Dairy products consumption in Brazil is associated with socioeconomic and demographic factors : Results from the National Dietary Survey 2008-2009

Objective To estimate the consumption of dairy products in the Brazilian population. Methods Data from two non-consecutive food records from 34,003 individuals (aged ≥10 years) in the first Brazilian nationally representative dietary survey (2008-2009) were used to estimate the dairy products intake. Dairy products were divided into four major subgroups: cow’s milk, cheese, yogurt and other dairy products. Absolute and relative frequencies of individuals in the population who reported consumption of dairy products in at least one of the two-day food record according to sex, age, geographic region and levels of per capita monthly family income were calculated. Moreover, the mean consumption of dairy products, as well as their nutritional contribution in total energy, protein, calcium, phosphorus and magnesium intake was estimated.


I N T R O D U C T I O N
According to the Household Budget Survey (HBS) 2008-2009, Brazilians over the age of 10 ingest an inadequate intake of micronutrients.In the specific cases of vitamins A, C, D and E and calcium, inadequate intake surpasses the levels of 65%, 30%, 98%, 99% and 84%, respectively [1][2][3] .A recent change in feeding habits could help explain these findings.Brazilians have substituted highly nutritious staples for nutrient-poor processed foods that have high total fat, saturated fat, sodium and sugar contents 4 .For example, in the period between 2002 and 2009, the energy contribution from carbonated sodas to average Brazilian diet increased by 20 percentage points, while the contribution from milk and dairy products decreased by 10 percentage points 5,6 .
In face of these beneficial effects, health authorities in Brazil and other countries recommend the consumption of dairy products as part of a healthy diet 7,11,23,24 .In spite of official recommendations, previous work has shown that consumption of dairy products remains below the adequate, especially among lower socioeconomic classes.A study conducted with adult and elderly individuals in the Southern region of Brazil revealed that only 46.2% of the participants consumed dairy products daily.Consumption was higher among those individuals with more years of education 25 .Andrade et al. 26 and Fisberg et al. 27 observed, with the use of the Dietary Quality Index, low scores for milk and dairy products among teenagers and adults in the city of São Paulo.According to these authors, the individuals included in the surveys consumed one serving of dairy products per day or less.Again, lower scores were attributed to these products for individuals in worse socioeconomic conditions.Nevertheless, further studies should characterize in greater detail the demographic and socioeconomic patterns of dairy consumption among Brazilians.Thus, the present study aimed at estimating the consumption of dairy products and their nutritional contribution to the Brazilian population according demographic and socioeconomic factors.

Study population and design
We analyzed 34,003 Brazilian individuals at the age of ten years or over, of both sexes, selected to participate in the population-based National Dietary Survey, a module of the 2008-2009 HBS.The HBS used a two-stage cluster sampling design.In the first stage, primary sampling units of the census were used, with probability proportional to the number of residences in each tract.In the second stage, the sample units were permanent private residences, selected using simple random sampling without replacement, from within each primary sampling unit.Of the 68,373 residences sampled, 13,569 (24.3%) were randomly selected for the National Dietary Survey.All inhabitants aged ten and over took part, totaling 34,003 individuals (22.4% adolescents, 64.9% adults and 12.7% elderly).More details on sampling, collection and processing of data were previously described 1 .

Food intake data
Individual dietary data were obtained with a food-intake form filled out by survey participants in two non-consecutive days, randomly distributed within a one-week period.Thus, both weekdays and weekends were contemplated in the survey 6 .
Study participants were instructed to take careful notes on the quantities, in common household units, of all foods and beverages consumed throughout the course of 24 hours, including how food was prepared, ingredients and commercial brands.In addition, participants described the times of day when meals took place and whether they ate at home or out.The forms were revised by survey agents along with the participants to detect potential mistakes and to make corrections as necessary.Subsequently, the data on the form was transferred to a specific application software developed by the Instituto Brasileiro de Geografia e Estatística, which includes a pre-formatted list containing 1,500 foods and beverages, 16 options of cooking methods, and 106 types of portions and measurements as well as items which were compulsory such as the place and time of consumption.Any food or beverage consumed that was not listed in the application was added Rev. Nutr., Campinas, 30(1):79-90, jan./fev., 2017 Revista de Nutrição https://doi.org/10.1590/1678-98652017000100008by the survey agent.During the survey, a total of 1,120 foods and beverages were reported 6 .
Tables of nutrition composition 28 and portion sizes 29 , specially created to analyze food and dishes cited in the 2008-2009 HBS, were used to convert the reported food items into values for micronutrients.The nutritional composition tables were constructed based on data from the Tabela Brasileira de Composição de Alimentos (Taco, Brazilian Table of Food Composition) 30 , as well as the Nutrition Data System for Research (NDS-R, version 2008, Nutrition Coordinating Center, University of Minnesota, Minneapolis, United States of America).The Nutrition Data System uses, as a primary source of data, the National Nutrient Database of the United States Department of Agriculture (USDA).

Dairy products consumption
The Household Budget Survey database provides information on the consumption of all kinds milk (whole, skimmed and semi-skimmed), cheeses (white or yellow), yogurt (whole or diet/light, natural or with fruit), and other dairy beverages (fermented milks, fruit vitamins).In the present study, these items were distributed into four categories: milk, cheese, yogurt, and other dairy products.

Socioeconomic and demographic variables
The variables evaluated were sex (male/ female), age group (adolescent: 10-19 years, adult 20-59 years, elderly: 60 or more), geographic region (North, Northeast, Midwest, Southeast, South) and family monthly income per capita.The latter corresponded to the sum of gross monetary and non-monetary incomes of each family member divided by the number of family members.Income was calculated as the number of Minimum Wages (MW) per capita at the time of the study, when the minimum wage was approximately US$ 220,00 per month 6 .Seven segments of family income per capita were defined (≤1 MW; 1-2 MW; 2-3 MW; 3-6 MW; 6-10 MW; 10-15 MW; >15 MW).

Statistical analysis
Descriptive analysis of mean, percentage and 95% Confidence Intervals (95%CI) was performed in Stata ® software (Statistics/Data Analysis, version 12.0, College Station, Texas, United States of America) considering the complex sampling design (svy commands) and significance level of 5%.
Were calculated absolute (n) and relative frequencies (%) of individuals in the population who reported consumption of each dairy product in at least one of the two-day food record according to sex (male and female), age (adolescents: 10-19 years, adult 20-59 years, elderly: 60 or more), geographic region (North, Northeast, Southeast, South and Middle-West) and levels of per capita monthly family income (in minimum wages).Differences in prevalence were evaluated through Pearson's Chi-square test, adjusted for the complex sampling design.Furthermore, we calculated the mean and 95% confidence interval of dairy products consumption (g/day) and the nutritional contribution of dairy products to daily total energy, protein, calcium, phosphorus and magnesium intake according to the same variables (sex, age, geographic region and levels of per capita monthly Family income).Difference of means was tested through lincom command (for two-group comparison) and pwcompare command after svy regress estimation.The Bonferroni's multiple comparison test was applied as an option of the pwcompare command for comparisons among three or more groups.

R E S U L T S
The overall prevalence of dairy product intake was 43.0%, and the main dairy products https://doi.org/10.1590/1678-98652017000100008consumed were milk and cheese at the frequencies of 21.1% and 20.7%, respectively.Yogurt and other dairy products were both consumed with a frequency of approximately 7.0% (Tables 1 and 2).Prevalence of consumption of all dairy subgroups was significantly higher among females, except for "other dairy products".Regarding age, the group of elderly subjects had the highest percentage of total dairy intake (49.4%), milk (26.5%) and cheese (25.2%), while adolescents displayed the highest frequency of yogurt consumption (8.3%) (p<0.05).Stratified analysis by geographic regions revealed that the Northern region had the lowest percentage of dairy, milk, cheese and yogurt intakes.The Southern region had the higher intake of dairy (53.1%), cheese (30.2%) and yogurt (8.3%), while the Midwest was the region with highest milk intake (26.4%) (p<0.05).
Significant differences in the prevalence of dairy product intake were observed among income classes for all product subgroups.Dairy products consumption ranged from 31.1% in the lowest income class (≤1 MW per capita) to 85.4% in the highest (>15 MW), while cheese and yogurt intake ranged from 9.5% in the lowest to 59.9% in the highest income class and from 4.2% in the lowest to 26.0% in the highest income class, respectively.Dairy products contribution very little to total energy intake (3.8%) and for the protein, phosphorus and magnesium intake (4.5%, 8.6% and 3.9%, respectively).The food group contributed more heavily to calcium intake (16.3%) (Table 3).
Dairy products represented a higher percentage of total energy and nutrient intake among females, the elderly, inhabitants of the Southern region and those individuals in the higher income categories.The only exception to this pattern regarded the contribution to magnesium ingestion, which was higher among adults.We also highlight the steep increase in the contribution of dairy products to energy and calcium intake with increasing family income.In the highest income class these contributions were, respectively, 3.3-and 2.7-fold greater than in the lowest income class (Table 3).
The average intake of dairy products in the study population was 101.8g,where milk represented an average of 54.5g.The highest intake averages for dairy products in general and milk were those of women, the elderly, Midwesterners, and those of individuals in the highest income class, whose average intake of dairy products was 3.3-fold higher than that of the lowest-income individuals (263.54gversus 79.2g).The average intake of cheese was higher among the elderly, Southeasterners, and in the highest income classes.Regarding yogurt, intake increased in the higher income class, among women, teenagers and Southerners (Tables 4  and 5).

D I S C U S S I O N
The prevalence of individuals who consumed dairy products in Brazil was only 43.0%, with significant variations depending on the demographic and socioeconomic variables evaluated.The highest percentages were found among women, the elderly, Southerners, and among those with highest income.Moreover, the average daily intake of these products was approximately 110g and their contribution to energy, protein, phosphorus and magnesium was below 10.0%.Dairy products contributed more intensely to the intake of calcium, with 16.3%.In the present study, we estimated the prevalence of dairy product consumers (%) as well as the amounts consumed, always looking at the overall product category and 4 subgroups.Previous studies also evaluated the intake of dairy products in different populations, using distinct methods 10,25,[31][32][33][34] .Regardless of methodological differences, all this studies, including the present study point to the low consumption of dairy products.In one work where global, national and regional estimates of milk and dairy beverage consumption were presented, the global daily average was similar to that found here (137mL/day).Higher values were found for citizens of wealthier countries (high income: 173mL/day; low-income: 72mL/day), and the elderly, especially women (163mL/day in women age 60 and over and 122mL/day in men aged 20-39).They explained the higher intake among elderly women as the possible result of medical recommendations, which prescribe milk intake as a way of avoiding bone loss and osteoporosis.For Brazil, these authors reported an average intake of 120mL/day.The Central Latin America was the region with highest milk intake (266mL/day).In Latin America (Southern), the milk consumption among adults averaged 112.8mL/day (95%CI=79.2,160.8) 31 .
In another study conducted in Switzerland, the authors observed that adults and the elderly did not consume the recommended amount of three to four servings of dairy products per day.The average number of servings found was 2.6, without significant differences between socioeconomic and demographic groups  considered.Moreover, 25.0% of the individuals reported reducing dairy consumption to decrease the ingestion of fat or cholesterol 34 .In a study based on data from the National Health and Nutrition Examination Survey (NHANES  2001-2004), the authors found that more than half of the Americans consume than the minimum recommended amount of three daily servings of dairy products 32 .In addition, Huth et al. 33 , who also worked with NHANES (2003-2006) data, observed that the energetic contribution of milk and cheese to the average North American diet was 9.2%.Regarding the contribution of these products to protein, calcium and magnesium intake the values corresponded to 17.1%, 46.3% and 7.9%, respectively, which are higher than found in the present study.
The Institute of Medicine recommends levels of daily calcium intake 35 that the average Brazilian would have difficulty consuming.A representative sample from the municipality of São Paulo showed that Brazilians consume dairy products only during breakfast, or at dinner by individuals who replace the meal with coffee and milk 36 .Another work conducted in the city of Campinas, near São Paulo, found that adolescents who reported sleeping less than seven hours per day ingested significantly more calcium than those who had more hours of sleep, probably because the former had breakfast with higher frequency 37 .Adolescents often skip their first meal of the day.Indeed, Leal et al. 38 observed that 21.0% of the adolescents evaluated did not have breakfast on a frequent basis 38 .Some factors help explain this pattern, including an exaggerated concern with body image, lack of time, lack of appetite during this period of the day, and lack of incentives from the family 39 .Breakfast and snacks offer an opportunity when the intake of foods that are not always provided in the large meals, such as milk and dairy products, should be prioritized.
In Brazil, the cost of dairy products 40 , unfamiliarity with the health benefits associated with their consumption 41 and current eating habits 4,42,43 are some factors that may be related to low intake of dairy products.The importance of family income to consumption of certain foods and food groups, including dairy products, is well documented 40,44 .Levy-Costa et al. 44 observed an increase in milk participation in Brazilian's diet with increasing household income, reaching a threefold difference in the comparison between the highest and the lowest income groups 44 .In the present study, we observed similar behavior, especially for the consumption of cheese and yogurt.Moreover, when comparing the percentage of consumption of these foods among the most economically developed regions (South, Southeast and Midwest) to less developed (North and Northeast), differences were also striking.
Dietary intake of Brazilians is often far from the nutritional recommendations, characterized by a low intake of fiber, vitamins and other minerals and high intake of saturated fat, added sugar and sodium [1][2][3]6 . Infact, several studies showed a low intake of fruits and vegetables and high intake of foods and beverages with high energetic value and low nutrient density as snacks, pizzas and sugar-sweetened beverages (sodas, soft drinks and juices) in Brazil 4,45,46 .The presence of these drinks, in addition, has been associated with low dairy products intake by the population 45,47 , which is worrying considering the substantial increase in consumption of soft drinks primarily in last years 44 .In addition, meals important to consumption of dairy products, like breakfast and intermediary snacks, have been omitted 38 or tend to have a predominance of foods and beverages with low nutritional value such sugar-sweetened beverages, sweets and desserts 46 .
The present study has a few limitations, including intrapersonal variability in the consumption of dairy products and potential errors built into consumption estimates, especially due to imprecisions in the calculations of energy intake, as previously discussed 48 .These imprecisions highlight the need for questionnaires that yield more accurate answers.Despite these limitations, the study is a representative sample of the Brazilian population and allows for a deeper understanding of dairy consumption among different population subgroups in different regions.
The results generate concern, given the nutritional importance of dairy products 7,9,23 and the health benefits associated to their consumption, including improved bone health, reduce risk of osteoporosis development 21 and the prevention of chronic disease such as obesity, diabetes, hypertension and cardiovascular disease.Moreover, the results unveil some of the reasons underlying the inadequate intake of nutrients such as calcium, magnesium and phosphorus observed among Brazilians 1-3 .

C O N C L U S I O N
The present study provides further indication of the low intake of dairy products by the Brazilian population.The diffusion of knowledge about the benefits associated with milk and dairy intake, the guidance on the different forms of consumption of these foods items and wider dissemination of the recommendations proposed by the Ministry of Health for consumption of less processed foods such as cow's milk, cheese curds and plain yogurt in versions without fat or with less fat 23 represent important strategies to increase the intake of dairy products Brazil.Further studies evaluating factors that are associated with the consumption of these foods, will enable the establishment of strategies to encourage and promote consumption to specific segments of the population.

R E F E R E N C E S
., Campinas, 30(1):79-90, jan./fev., 2017 Revista de Nutrição https://doi.org/10.1590/1678-98652017000100008 C O N T R I B U T O R SG POSSA participated in the analysis and interpretation of the data and elaboration of the article.R FISBERG contributed in the interpretation of the data, revision and approval of the final version of the article.MA CASTRO participated in the analysis and interpretation of the data and in the review and final approval of the article.R SICHIERI and M FISBERG participated in the review and final approval of the article.

Table 2 .
Prevalence of cheese, yogurt and other dairy products intake according to demographic and socioeconomic variables, Household Budget Survey.Brazil, 2008-2009.

Table 3 .
Nutritional contribution of dairy products to daily total energy and nutrients intake according to demographic and socioeconomic variables.Household Budget Survey.Brazil, 2008-2009.

Table 4 .
Mean and 95% Confidence Interval (95%CI) of dairy products and milk intake according to demographic and socioeconomic variables, Household Budget Survey.Brazil, 2008-2009.Significant differences between adults and older adults; b Significant differences between North and the other regions; c Significant differences between ≤1 MW and the other categories of family income; d Significant difference between ≤1 MW and >2

Table 5 .
Mean and 95% Confidence Interval of cheese, yogurt and other dairy products intake according to demographic and socioeconomic variables, Household Budget Survey.Brazil, 2008-2009.