Diversity of eating patterns and obesity in older adults—A new challenge

doi:10.4236/health.2013.58A3004

Paper Menu >>

Journal Menu >>

Vol.5, No.8A3, 23-28 (2013) Health

http://dx.doi.org/10.4236/health.2013.58A3004

Diversity of eating patterns and obesity in older

adults—A new challenge

Patricia Moraes Ferreira1, Silvia Justina Papini2, José Eduardo Corrente3*

1PhD Student, Department of Public Health, School of Medicine, São Paulo, Brazil

2Nursing Department, School of Medicine, São Paulo, Brazil

3Biostatistics Department, Bioscience Institute, São Paulo, Brazil;

*Corresponding Author: [email protected]

Received 19 June 2013; revised 19 July 2013; accepted 29 July 2013

License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

ABSTRACT

The increase in the variety of food choices in-

fluences the eating patterns of older adults,

which is in turn increases the occurrence of

obesity. This study aimed at identifying eating

patterns and their association with obesity in a

represent ati ve sample of older adults living in an

urban area and registered in the basic health

unit in the city of Botucatu, São Paulo, Brazil.

This is a cross-sectional study and data collec-

tion took place from March to June of 2011

through the application of a validated food fre-

quency questionnaire for older adults, a socio-

demographic survey and an anthropo metric eva-

luation. Eating patterns were identified through

principal component analysis. Scores of indi-

vidual consumption were divided in tertiles,

characterizing as low, moderate or high adhe-

rence of the individuals to each pattern. Logistic

regression models were fitted for the outcomes

“general obesity” and “abdominal obesity” and

the tertiles of consumption adjusting by socio-

demographic variables. Six eating patterns were

identified: Healthy foods, Snacks and weekend

meals, Fruits, Light and whole foods, Mild diet

and Traditional diet. It was found that the ad-

herence to healthy foods is protective against

obesity as well as adherence of snacks and

weekend meals are risk of obesity. Eating pat-

terns and their recognized influence on obesity

comprise an issue that deserves continuous

attention in order to evaluate collectively the

eating profile, and develop specific nutritional

guidelines for older adults.

Keywords: Eating Patterns; General Obesity;

Older Adults

1. INTRODUCTION

In Brazil, the most recent research regarding the nutri-

tional state of the population showed that, between 1974

and 2009, the prevalence of older individuals of below

average weight decreased, whereas the number of older

individuals who are overweight or obese has increased

steadily [1]. The rise in the number of obese older adults

may be associated with changes in nutrition in this age

group at a time when an increased consumption of foods

with high energy density and a reduction in the ingestion

of foods rich in fiber, and nutrients are becoming more

prevalent in society [2].

The importance of eating behaviors as a risk factor for

overweight and obesity is already well known [3,4]. The

relationship between eating and nutritional status in the

context of aging is frequently associated with nutritional

problems [5] and boredom with eating [6,7], since

physiological, economic and psychosocial factors may

limit the consumption of food [7-9]. Nevertheless, it may

be assumed that the older population, in addition to being

vulnerable, is also a heterogeneous group in terms of

many different aspects, including eating [9]. Due to the

increase in the variety of food choices offered by the

food industry and commerce, diversity in eating patterns

is emerging in the same population group.

In terms of epidemiology, the use of empirical analy-

ses of derivative standards may reflect the true eating

behaviors of a population and allow for improved plan-

ning and effective proposals to promote healthy eating

habits [10]. Therefore, from the hypothesis that the eat-

ing habits of older adults are diversifying and exerting a

strong influence on the increased prevalence of obesity,

this study aimed at identifying eating patterns of the older

P. M. Ferreira et al. / Health 5 (2013) 23-28

people and investigating their association with obesity.

2. METHODOLOGY

2.1. Study Population and Data Collection

The present study was carried out in the city of Botu-

catu, São Paulo, Brazil, which was chosen, in part, due to

the development of a growing number of studies on

population aging and to its high prevalence of older

adults (13.35%) [11], higher than the mean in the state

(11.6%) and in the country (10.8%) [12]. It is an epide-

miological cross-sectional study, with a representative

sample of individuals aged 60 years or older, living in the

urban area and registered in the basic health unit of the

municipality.

A validated food frequency questionnaire (FFQ) with

71 food items was provided to the participants [13]. The

sample size was calculated by considering five times the

number of items present in the FFQ, according to the

formula: if k > 15 ≥ n = 5 × k where k = number of items

of the FFQ [14], yielding a total of 355 individuals, who

were stratified among the sixteen basic health units in the

municipality. For the association study between general

obesity and the identified dietary patterns, we performed

a posteriori sample size calculation.

Data collection took place at households and at the ba-

sic health units from March to June of 2011. Also it was

applied a sociodemographic and lifestyle identification

questionnaire and anthropometric evaluation was made.

The weight it was measured using a calibrated portable

digital electronic scale (TANITA® TM), portable stadi-

ometer (ALTUREXATA®TM) for height. The waist cir-

cumference measurement was performed using inelastic

tape with 1 mm precision. Failing to perform the meas-

urement of height (older adults stooped posture), these

values were estimated from the measurement of the knee

based on the validated techniques [15], and waist cir-

cumference according to the techniques recommended

by the World Health Organization [16].

2.2. Ethical Approval

The study was approved by the Committee of Ethics in

Research of the Botucatu Medical School—São Paulo

State University (UNESP).

2.3. Data Analysis

Obesity was evaluated as either “general obesity” or

“abdominal obesity”. In order to define general obesity,

the authors used the Body Mass Index (BMI), with obe-

sity defined as ≥30 kg/m2; for abdominal obesity, a waist

Circumference (WC) ≥88 cm for women and ≥102 cm

for men were used as the defining criteria. Both criteria

are in agreement with recommendations from the World

Health Organization [16]. Other variables analyzed were:

consumption of alcoholic beverages, gender, education,

marital status and physical activity. The criteria estab-

lished by VIGITEL 2010 were used for the categoriza-

tion of physical activity [17].

The information regarding food consumption obtained

from the Food Frequency Questionnaire (FFQ) was ana-

lyzed and interpreted for the identification of eating pat-

terns, using the technique of exploratory factor analysis

(principal component analysis—PCA). The applicability

of the factor analysis was confirmed by the PCA method

with varimax rotation, through Kaiser-Meyer-Olkin

(KMO) and Bartlett’s tests of sphericity. Six factors were

extracted considering only the items with a factor load

greater than 0.3. Eating patterns were determined ac-

cording to the nutritional and functional characteristics of

the food items for each factor.

Following the factor analysis, consumption scores

were calculated in order to obtain participants’ adherence,

which were divided into tertiles classified as low (1st

tertile), moderate (2nd tertile) and high (3rd tertile) com-

pliance. Logistic regression models were fitted consider-

ing “general” and “abdominal” obesity as response vari-

ables and adherence to the eating patterns (low, moderate

and high) as explanatory variables adjusted for possible

confounding effects. All analyses were performed using

the SAS (Statistical Analysis System) program for Win-

dows (v. 9.2). For all tests, a significance level of 5% (or

the corresponding p-value) was considered.

3. RESULTS AND DISCUSSION

The study sample consisted of 163 male (45.9%) and

192 female individuals (54.1%), which is close to the

Brazil 2010 census results, in which 42.7% were men

and 57.3% were women [18], indicating that this is a

representative sample of the city’s population.

Age varied between 60 and 92 years, with a mean of

69.54 years old (standard deviation = 7.73 years old);

this population has a mean per capita family income of

1.76 minimum wages (approximately 500 US dollars).

The prevalence of general obesity was 15.95% in men

and 30.20% in women, while the prevalence of abdomi-

nal obesity was much higher: 42.94% in men and

74.47% in women.

The patterns identified through factor analysis were:

1) Healthy foods: raw endive/chicory; beetroot/chay-

ote/zucchini; broccoli/cauliflower/cabbage; cooked en-

dive/kale; carrots; extra virgin olive oil; tomatoes; lettuce;

fish; oats.

2) Snacks and weekend meals: processed meat; moz-

zarella/cheddar cheese; pizza/pancakes; baked snacks;

bacon/jerky; hamburger/chicken nuggets/meatballs; fried

snacks; regular butter; regular soda; bread rolls; pasta

with meat; cooked potatoes with mayonnaise; desserts/

P. M. Ferreira et al. / Health 5 (2013) 23-28

candies; fried potatoes/manioc.

3) Fruits: avocado; guava; papaya; apple/pears; melon/

watermelon; oranges/tangerines/pineapples; bananas.

4) Light and whole food: low-fat/non-fat milk; whole

wheat bread; natural juice without added sugar; oats; ex-

tra virgin olive oil.

5) Mild diet: cooked potatoes/manioc; soup; bread

rolls; whole milk; carrots; polenta.

6) Traditional diet: white rice; beans; lettuce; toma-

toes.

Until now, there have been few studies that have spe-

cifically included older adults and that have used statis-

tical models to identify eating patterns empirically [19-

21]. In Brazil, these statistical analyses have not still in-

cluded samples confined to older adults.

The knowledge of specific eating behaviors of older

adults is essential, since they comprise a group that is

highly vulnerable to nutritional problems, one in which

repercussions is much more severe than in other stages of

life [9].

Through the habits identified, it was possible to ob-

serve the existence of different consumption conditions

occurring in the eating culture of this population. Con-

trary to what was observed a few years ago in Brazil, the

older population does not seem to be limiting itself to the

consumption of a boredom diet, generally characterized

by the predominance of traditional food items of Brazil-

ian culture, such as rice and beans. The older adults in

this study also showed a preference for other eating pat-

terns which may reflect the local culture (characterized

by the consumption of pasta, cooked potatoes with may-

onnaise and desserts on the weekends), the Western diet

(characterized by a high concentration of carbohydrates

and fat), special diets (characterized by a consumption of

healthier food items, such as diet/light foods), and diets

consumed by individuals of a more advanced age (char-

acterized by a consumption of mild diet, with cooked and

easily digested food items).

It was observed that the high compliance with Stan-

dard 1—Healthy foods had an inverse and significant

relationship to general obesity, constituting a protective

factor. In other words, individuals showing high compli-

ance with this standard decreased their chance of devel-

oping general obesity by 63.3% (Table 1).

Table 1. Association between eating patterns and general obesity in older adults, according to the variables of interest (adjusted

model*), Botucatu, São Paulo, Brazil, 2011.

Variables Categories Estimate Standard errorp-value OR (IC 95%)**

High compliance −0.6071 0.2134 0.0044 0.367 (0.179 - 0.752)

Moderate compliance 0.2112 0.1932 0.2742 0.831 (0.434 - 1.592) 1-Healthy foods

Low compliance 0.0 - - 1.0

High compliance −0.00364 0.2165 0.9866 1.104 (0.514 - 2.370)

Moderate compliance 0.1059 0.1964 0.5897 1.231(0.612 - 2.478)

2-Snacks and weekend

meals

Low compliance 0.0 - - 1.0

High compliance 0.3048 0.2048 0.1367 1.334 (0.660 - 2.698)

Moderate compliance −0.3211 0.2133 0.1323 0.714 (0.343 - 1.484) 3-Fruits

Low compliance 0.0 - - 1.0

High compliance 0.1981 0.2018 0.3263 1.485 (0.728 - 3.029)

Moderate compliance −0.00053 0.2050 0.9979 1.218 (0.591 - 2.510)

4-Light and whole foods

Low compliance 0.0 - - 1.0

High compliance −0.0904 0.2069 0.6621 0.795 (0.393 - 1.610)

Moderate compliance −0.0480 0.1987 0.8090 0.830 (0.422 - 1.634)

5-Mild diet

Low compliance 0.0 - - 1.0

High compliance −0.3991 0.2113 0.0589 0.625 (0.307 - 1.270)

Moderate compliance 0.3275 0.1931 0.0899 1.292 (0.676 - 2.467)

6-Traditional diet

Low compliance 0.0 - - 1.0

Age - −0.0591 0.0236 0.0122 0.943 (0.900 - 0.987)

Partner 0.3910 0.1667 0.0190 2.186 (1.137 - 4.202)

Marital Status No partner 0.0 - - 1.0

Consumption 0.2046 0.1678 0.2225 1.506 (0.780 - 2.906)

Alcohol consumption No consumption 0.0 - - 1.0

*Logistic model with adjustment for gender, per capita family income, education and physical activity; obtained from full model, keeping all variables of inter-

st. **OR = Adjusted odds ratio. e

P. M. Ferreira et al. / Health 5 (2013) 23-28

Regarding abdominal obesity, it was observed that in-

dividuals who comply moderately with Standard 1—

Healthy foods decrease their risk of developing abdomi-

nal obesity by 41.2%. On the other hand, moderate com-

pliance with Standard 2—Snacks and weekend meals

increased 2.21 times the chance of developing abdominal

obesity (Table 2).

These results agree with the recommendations, since

Standard 1—Healthy foods has a high concentration of

protective elements for weight gain, while Standard 2—

Snacks and weekend meals has a high concentration of

fat and simple carbohydrates which, if ingested exces-

sively, can influence the development of obesity. Similar

eating patterns and results have been found in other

studies [22,23]. A standard termed “Healthy”, described

by Americans in the Baltimore Longitudinal Study of

Aging, was associated with lower gains in body mass

index and waist circumference [22]. A standard termed

“Meat and French Fries”, identified in Puerto Rican

adults and older adults, was associated with higher meas-

ures of waist circumference [23]. This similarity between

studies shows the consistency of the empirical methods

used to identify eating patterns.

The other identified eating patterns did not show a

significant association with obesity.

Some limitations of this study, which are common to

most studies regarding food consumption, must be con-

sidered. The first one refers to the cross-sectional design,

which does not allow for the establishment of a causal

relationship between risk factors and health outcomes.

However, since the purpose was to describe eating be-

haviors in order to guide immediate actions of health

promotion, the priority was not to determine whether the

relationship between food consumption and negative

health outcomes is indeed casual.

The influence of reverse causality in studies of obesity

is broadly recognized [24,25], since individuals may ei-

ther overestimate or underestimate the consumption of

certain food items. On the other hand, in this study the

results of associations between eating patterns and obe-

sity assumed the expected direction, and the control of

covariates was performed, increasing the probability

Table 2. Association between eating patterns and abdominal obesity in older adults, according to the variables of interest (adjusted

model*), Botucatu, São Paulo, Brazil, 2011.

Variables Categories Estimate Standard errorp-value OR (IC 95%)**

High compliance 0.2062 0.1781 0.2469 1.045 (0.570 - 1.914)

Moderate compliance −0.3686 0.1746 0.0348 0.588 (0.325 - 1.064)

1-Healthy foods

Low compliance 0.0 - - 1.0

High compliance −0.0935 0.1824 0.6082 1.293 (0.677 - 2.469)

Moderate compliance 0.4436 0.1808 0.0141 2.212 (1.164 - 4.203)

2-Snacks and weekend

meals

Low compliance 0.0 - - 1.0

High compliance 0.0981 0.1818 0.5895 1.455 (0.779 - 2.716)

Moderate compliance 0.1788 0.1793 0.3187 1.577 (0.852 - 2.920) 3-Fruits

Low compliance 0.0 - - 1.0

High compliance 0.1306 0.1818 0.4726 1.045 (0.562 - 1.943)

Moderate compliance −0.2172 0.1838 0.2373 0.738 (0.394 - 1.382) 4-Light and whole foods

Low compliance 0.0 - - 1.0

High compliance 0.1292 0.1805 0.4740 1.047 (0.565 - 1.939)

Moderate compliance −0.2130 0.1761 0.2265 0.743 (0.407 - 1.357) 5-Mild diet

Low compliance 0.0 - - 1.0

High compliance −0.2674 0.1755 0.1276 0.593 (0.325 - 1.081)

Moderate compliance 0.0115 0.1763 0.9478 0.783 (0.428 - 1.433) 6-Traditional diet

Low compliance 0.0 - - 1.0

Age - −0.0120 0.0190 0.5279 0.988 (0.952 - 1.026)

Partner 0.1793 0.1447 0.2152 1.431 (0.812 - 2.524)

Marital status No partner 0.0 - - 1.0

Consumption 0.1373 0.1464 0.3482 1.316 (0.741 - 2.336)

Alcohol consumption No consumption 0.0 - - 1.0

*Logistic model with adjustment for gender, per capita family income, education and physical activity; obtained from full model, keeping all variables of inter-

est. **OR = Adjusted odds ratio.

P. M. Ferreira et al. / Health 5 (2013) 23-28 27

that the associations found are valid.

In order to reduce the bias of memory, the assistance

of caregivers was permitted to answer the questionnaires,

and the questions from the Food Frequency Question-

naire (FFQ) referred only to frequency of consumption,

and did not include the portions consumed. Furthermore,

the apparent effect of report errors in the FFQ is fre-

quently discussed, since people are inclined to overesti-

mate, for instance, their consumption of vegetables [25].

However, a study observed that eating patterns extracted

from the FFQ are comparable to those obtained from the

24 h recall (the instrument which is considered to be the

standard) [26].

Since the eating patterns are derived empirically, it is

possible that other combinations of food items exist.

Nevertheless, in the present study, the authors chose the

best factor solution and verified the quality of interpreta-

tion of the eating standards using statistical criteria, de-

fining which factor solution was closer to the combina-

tion of food items observed among the sample individu-

als.

4. CONCLUSIONS

The authors concluded that there was a diversity of

eating patterns within a population of older adults, and

this current eating behavior was found to be a factor in-

dependently associated with obesity.

The results obtained support the premise that the eat-

ing standards and their recognized influence on obesity

are issues that deserve continuing attention in order to

evaluate collectively the eating profile of the older popu-

lation and orientate efforts for the development of spe-

cific nutritional guidelines for the older population.

5. ACKNOWLEDGEMENTS

The authors would like to thank to São Paulo Foundation Research

(FAPESP-Process no. 2010/12366-1) and National Counsel of Techno-

logical and Scientific Development (CNPq Process no. 301197/2011-3)

for the financial support.

REFERENCES

[1] Instituto Brasileiro de Geografia e Estatística—IBGE.

(2011) Análise do consumo alimentar pessoal no Brasil.

POF 2008/2009, Rio de Janeiro, 2011.

[2] Popkin, B.M. (1994) The nutrition transition in low-in-

come countries: An emerging crisis. Nutrition Reviews,

52, 285-298. doi:10.1111/j.1753-4887.1994.tb01460.x

[3] Ma, Y., Bertone, E.R, Stanek, E.J., Reed, G.W, Hebert,

J.R, Cohen, N.L., Merriam P.A. and Ockene I.S. (2003)

Association between eating patterns and obesity in a

free-living US adult population. American Journal of

Epidemiology, 158, 85-92. doi:10.1093/aje/kwg117

[4] Mccrory, M.A., Suen, V.M. and Roberts, S.B. (2002) Bio-

behavioral influences on energy intake and adult weight

gain. Journal of Nutrition, 132, 3830-3834.

[5] Ministério da Saúde. (2006) Obesidade: Determinantes

do sobrepeso e obesidade. Brasília.

[6] Freitas, A.M.P., Philippi, S.T. and Ribeiro, S.M.L. (2011)

Listas de alimentos relacionadas ao consumo alimentar de

um grupo de idosos: Análises e perspectivas. Revista

Brasileira de Epidemiologia, 14, 161-177.

doi:10.1590/S1415-790X2011000100015

[7] Campos, M.T.F.S., Monteiro, J.B.R. and Ornelas, A.P.R.C.

(2000) Fatores que afetam o consumo alimentar e a

nutrição do idoso. Revista de Nutrição, 13, 157-165.

doi:10.1590/S1415-52732000000300002

[8] Gollub, E.A. and Weddle, D.O. (2004) Improvements in

nutritional intake and quality of life among frail home-

bound older adults receiving home-delivered breakfast

and lunch. The Journal of the American Dental Asso-

ciation, 104, 1227-1235.

doi:10.1016/j.jada.2004.05.204

[9] Arbonés, G., Carbajal, A., Gonzalvo, B., Gonzales-Gróss,

M., Joyanes, M., Marques-Lopes, I., Martín, M.L., Mar-

tínez, A., Montero, P., Núñes, C., Puigdueta, I., Quer, J.,

Rivero, M., Roset, M.A., Sánchez-Muniz, F.J. and Vaque-

ro, M.P. (2003) Nutrición y recomendaciones dietéticas

para personas mayores. Grupo de trabajo “Salud pública”

de La Sociedad Espanhola de Nutrición (SEN). Nutrición

Hospitalaria, 18, 109-137.

[10] Newby, P.K. and Tucker, K.L. (2004) Empirically derived

eating patterns using factor or cluster analysis: A review.

Nutrition Reviews, 62, 177-203.

doi:10.1111/j.1753-4887.2004.tb00040.x

[11] Departamento de Informática do SUS—DATASUS. (2011)

Informações de Saúde.

[12] Instituto Brasileiro de Geografia e Estatística—IBGE.

(2011) Censo demográfico.

[13] Corrente, J.E., Marchioni, D.M.L., Fisberg, R.M. (2013)

Validation of a FFQ (Food Frequency Questionaire) for

older people. Journal of Life Sicence, in press.

[14] Hair, J.F., Anderson, R.E., Tatham, R.L. and Black, W.C.

(2005) Análise multivariada de dados. Artmed, Porto

Alegre.

[15] Chumlea, W.C., Roche, A.F. and Steinbaugh, M.L. (1985)

Estimating stature from knee height for persons 60 to 90

years of age. Journal of the American Geriatrics Society,

33, 116-120.

[16] World Health Organization (WHO). (1998) Obesity: Pre-

venting and managing the global epidemic, Geneva.

[17] Ministério da Saúde. (2010) Vigilância de fatores de risco

e proteção para doenças crônicas por inquérito telefônico

(VIGITEL). Brasília.

[18] Instituto Brasileiro de Geografia e Estatística—IBGE.

(2011) Censo Demográfico 2010. Características da Po-

pulação e dos Domicílios—Resultados do Universo.

[19] Pala, V., Sieri, S., Masala, G., Palli, D., Panico, S., Vineis,

P., Sacerdote, C., Mattiello, A., Galasso, R., Salvini, S.,

Ceroti, M., Berrino, F., Fusconi, E., Tumino, R., Frasca,

G., Riboli, E., Trichopoulou, A., Baibas, N. and Krogh, V.

(2006) Associations between dietary pattern and lifestyle,

P. M. Ferreira et al. / Health 5 (2013) 23- 28

anthropometry and other health indicators in the elderly

participants of the EPIC-Italy cohort. Nutrition, Meta-

bolism and Cardiovascular Diseases, 16,186-201.

doi:10.1016/j.numecd.2005.05.009

[20] Haveman-nies, A., Tucker, K.L., Groot, L.C.P.G.M., Wil-

son, P.W.F. and Staveren, W.A. (2001) Evaluation of

dietary quality in relationship to nutritional and lifestyle

factors in elderly people of the US Framingham heart

study and the European SENECA study. European Jour-

nal of Clinical Nutrition, 55, 870-880.

doi:10.1038/sj.ejcn.1601232

[21] Lin, H., Bermudez, O.I. and Tucker, K.L. (2003) Dietary

patterns of hispanic elders are associated with accultu-

ration and obesity. Journal of Nutrition, 133, 3651-3657.

[22] Newby, P.K., Muller, D., Hallfrisch, J., Qiao, N., Andres,

R. and Tucker, K.L. (2003) Dietary patterns and changes

in body mass index and waist circumference in adults.

The American Journal of Clinical Nutrition, 77, 1417-

1425.

[23] Noel, S.E., Newby, P.K., Ordovas, J.M. and Tucker, K.L.

(2009) A traditional rice and beans pattern is associated

with metabolic syndrome in puerto rican older adults.

Journal of Nutrition, 139, 1360-1367.

doi:10.3945/jn.109.105874

[24] Scagliusi, F.B. and Lancha Junior, A.H. (2003) Subno-

tificação da ingestão energética na avaliação do consumo

alimentar. Revista de Nutrição, 16, 471-481.

doi:10.1590/S1415-52732003000400010

[25] Togo, P., Osler, M., Sorensen, T.I.A. and Heitmann, B.L.

(2001) Food intake patterns and body mass index in ob-

servational studies. International Journal of Obesity, 25,

1741-1751. doi:10.1038/sj.ijo.0801819

[26] Hu, F.B., Rimm, E., Smith-Warner, S.A., Feskanich, D.,

Stampfer, M.J., Ascherio, A., Sampson, L. and Willett,

W.C. (1999) Reproducibility and validity of dietary pat-

terns assessed with a food-frequency questionnaire. The

American Journal of Clinical Nutrition, 69, 243-249.