7–9]. Physical inactivity and seden- physical activity profiles will differ across countries and
tary behavior are also a preventable behavior associ- across different environmental settings.
ated with obesity, but the evidence on this issue Up to now, there is no Latin American study using a
remains mixed, [10], and the results might not be central standard methodology across a group of partici-
generalizable to all world regions [11]. Furthermore, pating countries. Previous multicenter studies [23–26]
most of the information available about the role of dietary were important references to determine the rationale
and physical activity profiles in obesity has come from and design of ELANS.
studies that have been conducted in high income, The purpose of this report is to describe the design
developed countries, but similar evidence from developing and methodology of the ELANS. The study is currently
countries is limited. In the last decade studies started to ending the fieldwork data collection.
be conducted in developing countries such as Brazil [12]
and China [13]. However, there is insufficient evidence Overall design and methods
about whether the same dietary and physical activity The ELANS is a household-based multi-national cross-
patterns and similar associations with various sociodemo- sectional survey that was conducted over a period of one
graphic conditions can be found in different regions of year in eight Latin American countries (i.e., Argentina,
middle-income countries. Brazil, Chile, Colombia, Costa Rica, Ecuador, Perú, and
Although national surveys on nutrition already exist in Venezuela; Fig. 1) that represent approximately 60 % of
some Latin American countries [14–21], the majority of the total countries of Latin America. All of the study
the data available are based on Household Consumption sites are academy (universities and other academic
and Expenditures Surveys which provide consumption data institutions)-based and adhered to a common study
at the household level and are not accurate for calculations protocol for interviewer training, implementation of field-
of individual energy intake, or representative samples of work, data collection and management, and quality control
entire population in each country were not employed. procedures that will be simultaneously performed.
An overview of the most recent nutrition surveys
conducted at individual level in nationally representative Sampling procedures
household samples of Latin America is presented in A random complex, multistage sampling of 9000 adoles-
Table 1. Only a handful of countries, however, have cents and adults aged 15.0–65.0 years, stratified by
conducted surveys with face-to-face assessment of food geographical location (only urban areas), gender, age and
intake for reasons of cost, expediency and logistics [22]. socioeconomic status was performed to select a repre-
Determinations of the dietary patterns and energy and sentative sample of the urban household population of
nutrient intakes are critical for developing dietary eight Latin American countries. Only urban areas were
recommendations and policies to address the adverse included in order to keep homogeneous population in
consequences of inappropriate dietary patterns and the research, and based on the fact that almost all
physical inactivity. This information would be of the countries have at least 80–90 % of their population living
greatest actionable value to governments, the food and in urban areas. Individual quotas were defined for each
beverage industries and agriculture. of these variables, which allow the identification of the
Studies that combine nutrition and physical activity total numbers of interviews required to properly repre-
assessment in representative samples of Latin American sent the socio-demographic distribution under study.
countries are lacking. In this direction, the Latin American The survey was designed in order that no more than
Study of Nutrition and Health/Estudio Latinoamericano de one subject were selected from a household.
Nutrición y Salud (ELANS), which is randomized cross- The sampling size was calculated with a confidence
sectional multicenter investigation of the nutritional and level of 95 % and a maximum error of 3.49 %. A survey
physical activity statuses of adolescents and adults in eight design effect of 1.75 was estimated based on guidance
Latin American countries, was designed. The study aims from the U.S. National Center for Health Statistics [27],
to (1) provide up-to-date reliable and comparable data of and calculations of the minimum sample sizes required
dietary intake, physical activity, and its association with per strata (i.e., socioeconomic level (SEL), age, and
anthropometric profile among representative urban popu- gender) were performed for each country. Table 2 shows
lations of eight Latin American countries; (2) measure a detailed description of the case numbers required per
variation in overweight, dietary intake and physical activity strata in each of the 8 study sites.
Table 1 Cross-sectional household nutrition surveys conducted in representative samples of Latin America
Country Year of data collection Sample size Sample size that underwent Method Analysis of the dietary data Reference
dietary assessment
Argentina (National Survey of 2004–2005 36,354 (aged 6 m - 5 y 36,354 24-h Recall Food Composition database [16]
Nutrition and Health - ENNyS) and women 10–49 y) developed for ENNyS
Brasil (Household Budget 2008–2009 159,941 (aged≥ 0 y) 34,003 (aged ≥ 10 y) Two 24-h recall NDSR software [44] and Food [46]
Survey - POF) Composition database
developed for POF [45]
Colombia (National Nutritional 2008–2010 162,331 (aged 0–64 y) 17,897 (aged 5–64 y) Food-Frequency Questionnaire Qualitative (daily frequency of intake) [18]
Situation Survey -ENSIN)
Chile (National Food Consumption 2014 4920 (aged ≥2 y) 4920 Quantitative Food-Frequency PC-SIDE software [14]
Survey - ENCA) Questionnaire and 24-h Recall
Ecuador (Ecuadorian National Health 2011–2013 57,727 (aged 0–59 y) 19,932 (aged 1–59 y) 24-h Recall PC-SIDE software [47]
and Nutrition Survey - ENSANUT-ECU)
México (National Health and Nutrition 2012 96,031 (aged >0 y) 10,563 to 12,484 according Semi-quantitative Food Food Composition database [49]
Survey - ENSANUT) to method used Frequency and 24-h recall in developed by National Institute
11 % and 13 % of sample, of Public Health [48]
respectively
Perú (National Survey of Nutritional, 2006 4206 (aged ≥20 y) 4206 24-h Recall ANDREA software [50], [20]
Biochemical, Socioeconomic and developed by CENAN-INS
Cultural Indicators – ENINBSC)
Venezuela (Encuesta de Seguimiento 2012-2014 20,670 (aged ≥ 3 y) 6316 participants aged≥ 3 y Diet history and food Food Composition database [51]
al Consumo de Alimentos - ESCA) frequency questionnaire developed for ESCA
Fisberg et al. BMC Public Health  (2016) 16:93 Page 3 of 11
Fisberg et al. BMC Public Health  (2016) 16:93 Page 4 of 11
Fig. 1 Latin American countries included in ELANS
Table 2 Estimated distributions of country-specific samples according to age range, gender, and SEL
Age range Gender SEL Total sample Margin
15–19.9 years 20–34.9 years 35–49.9 years 50–65 years Male Female High Middle Low size of error
(%)
Argentina 144 444 336 276 576 624 72 504 624 1200 2.83
Brazil 220 760 600 400 940 1060 520 940 540 2000 2.19
Chile 113 305 270 183 426 444 61 348 461 870 3.32
Colombia 148 443 357 283 615 615 62 357 812 1230 2.79
Costa Rica 111 300 229 150 395 395 87 450 253 790 3.49
Ecuador 128 312 224 136 400 400 104 576 120 800 3.46
Peru 165 440 308 187 528 572 209 374 517 1100 2.95
Venezuela 143 418 319 220 528 572 55 154 891 1100 2.95
Fisberg et al. BMC Public Health  (2016) 16:93 Page 5 of 11
The demographic data were obtained from the appro- As exclusion criteria were considered pregnant and
priate national statistics institutes of each country. The lactating women (in the first 6 months postpartum),
selection of the households was design in four stages. In individuals with major physical or mental impairments
the first stage, total urban population was employed to that affect food intake and physical activity (e.g., mus-
define main regions proportionally in each country first, culoskeletal disease, recent surgery, severe asthma, de-
and then select cities representing each region - main mentia, major depression), individuals below 15 or
cities and other cities representatives of the region, mix- over 65 years old, adolescents without assent and con-
ing a random method and sampling criteria, and trying sent of a parent or legal guardian, individuals living in
to fulfill urban population coverage to the maximum any residential setting other than a household (e.g.,
possible. This mixed criterion enabled an increase in the hospitals, regiments, and nursing homes), and individ-
efficiency of the fieldwork according to the study charac- uals unable to read.
teristics. In the second stage, sampling points (census
tracts) of each city were randomly selected. In the third
Ethical issues
stage, a cluster of households was selected from each
The overarching ELANS protocol was approved by the
sampling unit. Addresses were chosen systematically
Western Institutional Review Board (#20140605) and is
using standard random route procedures, beginning with
registered at Clinical Trials (#NCT02226627). Each site-
an initial address selected at random and selection of
specific protocol was also approved by the ethical review
households with three systematic jumps, that is, the
boards of the participating institutions. All participants
selection of a given household was made by randomly
provided informed consent/assent for participation in
picking the first home and subsequently skipping 3
their country-level study. Participant confidentiality for
households. Finally, in the fourth stage, selection of
the pooled data is maintained via the use of numeric
respondent within a household was performed using
identification codes rather than names. All data transfer
two criteria: in each sampling unit, half of the house-
was done with a secure file sharing system.
holds the participant was selected by the next birth-
day criteria; in the other half selection was by quotas
of gender, age, and SEL. Data collection instruments
The field interviewer arrived at the home, show an The ELANS protocol includes data collected via ques-
official identification badge, a letter introducing ELANs tionnaires and objective measurements. The question-
study, and briefly explain the survey’s purpose. A naires were administered in 2 household visits (Fig. 2).
screener questionnaire to enumerate the household and The first visit involved the selection of a respondent
to determine eligibility to participate further in ELANS within the household. Additionally, the first visit also
was fulfilled. Thereafter, the selected participant signed included application of the SEL questionnaire and 24-h
the informed consent form. dietary recall (24-h), and assessment of anthropometric
A telephone number was available for the participants measurements. In the first visit, a subsample received
in order to answer any questions about the research, and instructions regarding the use of an accelerometer with
use of the accelerometer device. a diary to be filled out for 7 consecutive days. The sec-
In the following circumstances - unsuccessful attempts ond visit was performed 8 days after the first contact for
to contact the target individual, total refusal to partici- the participants who were given accelerometers and
pate, obstruction by a family member, or inability to 4 days later for the participants who were not given
participate for a specified reason (e.g., travel, agenda, accelerometers. The second visit also included the
hospitalization) - substitutes were chosen in the home administration of a second 24-h, the IPAQ-Long
next door, following the same random selection criteria Questionnaire, and a beverage intake questionnaire
described above. (BEVQ) and the retrieval of the accelerometer.
Allocation First Visit Second Visit
8 days
Sample stratification Selection of respondent 24-hour dietary recall 
Selection of households Eligibility assessment IPAQ-long form questionnaire
Informed consent form signature Food frequency questionnaire
SEL questionnaire Accelerometers withdrawal 
24-hour dietary recall 
Anthropometric evaluation
Accelerometer use in a sub-sample
Fig. 2 Study design
Fisberg et al. BMC Public Health  (2016) 16:93 Page 6 of 11
Dietary assessment The MSM method is a web-based statistical modeling,
The 24-h method was selected because of its nearly available at MSM website (https://msm.dife.de/tps/en).
universal applicability across populations with varying
literacy skills and its relatively low burden for partici- Physical activity measures
pants [28]. Each recall was conducted using the Multiple Self-reported activities
Pass Method [29] and ascertained all foods, nonalcoholic Self-reported physical activity was assessed using the
and alcoholic beverages, water and dietary supplements International Physical Activity Questionnaire (IPAQ)-
consumed over the prior 24 h in depth. The BEVQ was long version, a validated self-report measurement tool
designed to obtain the frequency of beverage intake for physical activity in Latin America [32]. The Mexican
across 10 beverage categories (water, flavored water, soft (Spanish) version of IPAQ [33] was adapted for all coun-
drinks, fruit drinks, sport drinks, energy drinks, tea and tries of ELANS, using culturally appropriate wording
coffee drinks, other non-alcoholic drinks, and alcoholic and examples. Only the sections leisure-time and trans-
drinks). For each beverage, participants answered port physical activity (LTPA and TPA) were included,
whether they consume the specific category of beverage, due to greater importance of these domains in public
the frequency of intake (daily, weekly, monthly), and health and poor validity of the occupational and home-
how often they drink the beverage during the selected based PA IPAQ sections in Latin American urban
unit (1–10 occasions). The list of beverages included in settings. These sections are the most relevant for
the questionnaire was standardized as much as possible categorizing population levels of physical activity and for
across the ELANS countries; however, regional varia- guiding public health policies and programs [32].
tions in beverage consumption patterns required some A domain-specific activity score is calculated separ-
cultural and regional adaptations for some items within ately for each domain of physical activity (transportation
beverage categories. and leisure-time). Total times engaged in walking,
The food and beverages intakes recorded with the 24-h moderate physical activity and vigorous physical activity,
is transformed into energy, macronutrients and micro- all expressed in min/week, are scored using established
nutrients values using the Nutrition Data System for methods posted at the IPAQ website (www.ipaq.ki.se).
Research version 2013 software (NDS-R, University of Additionally, information not included as part of the
Minnesota, MN). A food matching standardized proced- summary score of physical activity, such as sedentary
ure involving nutritional equivalency of local food items activities (reading, television viewing and sitting at a
reported by the study participants into foods available desk), will be analyzed.
in NDS-R database was strictly conducted by each
country, and describe in detail elsewhere [30]. A con- Objectively assessed activity
cordance rate of at least 80 % to 120 % for energy To objectively monitor physical activity and inactivity,
and macronutrient content was required to establish 40 % of the sample was asked to wear a triaxial accelerom-
a nutritional equivalency of local food items to foods eter (model GT3X+, ActiGraph, Pensacola, FL, USA) on
available in NDS-R database. an elasticized belt at hip level on the right mid-axillary line
The Multiple Source Method (MSM) was used to esti- for 7 days. The participants were asked to wear the device
mate the usual intake of each nutrient, food and food while they are awake and to take it off for sleeping,
groups. The method was developed by researchers at the showering or swimming. Verbal (in person and by demon-
European Prospective Investigation into Cancer and stration) and written instructions on how to wear the
Nutrition (EPIC) The MSM method is a mixed model accelerometer were provided. To further ensure protocol
comprised of three parts which requires at least two days compliance, participants filled in an accelerometer log in-
of short-term dietary measurements (such as 24-h) on a dicating the start- and end-time of use per day. Following
random subsample of the target population. In the first the final day of data collection, the accelerometers was
part, the probability of consumption of a food/nutrient returned to the study sites, and the research team verified
on a day is estimated using logistic regression with ran- the data for completeness using the ActiLife software
dom effects (probability model). Second, data that has version 6 (ActiGraph, Pensacola, FL). At least 5 days of
been transformed for normality is used to estimate usual recording with a minimum of 10 or more hours of regis-
amount of food intake on days of consumption using tration per day including at least one weekend day were
linear regression, also with random effects (quantity required for data inclusion and analysis. The sampling
model). In the final part, the individual usual food/nutri- interval (epoch) was set at 30 records per second.
ent intake is calculated by multiplying the probability of
consumption of a food/nutrient (part 1) with the usual Anthropometric measurements
amount of food intake (part 2) [31]. The means and In each country, the anthropometric measurements of body
percentiles of the intakes will be estimated for each sex. weight, height and waist, hip and neck circumferences were
Fisberg et al. BMC Public Health  (2016) 16:93 Page 7 of 11
collected according to standardized procedures. The partic- performed on 50 participants from each country
ipants were measured after all heavy clothing, pocket items 2 months before starting data gathering for the full study
and shoes and socks are removed. Body weight was mea- protocol. This pre-test included all procedures from
sured with a calibrated electronic scale up to 200 kg with the selection of volunteers to the analyses of data
an accuracy of 0.1 kg. Height was measured with a portable consistency. Principal investigators of each country
stadiometer up to 205 cm with an accuracy of 0.1 cm. The provided information, their opinion and experience
measurements were taken during inspiration, with the base with regard to data collection. This pilot study allowed for
of the stadiometer lightly touching the upper reaches of the the fine-tuning of the study protocol, among other things,
head and with the participant’s head in the Frankfort Plane the pilot test allowed measurement of the time required
[34]. The circumferences were measured with an inelastic for the execution of various activities, for example the first
tape to the nearest 0.1 cm. Waist circumference was mea- and second visit, data transfer from the pollster to the in-
sured according to World Health Organization recommen- vestigator, the review and survey data entry. It also allowed
dations, i.e., with the participants standing, after a regular verification of the exactitude of the data collected and
expiration, to the nearest cm, midway between the lowest identification of the issues that should be emphasized for
rib and the iliac crest [35]. Hip circumference was recorded the second training session for pollsters, prior to the
at the level of the greatest posterior protuberance of the beginning of the study. The results of this process did not
buttocks, with the tape held horizontally flat without press- reveal the need to implement any changes in the proposed
ing the soft tissues. Neck circumference was measured at methodology. Due to the lack of security in many cities, it
the point just below the larynx (thyroid cartilage) and per- was verified the need for greater identification of inter-
pendicular to the long axis of the neck (with the tape line viewers (badges, cover letter, apron lab coat).
in the front of the neck at the same height as the tape line
at the back of the neck) [36]. Study management
The interviewers were trained to collect all measure- The management of ELANS was designed to ensure
ments by certified nutritionists/dietitians who will simul- effective collaboration and communication between the
taneously operate as supervisors of the fieldwork. Each eight study centers involved in this cross-sectional study.
measurement was repeated twice to ensure accuracy, Investigators from each participating center were in-
and the average used for the analyses. If the two readings volved in the planning and development of the protocol,
differ by more than the previously established set point which included the study design.
(0.1 kg for weight, 0.5 cm for height, 0.5 cm for neck Two chairs (from Argentina and Brazil) and a co-chair
circumferences, and 1 cm for waist and hip circumfer- (from Costa Rica) are responsible for the overall coord-
ences), then a third measurement was taken. All three ination of the study. Each study site is managed by a
measurements were recorded, and the outlier excluded local principal investigator (PI), who is responsible for
during the data cleaning process. all aspects of data collection at the local level.
The body mass index (BMI; weight (kg)/height A database center is responsible for the creation and
(m2)) and waist-to-hip ratio were calculated. The ab- management of a central database in Argentina. Two re-
solute values of each circumference measurement searchers in Brazil are responsible for the management
were compared to predefined cutoff points according of a central database and for the analysis of the acceler-
to age and gender. ometer data. A statistician and his team in Brazil are re-
sponsible for the data analysis and the generation of
Demographics and SEL preliminary reports.
A questionnaire was used to collect information about To facilitate data collection, entry and management,
demographics such as age, gender, years of education, a secured web-based system was used. Data from
number of people in the household, race/ethnicity, mari- each study site entered remotely using a standard File
tal status, and number of years living in the country. Transfer Protocol (FTP) web-browser, and the system
Socioeconomic level was also evaluated by question- allowed both the study site staff and the coordinating
naire using a format that will be country-dependent and center to monitor the progress and generate missing data
based on the legislative requirements or established local reports in real time.
standard layouts. SEL data was divided into three strata A steering committee of four external advisors with
(high, medium and low) based on the national indexes extensive and diverse scientific background was invited
used in each country [37–43]. to assess the overall progress of the study and to provide
guidance regarding the overall study direction and study
Pilot study goals. External advisor had expertise in designing multi-
To examine all procedures and verify that the planned centric epidemiological studies, nutrition and physical
activities could be adequately executed, a pilot study was activity survey, and statistical analysis.
Fisberg et al. BMC Public Health  (2016) 16:93 Page 8 of 11
With the exception of requiring that the study be Quality control strategies
global in nature, the study sponsor had or will have Quality control strategies were applied using a frame-
no role in the study design, data collection and ana- work that comprehensively considered each phase of the
lysis, decision to publish, presentation and/or manuscript study. To ensure accurate, standard and consistent mea-
preparation. surements throughout this multicenter study, a variety
Once the study is complete, the data will be available of procedures were used and are described in Table 3.
for sharing with the international scientific community. One the key aspects of the quality control was the devel-
opment of the manual of operations to keep same meth-
Personnel training odology and procedures among all sites. The crucial
To validate and harmonize the methodologies, a general roles of the PIs were focused on supervising of pilot test
training meeting was held with all of the PIs and the and fieldwork, identify and decide on the necessary
ELANS coordination center. The scientists in charge of amendments to the study protocol, coordinate all the
every research tool attended the meeting. At the end of trainings, ensuring accuracy of data entry and identifying
this training, national teams were placed in charge of inconsistencies and standardized procedures across the
translating the field protocol into the local language and sites. The coordinating center audited the complete pro-
submitting the protocol to the local ethical review cedures for the countries.
boards of the participating institutions. Each study site
was also responsible for training their personnel for the Analysis plan
pilot and field studies. The interviewers were required to The primary outcome of interest will be to describe the
have completed at secondary education. Training lasted distributions of overweight and obesity, and energy
from 7 to 10 days at each site. intake and expenditure intake across the countries and
Table 3 Quality assurance strategies
Levels of Phases
quality control Design and planning Pilot testing Data collection Data analysis
Coordinating • Critical review of protocols • Smoothness and • Monitoring field activities • Audit and evaluate
center • Harmonization of manual of feasibility of field validity of findings prior
operations for eight study sites operations assessed to publication
• Coordination of timelines and • Internal peer reviews
activities prior to publication
Principal • Review of design and planning • Audit after completion • Supervising and ensuring • Validity checks
Investigators of the study of the pilot accuracy of data entry • Results review
• Regular meetings with coordinating
center
Field Personnel • Extensive training over a period • Evaluated all field • Field coordinator will assure
of 7 to 10 days-theory and and documenting that procedures for data collections
practical-by the study managers difficulties and quality control are followed
• Additional training when necessary
Survey • Peer-reviewed • Consistency in small • Regular checks done to assess • Incomplete questionnaires
Questionnaires • Validated pilot study will be completeness identified and discarded
• Translated to local languages established
Measuring • Standardization of equipment • Evaluation of • Regular calibration of equipment;
Equipment and measurements calibration techniques, faulty equipment replaced when
• Acquired by each country acceptability of use required
• Development of anthropometric in field
procedures manual
Documentation • Assurance of standardized • Recording legibility • Audit recordings
procedures across the sites assessed
• Training in appropriate and
legible documentation
Data Storage & • Data back-up and protection • Accessibility of software • Identify inconsistencies • Datasets identified
Confidentiality policies established assessed • Corrective actions • Access to personal
• Locked and password protected identifiers
data storage limited
• Active back-up
Data Entry • Training of staff • Variability assessments • Interim analyses to identify • Reporting of outliers
• Protocols, consistent data cleaning conducted duplicate entries • Validity checks
methods and verification systems • Database errors tracked
established
Fisberg et al. BMC Public Health  (2016) 16:93 Page 9 of 11
regions of the study. A secondary objective of the study Further specific sub-projects will depend on the results
will be to identify the individual characteristics associ- and associations found for the initial analyses. Some of
ated with the two main outcomes of interest (obesity these sub-projects include analyses of micronutrient
and energy intake/expenditure) and to test whether ingestion, snack consumption and the determinants of
these associations vary across the different regions by regular physical exercise levels. Other projects will fol-
applying the statistical methodologies described below. low depending on the results and associations observed
These individual characteristics of interest will include, in the main analyses. Some of the individual variables
but are not be limited to, education, marital status, age, regarding dietary and physical activity patterns will likely
sex and individual income. The results of the three main be aggregated, depending on their interdependent rela-
outcomes will also be presented after stratification for tionships and previous studies in the literature, by apply-
each of the individual characteristics. ing factor analysis. The final datasets compiled from
Initially, basic statistical models will be applied to each country will be standardized into groups of vari-
analyze the aggregated results for all the individuals in ables and categories and sent to the coordinating center
the multicenter study to identify the individual charac- in Brazil. Descriptive statistical analyses, multilevel ana-
teristics (independent of area of residence) associated lyses, and structural equation models will be performed
with each of the outcomes, with the inclusion of all the with R 3.2.0, MLwiN 2.29, and Stata 13, respectively.
projected 9000 individuals. Separate individual models
for each of the countries will also be performed. Discussion
We will also test the presence of statistically significant The ELANS is a comprehensive cross-sectional multi-
differences between regions of residence and countries center investigation of the nutritional and physical ac-
by adjusting multilevel models (also known as hierarqui- tivity statuses of adolescents and adults in 8 Latin
cal linear models), where the individual results will be American countries.
included as the first level of the model and the region of An important strength of the ELANS study concerns
residence as the second level of the model. Overweight the number of participating countries from different
and obesity will be analyzed separately as dichotomous regions in Latin America including countries that lack
dependent variables (vs. normal weight) by adjusting data on dietary intake and physical activity level of its
logistic multilevel models, while energy intake and ex- population. In addition, the data set allows unique com-
penditure will be treated as continuous dependent vari- parisons of dietary and physical activity patterns, as all
ables and multilevel linear models will be adjusted. We measurements were obtained according to standard
will test for a statistical difference of the dependent vari- methodology and protocols in all participating countries
ables according to region of residence (a two-level and all countries started the fieldwork in very near time.
model) and also by including the country of residence as The objective measures of physical activity with acceler-
the third level of the previous multilevel models. Statis- ometer from subgroups of respondents, further enriches
tical difference between the regions and the countries the data set. Another strength of the study is the selec-
will be tested by analyzing the area-level variance (be- tion and adaptation of a food composition database to
tween-group variation) of each multilevel model. make cross-country nutritional intake comparisons.
Another objective will be to identify the pathways that Standardization at the food and nutrient levels will likely
could lead to overweight/obesity and higher energy in- minimize systematic and random errors in nutrient
take/expenditure. Some of the candidates include lower intake estimations because between-country compari-
income leading to higher consumption of nutritionally sons are particularly prone to error when different food
poor products leading to obesity, marital status leading composition tables are used to estimate dietary intake.
to higher energy intake, among many other possibilities. The study also has several potential weaknesses. The
These options will be tested by adjusting Structural ELANS is currently a cross-sectional study with all of
Equation Models (SEM), where each of the pathways the inherent limitations of this type of design. This
that lead to the dependent variables will be tested inde- means that it will be able to explore correlates of dietary
pendently for statistical significance. We will also calcu- intake, physical activity and obesity, but not its causal
late the Standardized Root Mean Square Residual determinants. One of the issues of the ELANS is the
(SRMR) for each of the final structural models to assess variations in SEL questionnaire between countries since
its goodness-of-fit, with the objective of selecting the its format was based on the legislative requirements or
models with the pathways that best fit the data. Finally, established local standard layouts. That may reduce the
given the multicentric nature of the sample, we will also validity of cross-country comparisons. The ELANS
fit structural models with clustered standard errors to countries span a wide range of health, social and economic
relax the assumption of independence of observations indicators; however, the results of the study may not be
within clusters of regions and countries. directly generalizable to other countries. Nevertheless, no
Fisberg et al. BMC Public Health  (2016) 16:93 Page 10 of 11
study has evaluated the nutritional statuses and physical de Investigación Nutricional de Peru. The funders had no role in study
activity patterns of adolescent and adult populations in design, data collection and analysis, the decision to publish, or the
preparation of this manuscript. KLT received consulting fees from the Coca
Latin American using a standardized methodology across Cola Company to participate. MF is member of the directory of Danone
a consortium of several participating countries. This study Institute International.
will provide a unique dataset from Latin America that will
Author details
enable cross-country comparisons of nutritional status 1Instituto Pensi, Fundação Jose Luiz Egydio Setubal, Hospital Infantil Sabara,
that focus on both energy intake and expenditure. The São Paulo 01239-040, Brazil. 2Universidade Federal de São Paulo, São Paulo
findings of this study should affect the planning of health 04023-062, Brazil.
3Commitee of Nutrition and Wellbeing, International Life
Science Institute (ILSI-Argentina), Buenos Aires C1059ABF, Argentina.
policies and programs that are designed to control nutri- 4Departamento Nutricion, Facultad de Ciencias Medicas, Universidad
tional inadequacies and low levels of physical activity, and Favaloro, Buenos Aires C1078AAI, Argentina. 5Departamento de Bioquímica,
their consequences, as well as the local and cultural Escuela de Medicina, Universidad de Costa Rica, San José 11501, Costa Rica.6Departamento de Nutrición, Diabetes y Metabolismo, Centro de Nutrición
adaptation of these policies and programs for implementa- Molecular y Enfermedades Crónicas, Escuela de Medicina, Pontificia
tion in Latin American countries. Universidad Católica, Santiago 833-0024, Chile. 7Departamento de Nutrición y
Bioquímica, Pontificia Universidad Javeriana, Bogotá, Colombia. 8Centro de
Competing interests Estudios del Desarrollo, Universidad Central de Venezuela (CENDES-UCV)/
Fundación Bengoa, Caracas 1010, Venezuela. 9All authors declare that they have no competing interests. Colegio de Ciencias de la
Salud, Universidad San Francisco de Quito, Quito 17-1200-841, Ecuador.
10
Authors’ contributions Instituto de Investigación Nutricional, Lima 15026, Peru.
11Departamento de
All authors were involved in the conception and design of the study. IZZ, IK Epidemiologia, Faculdade de Saúde Pública, Universidade de São Paulo, São
and MF researched the literature and drafted the manuscript. GG, AR, LYC, Paulo 01255-000, Brazil.
12Nutrition and Health Sciences Program, Hubert
MHC, MCY, RGP, VG, ADPCF, MP, BK, KLT critically reviewed the manuscript Department of Global Health, Rollins School of Public Health, Emory
and approved the final version. University, Atlanta 30322, USA.
13Division of Metabolic and Nutritional
Medicine, Dr. von Hauner Children’s Hospital, University of Munich Medical
14
Authors’ information Center, D-80337 Munich, Germany. Department of Clinical Laboratory and
The following are members of ELANS Study Group: Nutritional Sciences, University of Massachusetts Lowell, Lowell 01854, USA.15
Chairs: Mauro Fisberg and Irina Kovalskys Rua Borges Lagoa, 1080, Vila Clementino, São Paulo CEP 04038-002, Brazil.
Co-chair: Georgina Gómez Salas
Core Group members: Mauro Fisberg, Irina Kovalskys, Attilio Rigotti, Lilia Yadira Received: 9 July 2015 Accepted: 21 January 2016
Cortés Sanabria, Georgina Gómez Salas, Martha Cecilia Yépez García,
Rossina Gabriella Pareja Torres, and Marianella Herrera-Cuenca
External advisory board: Berthold Koletzko, Luis A. Moreno, Michael Pratt, and
Katherine L. Tucker References
Project Managers: Viviana Guajardo and Ioná Zalcman Zimberg 1. Stevens G. Global health risks: progress and challenges. Bull World Health
International Life Sciences Institute (ILSI) – Argentina: Irina Kovalskys, Organ. 2009;87:646.
Viviana Guajardo, Maria Paz amigo, Ximena Janezic, and Fernando Cardini 2. World Health Organization (WHO). Global status report on
Instituto Pensi- Hospital Infantil Sabara – Brazil: Mauro Fisberg, Ioná Zalcman noncommunicable diseases 2014. Geneva: World Health Organization
Zimberg, and Natasha Aparecida Grande de França (WHO); 2014.
Pontificia Universidad Católica de Chile: Attilio Rigotti, Guadalupe Echeverría, 3. Rivera JA, de Cossío TG, Pedraza LS, Aburto TC, Sánchez TG, Martorell R.
Leslie Landaeta, and Óscar Castillo Childhood and adolescent overweight and obesity in Latin America: a
Pontificia Universidad Javeriana - Colombia: Lilia Yadira Cortés Sanabria, systematic review. Lancet Diabetes Endocrinol. 2014;2:321–32.
Luz Nayibe Vargas, Luisa Fernanda Tobar, and Yuri Milena Castillo doi:10.1016/S2213-8587(13)70173-6.
Universidad de Costa Rica: Georgina Gómez, Rafael Monge Rojas, and Anne 4. Kelly T, Yang W, Chen CS, Reynolds K, He J. Global burden of obesity in
Chinnock 2005 and projections to 2030. Int J Obes (Lond). 2008;32:1431–7.
Universidad San Francisco de Quito - Ecuador: Martha Cecilia Yépez García, doi:10.1038/ijo.2008.102.
María Elisa Herrera Fontana, Mónica Villar Cáceres, and María Belén Ocampo 5. Barria RM, Amigo H. Nutrition transition: a review of Latin American profile.
Instituto de Investigación Nutricional - Peru: Rossina Pareja Torres, Maria Reyna Arch Latinoam Nutr. 2006;56:3–11.
Liria, Krysty Meza, Mellisa Abad, and Mary Penny 6. Rivera JA, Barquera S, Gonzalez-Cossio T, Olaiz G, Sepulveda J. Nutrition
Universidad Central de Venezuela: Marianella Herrera-Cuenca, Maritza Landaeta, transition in Mexico and in other Latin American countries. Nutr Rev.
Betty Méndez, Maura Vasquez, Guillermo Ramirez, and Pablo Hernández 2004;62:S149–57.
Statistical analysis: Alexandre DP Chiavegatto Filho 7. Swinburn BA, Sacks G, Hall KD, McPherson K, Finegood DT, Moodie ML, et
Accelerometry analysis: Priscila Bezerra Gonçalves and Claudia Alberico al. The global obesity pandemic: Shaped by global drivers and local
Physical activity advisor: Gerson Luis de Moraes Ferrari environments. Lancet. 2011;378:804–14. doi:10.1016/S0140-6736(11)60813-1.
Dietary intake advisor: Ágatha Nogueira Previdelli 8. Bermudez OI, Tucker KL. Trends in dietary patterns of Latin
American populations. Cad Saude Publica. 2003;19 Suppl 1:S87–99.
Acknowledgments doi:10.1590/S0102-311X2003000700010.
We would like to thank the following individuals at each of the participating 9. Popkin BM, Adair LS, Ng SW. Global nutrition transition and the pandemic
sites who made substantial contributions to the ELANS: Luis Moreno, Beate of obesity in developing countries. Nutr Rev. 2012;70:3–21.
Lloyd, Brenda Lynch, Mariela Jauregui, Alejandra Guidi, Luis Costa, and 10. Pearson N, Braithwaite RE, Biddle SJH, van Sluijs EMF, Atkin AJ. Associations
Regina Fisberg. between sedentary behaviour and physical activity in children and
adolescents: A meta-analysis. Obes Rev. 2014;15:666–75. doi:10.1111/obr.12188.
Funding 11. Joint WHO/FAO Expert Consultation. Diet, nutrition and the prevention of
The ELANS study and authors were partially supported by a scientific grant chronic diseases, WHO technical report series, No. 916. vol. 916. Geneva:
from the Coca Cola Company and by different grants and support from the Joint WHO/FAO Expert Consultation; 2003.
Instituto Pensi/Hospital Infantil Sabara, International Life Science Institute of 12. Olinto MTA, Willett WC, Gigante DP, Victora CG. Sociodemographic
Argentina, Universidad de Costa Rica, Pontificia Universidad Católica de Chile, and lifestyle characteristics in relation to dietary patterns among
Pontificia Universidad Javeriana, Universidad Central de Venezuela (CENDES- young Brazilian adults. Public Health Nutr. 2011;14:150–9.
UCV)/Fundación Bengoa, Universidad San Francisco de Quito, and Instituto doi:10.1017/S136898001000162X.
Fisberg et al. BMC Public Health  (2016) 16:93 Page 11 of 11
13. Lv J, Liu Q, Ren Y, Gong T, Wang S, Li L. Socio-demographic association 34. Lohman TG, Roche AF, Martorell R. Anthropometric standardization
of multiple modifiable lifestyle risk factors and their clustering in a reference manual. vol. 24. 3rd ed. Champaign: Human Kinetics Press; 1988.
representative urban population of adults: a cross-sectional study 35. World Health Organization (WHO). Waist Circumference and Waist-Hip Ratio:
in Hangzhou. China Int J Behav Nutr Phys Act. 2011;8:40. Report of a WHO Expert Consultation. Geneva: World Health Organization; 2008.
doi:10.1186/1479-5868-8-40. 36. Cornier M-A, Després J-P, Davis N, Grossniklaus DA, Klein S, Lamarche B, et al.
14. de Salud de Chile M. Encuesta Nacional de Consumo Alimentario. 2014. Assessing adiposity: a scientific statement from the american heart association.
Santiago. Circulation. 2011;124:1996–2019. doi:10.1161/CIR.0b013e318233bc6a.
15. Instituto Brasileiro de Geografia e Estatística. Pesquisa de Orçamentos 37. Associação Brasileira de Empresas de Pesquisa (ABEP). Critério padrão de
Familiares: 2008–2009. Rio de Janeiro: Análise do Consumo Alimentar classificação econômica Brasil. 2013.
Pessoal no Brasil; 2011. 38. Comisión de Enlace Institucional AAM-SAIMO-CEIM. Nivel Socioeconómico.
16. Ministerio de Salud de Argentina. Alimentos Consumidos en Argentina. Antecedentes, marco conceptual, enfoque metodológico y fortalezas.
Resultados de la Encuesta Nacional de Nutrición y Salud -ENNyS 2004/5. Buenos Aires: Comisión de Enlace Institucional AAM-SAIMO-CEIM; 2006.
Buenos Aires: Ministerio de Salud de Argentina; 2012. 39. Departamento Administrativo Nacional de Estadisticas de Colombia.
17. Ministerio del la Salud Publica de Ecuador. Encuesta nacional de salud y Proyecciones nacionales y departamentales de poblacion 2005–2020,
nutricion, ENSANUT 2011–2013. Quito: Ministerio del la Salud Publica de Estudios Postcensales No. 7. 2009.
Ecuador; 2013. 40. Instituto Nacional de Estadística y Censos de Ecuador. Encuesta de
18. Fonseca Z, Heredia A, Ocampo R, Forero Y, Sarmiento O, Álvarez M, et al. Estratificación de Nivel Socioeconómico. 2011.
Encuesta Nacional De La Situación Nutricional En Colombia 2010 - ENSIN. 41. IPSOS. Estudio general de medios. 2012.
Instituto Colombiano de Bienestar Familiar: Bogotá; 2011. 42. Asociacion Investigadores de Mercado. Grupos Socioeconómicos Chile.
19. Costa Rica. Ministerio de Salud. Encuesta nacional de nutrición 2008–2009. Chile: Asociacion Investigadores de Mercado; 2012.
San José: Costa Rica. Ministerio de Salud; 2012. 43. Instituto Nacional de Estadística. República Bolivariana de Venezuela,
20. Peru. Ministerio de Salud. Encuesta nacional de indicadores nutricionales, Síntesis Estadística de Pobreza e Indicadores de Desigualdad. 2011.
bioquímicos, socioeconómicos y culturales relacionados con las 44. University of Minnesota. Nutrition Coordinating Center, Nutrition data
enfermedades crónicas degenerativas. Lima: Peru. Ministerio de Salud; 2006. system for research (NDSR). 2008.
21. Instituto Nacional de Estadística de Venezuela. Encuesta Nacional de 45. Núcleo de Estudos e Pesquisas em Alimentação. Tabela brasileira de
Consumo de Alimentos (ENCA) 2013. Caracas: Instituto Nacional de composição de alimentos. 4th ed. 2011.
Estadística de Venezuela; 2014. 46. Instituto Brasileiro de Geografia e Estatística. Pesquisa de Orçamentos Familiares:
22. Ferro-Luzzi A. Keynote Paper: Individual food intake survey methods. Meas. 2008–2009. Rio de Janeiro: Antropometria e Estado Nutricional; 2010.
Assess. Food Deprivation Undernurition. Rome: Food and Agriculture 47. Freire W, Ramírez-Luzuriaga M, Belmont P, Mendieta M, Silva-Jaramillo M,
Organization of the United Nations 2003. Romero N, et al. Tomo I: Encuesta Nacional de Salud y Nutrición de la
23. Moreno LA, De Henauw S, González-Gross M, Kersting M, Molnár D, población ecuatoriana de cero a 59 años, ENSANUT-ECU 2012. Quito:
Gottrand F, et al. Design and implementation of the healthy lifestyle in ENSANUT-ECU; 2014.
Europe by nutrition in adolescence cross-sectional study. Int J Obes. 48. Safdie M, Barquera S, Porcayo M, Rodríguez S, Ramírez C, Rivera J, et al.
2008;32 Suppl 5:S4–11. doi:10.1038/ijo.2008.177. Bases de datos del valor nutritivo de los alimentos, Compilación del
24. Katzmarzyk PT, Barreira TV, Broyles ST, Champagne CM, Chaput J-P, Instituto Nacional de Salud Pública. 2004.
Fogelholm M, et al. The International Study of Childhood Obesity, Lifestyle 49. Romero-Martínez M, Shamah-Levy T, Franco-Núñez A, Villalpando S, Cuevas-
and the Environment (ISCOLE): design and methods. BMC Public Health. Nasu L, Gutiérrez JP, et al. Encuesta Nacional de Salud y Nutrición 2012:
2013;13:900. doi:10.1186/1471-2458-13-900. diseño y cobertura. Salud Publica Mex. 2013;55:S332–40.
25. van Stralen MM, te Velde SJ, Singh AS, De Bourdeaudhuij I, Martens MK, van 50. Asociación Benéfica PRISMA. Análisis Nutricional de la Dieta según
der Sluis M, et al. EuropeaN Energy balance Research to prevent excessive Requerimientos y Adecuación (ANDREA®). Lima: Asociación Benéfica
weight Gain among Youth (ENERGY) project: Design and methodology of PRISMA; 2003.
the ENERGY cross-sectional survey. BMC Public Health. 2011;11:65. 51. Venezuela. Instituto Nacional de Estadística. Encuesta de Seguimiento al
doi:10.1186/1471-2458-11-65. Consumo de Alimentos (ESCA). Caracas: Venezuela. Instituto Nacional de
26. Kerr J, Sallis JF, Owen N, De Bourdeaudhuij I, Cerin E, Sugiyama T, et al. Estadística; 2014.
Advancing science and policy through a coordinated international study of
physical activity and built environments: IPEN adult methods. J Phys Act
Health. 2013;10:581–601.
27. National Center for Health Statistics (NCHS). Analytic and reporting
guidelines: the third national health and nutrition examination survey,
NHANES III (1988–94), Prevention. 1996. p. 1–47.
28. Dennis B, Stamler J, Buzzard M, Conway R, Elliott P, Moag-Stahlberg A, et al.
INTERMAP: the dietary data–process and quality control. J Hum Hypertens.
2003;17:609–22. doi:10.1038/sj.jhh.1001604.
29. Moshfegh AJ, Rhodes DG, Baer DJ, Murayi T, Clemens JC, Rumpler WV, et al.
The US department of agriculture automated multiple-pass method
reduces bias in the collection of energy intakes. Am J Clin Nutr.
2008;88:324–32.
30. Kovalskys I, Fisberg M, Gómez G, Rigotti A, Cortés LY, Yépez MC, et al. Submit your next manuscript to BioMed Central 
Standardization of the food composition database used in the Latin and we will help you at every step:
American nutrition and health study (ELANS). Nutrients. 2015;7:7914–24.
doi:10.3390/nu70x000x. •  We accept pre-submission inquiries 
31. Harttig U, Haubrock J, Knüppel S, Boeing H. The MSM program: •  Our selector tool helps you to find the most relevant journal
web-based statistics package for estimating usual dietary intake
using the multiple source method. Eur J Clin Nutr. 2011;65 Suppl 1:S87–91. •  We provide round the clock customer support 
doi:10.1038/ejcn.2011.92. •  Convenient online submission
32. Hallal PC, Gomez LF, Parra DC, Lobelo F, Mosquera J, Florindo AA, et al. •  Thorough peer review
Lessons learned after 10 years of IPAQ use in Brazil and Colombia. J Phys
Act Health. 2010;7 Suppl 2:S259 64. •  Inclusion in PubMed and all major indexing services –
33. Salvo D, Reis RS, Sarmiento OL, Pratt M. Overcoming the challenges of •  Maximum visibility for your research
conducting physical activity and built environment research in Latin
America: IPEN Latin America. Prev Med. 2014;69:S86–92. Submit your manuscript at
www.biomedcentral.com/submit