INTRODUCTION

Chronic Non-Communicable Diseases (CNCD) are the leading cause of death worldwide and are responsible for 63% of fatalities, although the mortality rate has declined in developed countries in recent years^{1 , 2}. Like other developing countries, Brazil is currently experiencing a triple burden of disease, where chronic conditions account for about 72% of deaths^{3 , 4} and coexist with infectious disease, maternal disorders, perinatal conditions and vitamin deficiencies, as well as the growth of violence and external causes^{5 , 6}.

Hypertension (H), prevalent in 36.0% of people aged over 40 years, and Diabetes Mellitus (DM), present in 5.7% of adults and almost 18.0% of the elderly, are chronic conditions and primary risk factors for cardiovascular disease. Together, they are also the leading cause of hospitalization in Brazil^{3 , 4 , 7}.

Among the determinants governing the prevalence of these diseases are rapid demographic and nutritional transition and changing lifestyles^{2 , 4 , 6 , 8}. For example, there has recently been an increase in the consumption of high energy foods rich in fat, salt and sugar with a negative impact on the control of CNCD^{2 , 9 - 11}. Food consumption stands out among modifiable risk factors and research into eating habits allows the creation of strategies to promote health through diet^{2 , 11}.

In 2006 the Ministry of Health published the Food Guide for the Brazilian Population as an official instrument that establishes guidelines for healthier food choices. The Guide was distributed among primary care staff with the goal of advising professionals on health promotion and the prevention of diet-related diseases^{9 , 11}, with a new edition set for publication in 2014. The aim of this study was to describe and classify, according to recommendations in the Food Guide for the Brazilian Population, the adequacy of the eating habits of hypertensive and diabetic patients treated at a primary care facility and determine the association between diet suitability and the health problem.

METHODS

Study design and sampling: this is a crosssectional study that used baseline data collected between March and April 2011 from longitudinal study designed to assess the effects of an intervention with staff from health care units. The population studied was composed of hypertensive and diabetic adults from the catchment area of the Serviço de Saúde Comunitária (Community Health Care Service) of the Grupo Hospitalar Conceição who had doctor's appointments through the service in the previous three years. The service is a reference for 108,000 people and consists of 12 health care units located in the northern and eastern districts of Porto Alegre, Rio Grande do Sul State. Staff consists of doctors, nurses, dentists, psychologists, social workers, nursing, administrative and oral health technicians, and community health care agents, with additional support provided by dieticians, pharmacists and medical specialists.

The size of the baseline sample was defined based on the objectives of the primary study that this investigation is part of, totaling 3,784 subjects chosen randomly from the records of diabetic and hypertensive adults registered on Information System. Among those selected there were 65 refusals and 281 people did not take part because they were not hypertensive or diabetic, no longer lived in the catchment area or had passed away. There were 51 people that could not be interviewed as they were on vacation or hospitalized and 889 people could not be located either because there was no recorded address, no one was home after four attempted visits, or none of the neighbors reported knowing the individual. It is quite plausible to think that the proportion of subjects who were not part of the target population was higher among those not located than those who provided information. As such, were estimated a response rate of around 75% with 2,482 people interviewed.

Variables: The baseline used a structured questionnaire applied at subjects' homes by qualified interviewers. Dependent variables for the study were: gender, age, schooling level, marital status and socioeconomic classification determined according to the criteria of the Associação Brasileira de Empresas de Pesquisa (ABEP, Brazilian Association of Market Research Companies)¹², grouping social classes into three categories (A/B, C and D/E). There was a flaw in the questionnaire and the schooling level of the head of the household, an item in the proposed algorithm, was not collected. In order to remedy this problem the information was collected by telephone, with a response rate of 50%. Were estimated the correlation between the ABEP classification calculated using the schooling level of the household head and that determined using the schooling level of interviewee, establishing a Kappa index of 0.862, (95% Confidence Interval [95%CI]=0.834-0.893) for the ABEP rating grouped into 3 categories. The variable health problem was categorized as "only Hypertensive" (H), "only Diabetes Mellitus" (DM) or "Hypertension and Diabetes Mellitus" (H*DM).

Consumption of markers of a healthy diet was identified using questions from the test entitled "How is your diet?", from the pocketsized Food Guide¹³. These enquired about the consumption of vegetables, fruit, legumes, fish, fried foods, sausages, candy, soda and dairy products, and about the habit of removing any visible fat from poultry or meat.

Diet adequacy was classified into three levels according to the scoring system proposed in the Food Guide test, adapted for this study. Questions on physical activity and alcohol consumption were replaced by the International Physical Activity Questionnaire (IPAQ)¹⁴ and Cut-down; Annoyed; Guilty & Eye-opener (CAGE)¹⁵, whose results were included in the diet classification score. The question about reading food labels was excluded. As such, the cutoff points for classifying diet adequacy were adjusted slightly.

The Food Guide classifies diet into three levels: up to 28 points, from 29 to 42 points and 43 points or more. In this study these were altered to first quartile, interquartile range and third quartile (up to 28, from 28 to 34, over 34), respectively, to rate the diet as "inadequate", "partially adequate" and "adequate". This change in the cutoff points was based on point distribution, that is, distribution close to the normal curve. This was also used as a dichotomous variable where the categories "partially adequate" and "adequate" were combined for analysis purposes.

The Chi-square test was used with the socioeconomic variables and health problem variable. The description of the consumption of food markers and the prevalence of adequate diet were obtained using the Chi-square test as a function of the health problem. Were considered significant associations with a type I error <5% (p<0.05).

Poisson regression was used to determine the prevalence ratio, with the chronic form of the disease considered a factor when calculating the prevalence of adequate diet in each stratum. Analyses were conducted to test the independence of the groups without considering causality. PASW Statistics 18 R software was used¹⁶.

The research project was submitted to the Research Ethics Committee of Hospital Nossa Senhora da Conceição and approved on June 13, 2012, under Protocol number 39056.

RESULTS

A total of 2,483 subjects were analyzed, of which 68,0% were women. Among the variables studied, the proportion of missing data varied from 0.7% to 7.0% and was above 5.0% for two variables (schooling level and ABEP classification). Average age was 63.2 years (95%CI=62.7-63.7), 31.0% of participants completed high school and 52.0% lived with a partner. The predominant socioeconomic classification was class C (64.0%), with approximately 25.0% classed as A/B and only 11.0% D/E.

Table 1 shows that 66.5% (95%CI=64.568.3) exhibited only H, 6.5% (95%CI=5.6-7.5) only DM and 27.1% (95%CI=25.3-28.9) presented H*DM. Higher prevalence of H was recorded among women and DM in men. The prevalence of patients with H*DM doubled in the 40-59-year-old age range in relation to 20 to 39-year-old.

The results shown in Table 2 indicate the distribution of eating habits and their relationship with the health problem. Approximately 50.0% ate less than 3 soup spoons of vegetable or legumes a day and 80.0%, less than 3 fruits per day. There was no intergroup difference in pulse consumption, with 72.8% of subjects eating pulses on a daily basis. Around 50.0% of participants ate fish at least once a week, with higher proportions (61.0%) observed among those classed as only diabetic. Consumption of full fat milk and dairy products was reported by 64.0% of the sample, with low fat dairy products consumed more by people categorized as H*DM. Failure to remove visible fat from poultry and meat was observed in 22.5% and was more common among hypertensive subjects. Fried food and sausages were consumed twice a week or more by around 26.0% of participants. Approximately 40.0% of diabetics reported daily consumption of candy, soda or artificial fruit juices.

Table 3 shows the classification of diet adequacy and its association with the health problem. Inadequate eating habits were prevalent in 29.6% (95%CI=27.7-31.6) of the sample, with higher levels recorded (32.5%) among those suffering from H. Adequate diet was observed in 23.4% (95%CI=21.7-25.3) of the sample and was more prevalent in those with H*DM (26.8%). The prevalence ratio for inadequate diet was 30% higher in hypertensive subjects compared to diabetics or those suffering from both diseases, as shown in Table 4.

DISCUSSION

The greater prevalence of hypertension among women recorded in this study was also observed in the telephonic inquiry¹⁷. On the other hand, for this sample diabetes was more frequent among men, unlike the results of the telephonic inquiry, which found that DM was more common in women¹⁷.

With respect to schooling level, data from this study indicate that people who have completed high school or tertiary education exhibit lower comorbidity and high DM levels, similar to Vigilância de Fatores de Risco e Proteção para Doenças Crônicas por Inquérito Telefônico (Vigitel, Vigilance Risk Factors and Protection for Chronic Diseases by Telephonic Inquiry) data, in which H and DM were twice as prevalent in people with less than 8 years of schooling¹⁷. Socioeconomic classification showed a predominance of class C individuals and only 11% categorized as D/E, which may have been underestimated due to losses.

The increased prevalence of comorbidity associated with higher age was also observed in 2010 Instituto Brasileiro de Geografia e Estatística (IBGE, Brazilian Institute of Geography and Statistics) data, which showed a 45% rise in the presence of chronic disease between the ages of 40 and 59 years and a prevalence of 65% from 50 years, reaching almost 80% in those aged 65 years or more. When asked about the number of chronic diseases, comorbidity doubled from 40 years old and tripled among those aged 65 years or more⁷. The Vigitel 2010 recorded an almost three-fold increase in the prevalence of H and DM among those older than 45 years, climbing to almost six times higher for H and ten times for DM from 65 years old¹⁷.

Most respondents did not follow the dietary guidelines recommended by the Ministry of Health for the general population¹¹. In regard to food markers of a healthy diet, insufficient fiber consumption was observed as a result of low fruit and vegetable intake and excessive consumption of soda, sugar, salt and saturated fat. A baseline study in 16 countries that included approximately 2,790 overweight or obese people with type 2 diabetes (Action for Health and Diabetes Group [Look AHEAD]) also concluded that most people do not adhere to dietary recommendations¹⁸.

Fruit and vegetable intake in this study was far below that recommended by the World Health Organization of at least five daily servings for the general populationl² or the level stipulated in the Food Guide of at least 3 servings of fruit and 3 of legumes or vegetables every day¹¹. Nevertheless, although lower than recommended levels, consumption of these foods was still higher than in the Brazilian population in general according to data from a population-based survey that found that 9.2 to 24.0% of people ate five servings of fruit and vegetables every day¹⁷. The study by Levy-Costa et al. ¹⁹ evaluated food availability in urban areas from 1970 to 2003 and found a lack of fruit and vegetables in peoples' diets.

The results of the "Look AHEAD" baseline study assessing the eating habits of diabetics showed adequate vegetable consumption of 38%¹⁸. The diabetic subjects in the present study reported less satisfactory vegetable and legume intake. Considering that diabetics from both studies are monitored by health care services and that vegetable and legume consumption are emphasized for their role in glycemic control²⁰, the findings of the present study are less satisfactory.

The habit of eating one serving of pulses at least five times a week, in line with Ministry of Health¹¹ recommendations, observed in diabetic and hypertensive subjects in comparison to the general population was similar to the Brazilian average (66.7%) and higher than the percentage recorded in Porto Alegre (49.7%)¹⁷.

The Food Guide recommends eating fish at least twice a week¹¹, however, our study found that almost half the subjects never or hardly ever did so. According to the study by Levy-Costa et al. ¹⁹, based on the Family Budget Survey, fish consumption declined by 50% from 1970 to 2003.

The consumption of full fat milk and dairy products in the present study exceeded the estimates of the 2010 telephonic inquiry, which indicated a national average of 56.0% and 44.3% in Porto Alegre (RS)¹⁷. Failure to remove visible fat from meat and poultry was less common in this study than in the telephonic inquiry, which estimated that in Porto Alegre (RS) 35.0% of the adult population do not remove visible fat, a percentage similar to the national average¹⁷. The Food Guide suggests low fat milk and dairy products and recommends that visible fat be removed from meat and poultry to lower saturated fat intake¹¹.

Eating fried food, sausages and sweets and drinking soda or artificial fruit juice more than twice a week was very common among interviewees. The Ministry of Health highlights the need to lower salt intake and emphasizes high levels as one of the causes of hypertension, limiting daily salt consumption to five grams for the general population¹¹. A single serving of the above mentioned foods typically contains the recommended salt intake for an entire day. The Brazilian Food Guide recommends a maximum of one daily serving of these foods for the general population, while the consumption of candy by diabetics should be monitored according to the glycemic load^{21 , 22}.

A distinctive characteristics found among diabetic subjects was that more than half reported they did not eat candy regularly, a habit considered appropriate, while more than one third ate candy on a daily basis, although this amount decreased among those with H*DM. The AHEAD study found candy consumption above recommended levels in 28% of the diabetics assessed, a lower percentage than that recorded in the present study.

Four major studies on food availability between 1970 and 2003 reported high consumption of sugar, fats in general and processed foods^{11 , 23 - 25}. In the presence of a chronic condition, reduced consumption of these nutrients is highly recommended given their relationship with H²³ and DM^{22 , 26} and risk of cardiovascular diseases^{2 , 11}. Nevertheless, despite being monitored by a health care service, the eating habits of hypertensive and diabetic individuals in this study were not satisfactory.

There was a significant tendency towards better eating habits as the health problem worsened, although causality cannot be confirmed. Considering that the disease burden for diabetes (680 DALY per 100,000) is greater than that of hypertension (222 DALY per 100,000)²⁷ and that the presence of H*DM is worse than suffering from only one of the conditions, it can be suggested that the severity of a chronic condition is a motivating factor in dietary changes and greater contact with health care services.

A higher prevalence ratio of inadequate diet was found in the group consisting of only hypertensive participants (Table 4), which can also be observed in more detail in the differences in fruit and fish intake and removal of visible fat from meat and poultry, shown in Table 2. The detection of poorer eating habits among hypertensive subjects indicates the need for strategies aimed at preventing comorbidities and complications in this population. Diabetics reported eating more fish and less candy, but consumed full fat milk and dairy products (Table 2). Those with H*DM and a more adequate diet (Table 3) tended to remove visible fat from meat and consume dairy products with a lower fat content (Table 2).

Limitations of the present study were related to the questions on eating habits, in that the test from which they were taken was not validated. As such, like most food surveys there is potential for measurement bias, with one of the problems encountered being serving sizes. In order to reduce this bias, tables of homemade servings were described in the manual and examples were given in the questionnaire.

Another point to be considered is that the study deals with people suffering from diseases who frequently receive diet-related information from either the media or health care professionals. An additional limitation was that part of the sample could not be located despite efforts to compensate for the loss. Nevertheless, for the purposes of this study the sample size was sufficient for analyses. It is quite plausible to consider that the proportion of subjects who were not part of the target population was higher among those not located than those who provided information.

CONCLUSION

The study demonstrated that eating habits among diabetic and hypertensive subjects, monitored by primary health care services, were inadequate according to Ministry of Health guidelines. The association between better diet and the presence of both diseases indicates a delayed change in eating habits, since poor diet is a known risk factor for chronic diseases and their complications. It is important to work with the population in applying preventive measures that promote the adoption of healthy eating habits before health deteriorates. Further studies are suggested to identify the most effective and feasible techniques for promoting a healthy diet in primary health care.