Perceptions about men ’ s health in a gender relational perspective

The goal of this article is to assess the perception of male health from the viewpoint of men and women. The study drew from a random sample of men aged 20 to 59 who were SUS (Unified Healthcare System) users and resided in the capitals of Brazilian states and the Federal District. Participants were interviewed by phone. Sociodemographic variables and variables related to perception of healthcare services, health status and health care were recorded. Logistic regression was used to assess failure to seek service and good and very good self-care. The majority of the study population considers it has no health problem, especially men. The main reason for failure to seek treatment is no access to services, although both men and women claim the healthcare services receive them adequately. Over 40% of men and almost 30% of women self-medicate. Men believe they are taking good care of their health, while women have a different perception of this. Perceptions differ when we take into consideration age, years of schooling, ethnicity and occupation.


Introduction
The first studies of men's health emerged in the US last century, in the late 1970s.Most of these studies centered on health risks 1 .In the nineties, the topic gained new approaches, among them an attempt to focus on male health and disease from a gender relational point of view 2,3 .
In the field of knowledge produced on this theme, we call attention to the review by McKinlay 4 , in New Zealand.In this study gender differences regarding morbidity, mortality and life expectancy were explained based on the biological-genetic specificities of men and women, on social differences and inequalities, on different social expectations on the part of men and women, on the search and use of healthcare services and professional healthcare focused on men.
Looking at the literature on this topic produced in Brazil, the milestone would be the special issue of Ciência & Saúde Coletiva (Science and Group Health) published in 2005.The editorial of this number stated that promoting the health of both men and women required a deeper exploration of the hegemonic ideologies of masculinity and the health-disease process 5 .
The important milestone in Brazil was the National Policy for Integrated Healthcare for Men, created by the Ministry of Health to promote health activities focusing on the unique situation of men in the different sociocultural and political contexts 6 .
In a discussion about men's health, there are studies that focus on the need to address the topic from a gender relational perspective 5,7 .This means that gender models are built using an inter-relation, just as what is culturally viewed as "male" only makes sense from the female perspective, and vice-versa 8 .The adjective "relational" is different from "complementary".This category creates and reproduces power asymmetries 7 .Perceptions of men's health may be viewed as a synthesis of the interactions between the male and female perceptions in a context of asymmetries.
Perceptions regarding health -understood here as subject interpretations of their own health -have been used in long-term studies.These have been taken as an important indicator for, among other purposes, understanding how individuals perceive their well-being 9 and understand the health situation of the populations 10 .Individual perceptions have also been associated with health self-assessment studies 11 .In epidemiological terms, these studies have predictive pow-er on the mortality of different socioeconomic groups, despite being limited, especially in terms of measurements 12 .
Based on these initial considerations, this study attempts to analyze the male and female perceptions of men's health and their access to healthcare services.Analysis of the differences and convergences in the perceptions of men and women may provide subsidies for expanded understanding of male health, from a gender relational perspective.

Method
This cross-sectional study was performed in 2014 on a random sample of men aged 20 to 59 who were SUS (Unified Healthcare System) users and resided in the capitals of Brazilian states and the Federal District.Participants were interviewed by phone.
The sample size was calculated using as a parameter the fact that 45.7% of fathers are present during delivery 13 (5% error and 95% confidence interval).This determined a total of 3,810 interviews in Brazilian state capitals and the Federal District, which concentrate the majority of the adult population.
Because the survey was done by phone, the informed consent form was replaced by verbal consent obtained from the interviewees during the telephone conversation.The study entitled "Male Care Focused on Sexual Health, Reproduction and Paternity from the Gender Relational Perspective" (in Portuguese Os cuidados masculinos voltados para a saúde sexual, a reprodução e a paternidade a partir da perspectiva relacional de gênero) 14 was approved by the Ministry of Health, Fernandes Figueira National Institute of Women's Children's and Adolescent Health Ethics Committee for Research with Human Beings.This institute is part of the Osvaldo Cruz foundation in Rio de Janeiro.
This study used sociodemographic variables, and perception of healthcare services, health status and health care.
Sociodemographic variables were region of residence, age, schooling, ethnicity, religion, marital status and employment status.Regions were North, Northeast, Southeast, South and Middle West.Age was split into four groups: 20 to 29, 30 to 39, 40 to 49 and 50 to 59; years of schooling were split as follows: < 9, 9 to 11 and ≥ 12 years of schooling; ethnicity were while, brown, African descent, Oriental and native Indian; religion was yes or no.Marital status could be non-stable union or stable union; employment status could be yes or no.
Health variables were perception among the priority population of the primary healthcare units, presence of any disease or health problem, incidence and location of treatment, reason for not seeking treatment.Responses were subsequently opened and encoded.We also analyzed perceptions of how Healthcare Services service men and self-care among men.In both cases the options were very well/very good, well/good, fair, bad/poor and very bad/very poor.
We also explored the association between sociodemographic (independent) variables and failure to seek treatment (dependent variable) in the presence of a health issue among men and women, and between sociodemographic (independent) variables and good and very good selfcare (dependent variable) from the viewpoint of men.We used Poisson's regression to calculate the gross and adjusted prevalence ratios.
Estimates were calculated bearing in mind the complex nature of the sample, using two post-stratification weighting factors.The first factor (ratio) was the ratio between the relative frequency of individuals found in the 2010 Demographic Census, and the relative frequency found in this study using the 24 categories of gender (male and female), age (20-29, 30-39, 40-49 and 50-59 years of age) and years of schooling (<9, 9-11 and ≥ 12).The purpose was to achieve an even sociodemographic composition.The second factor (sample fraction) was the ratio between the total number of adults living in each region and the total in this study, to correct any differences between the population of each region and the study population.The final weighting factor for the total population was the product of these two factors, and for each region only the ratio.Thus, the data was expanded to represent approximately 12 million men and 14 million women living in the study locations.
Data was processed using Stata, bearing in mind the specificities of men and women, and a significance level of 4% (p < 0.05), and the 95% Confidence interval (CI 95% ) for comparing genders, as well as the representativeness of each interviewee in the adult population assessed.

Results
We completed 3,885 interviews, 1,894 with men and 1,991 with women, and found no difference in the majority of the sociodemographic charac-teristics we looked at (Table 1).Differences were found only in religion and employment status, women are more religious and more men are employed.
In general, both men and women feel that primary healthcare services focus on everyone, regardless of gender, age, social status or health, with a larger percentage of women than men sharing this view (Table 2).Among the men, 14.1% claimed that, at the time of the interview, they had some disease or health problem, slightly more than half the corresponding percentage among women.Of these, most were being treated, most often with self-medication, in particular among men (42.1%).Among the women, we found an equal prevalence of healthcare services (private or primary care unit) and self-medication.The men who claimed not to be in treatment reported not having access to care or the problem not being important as the reasons for not seeking treatment.In the case of women, no access to care was the reason given most often.
Figure 1 shows the health problems mentioned by men and women.We find that the three main groups in both genders mentioned circulatory system disorders, osteomuscular system and connective tissue disorders, blood disorders, disorders of the hematopoietic organs and a number of immune disorders.Neoplasias and endocrine, nutritional and metabolic diseases were more frequent among women.
Regarding male perception of the quality of health provided by primary healthcare units, most of the men or women claimed it is good or fair, with fair being mentioned most often by the men in the study (Table 2).Almost twice the number of women as men claimed they did not know.When we asked how men care for their own health, the answers given by men and women matched.However, men were more likely to answer good or very good, and women more likely than men to answer bad and very bad.
The regression analysis adjusted for sociodemographic variables (Table 3) shows that for both genders, advancing age is a factor of protection in terms of seeking care in the presence of some health problem, as is more years of schooling.Among women, religion is a risk factor as not seeking care is 20% more prevalent among women claiming to have some kind of religion.We call attention to the fact that only 7.3% of women aged 50 to 59 did not seek medical care, compared to 40.8% of the younger women.The same is true for women with more schooling (9.2%) compared to those with less schooling (28.0%).We observed the same trend among men for these variables, although with overlapping confidence intervals.
From the viewpoint of men (Table 4), healthcare self-assessment is directly linked to years of schooling, with the group having more education being 32% more likely to consider care as good or very good, compared to the groups with fewer years of schooling.Native Indian men have more than a 50% chance of considering their healthcare as fair, poor or very poor.From the viewpoint of the women in this study, women of Oriental or native Indian ethnicity believe the chance that men take care of themselves well or very well is higher (by as much as 90%) than it is among other women.Employment is associated with a more positive vision of male self-care among women.They are 69% more likely to consider male self-care as good or very good than women who are not part of the job market.We point out that among almost all categories of women, the prevalence of male self-care of health as good or very good did not exceed 10%, remaining always below how men assess themselves.

Discussion
Internal validation of the results may be analyzed during the course of the project, from training the teams through statistical data analysis.The sample plan was designed to include all locations, and phone numbers were drawn at random.Success rate was 76.2% and the interviewers reported no problem identifying eligible subjects.
External validation of this study is primarily through the comparison of the distribution of the sociodemographic categories of the study population, which does not differ from official estimates based on census data by more than 10% 15 .Specifically regarding the variables that in this study differ among men and women, 24% of the men and 14% of the women claimed to have no religion.This ratio (1.7) is close to the ratio obtained in the 2010 census (1.4 in rural areas and 1.5 in urban areas).This small difference may be explained as the study population lives in state capitals and the Federal District, where there may be a higher frequency of people with no religion.Regarding employment, the census also shows more men than women are employed, in a ratio of 1.4, similar to that found in this study (1.3).
In other national studies, data comparison is hindered by differences in methods, from the study population through analyses.Nevertheless, some analogy is possible, as shown below.
A study conducted in the south of Brazil 16 shows that women over the age of 14 are three times as likely to mention some health problem in the past two months than are men of the same age group.This corroborates the estimates in this study, which indicate that almost twice the number of women mentioned some health problem.However, these statements must be put into context, as the survey population is of a different age, and the period of study is also different.
Regarding the healthcare problems reported, the data coincides with that obtained during the study conducted in Minas Gerais 17 , in which diseases of the circulatory system were mentioned most often, followed by osteomuscular and connective tissue disorders.However, the third cause differs.As the third cause, our survey found dis-orders of the blood and hematopoietic organs, as well as some immune disorders, while the Minas Gerais study listed endocrine, nutritional and metabolic disorders among men, and respiratory system disorders among women.These differences may be explained by the difference in age, region and especially study population, as the Minas Gerais study was limited to healthcare workers.
A study conducted in the South 16 of the country also shows a higher risk (1.78) of poor perception of risk among women.Although not within the scope of this study, it is possible to draw a parallel with the answers about male care of their own health, where most men considered their self-care to be good or very good, and women believe it to be poor or very poor.This perception may explain why men mention fewer health problems and thus seek healthcare services less frequently, which is in line with their view of fair service at healthcare units.However, several studies have mentioned that men are more likely to engage in risky behavior than women, and have worse biochemical indicators 18,19 .
A systematic review estimates the prevalence of self-medication among the adult Brazilian population as 35% 20 , with no gender specified, *weighted to represent the male and female population aged 20 to 59 living in state capitals and the Federal District.0 = does not know; 1=some infectious and parasite disorders; 2 = neoplasias; 3 = blood and hematopoietic organ disorders and some immune disorders; 4 = endocrine, nutritional and metabolic disorders; 5 = mental and behavioral disorders; 6 = nervous system disorders; 7 = eye and related disorders; 8 = ear and mastoid apophysis disorders; 9 = circulatory system disorders; 10 = respiratory system disorders; 11 = digestive system disorders; 12 = skin and subcutaneous tissue disorders; 13 = osteomuscular and connective tissue disorders; 14 = genitorurinary disorders; 15 = pregnancy, delivery and puerperium; 17 = congenital malformations, deformities and chromosomal anomalies; 19 = lesions, poisonings and other consequences of external causes.further corroborating the data from this study, which estimates that the percent of men who self-medicate is over 40%, compared to 30% of the women.Here we point out that although the percentage is high, these people consider themselves to be under treatment for an existing health problem.Among those mentioning this practice as non-treatment and a reason for not seeking healthcare, the percent self-medication is small (about 3%).
In the domestic literature [21][22][23][24][25] , it is clear that men seek out healthcare services less than women.Reasons for not seeking healthcare, even though they have a health problem, men mentioned not having access to healthcare and non-importance of the problem as the main reasons, and women claimed lack of access to care.Among the general population, in 2008 26 the main reason given was not having money to pay for healthcare, followed by difficult access and delays.This data puts in question the role of SUS, in that a large percentage (20.5%) of the population sampled in 2008 considered private care as the location they go in search of care.This study differed, to the extent that one of the selection criteria was the use of primary services provided by SUS, with universality being mentioned as one of the characteristics of service.
As mentioned by Aquino et al. 27 three decades ago, men continue displaying the same health characteristics, showing that in spite of the SUS, little progress has been made in terms of men's healthcare and expanded service for this popula-

Collaborations
EC Moura helped conceive the project, analyze and interpret the data, write the article, and was responsible for all aspects of the work and for ensuring the accuracy and integrity of the entire work.R Gomes helped interpret the data, draft the article and with the final approval.GMC Pereira helped with a critical review of the intellectual content and final approval.
tion.The situation is even worse when we realize that self-medication is a common practice, and can be considered an indicator of not having access to primary healthcare services.
In general, study data serves to underlie discussions about cultural gender models in healthcare.These models may influence the perceptions men and women have, and their search for Healthcare Services.To this end, perhaps because of the common association between men and strength, women and weakness, influenced our male subjects, leading them to see themselves as healthier than women.The same argument applies in terms of the search for healthcare services.
Our data also reinforces the need to, beyond cultural gender models, take into consideration other aspects such as age, years of schooling and socioeconomic situation.The influence of these models may take on different tones when these aspects are included.Another aspect that should be further explored is the relationship between gender and ethnicity, as the perceptions of the native Indian men we interviewed are different, as are those of Oriental and native Indian women.
This topic deserves further study, as there are few population studies in Brazil that look into the different perceptions of health among men and women.
Lastly, we mention that one of the limiting factors in this study is the absence of a gold standard for assessing health perception, the small size of the sample and the consequence impossibility of further stratifications.
*weighted to represent the male and female population aged 20 to 59 living in state capitals and the Federal District.1.No data: 4 men and 1 woman.2. No data: 4 women.

Table 2 .
Distribution * and (% and 95% Confidence Interval) of male and female perception of healthcare services, health status and health care.Brazil, 2014. the male and female population aged 20 to 59 living in state capitals and the Federal District.1.Only for those claiming a healthcare problem.2. Only for those with a healthcare problem and not in service * weighted to represent

Table 3 .
Prevalence, prevalence ratio (PR), adjusted prevalence ratio (PRadj) for all variables and their respective 95% confidence interval (CI 95% ) for the reasons given by men and women for not seeking healthcare in the presence of a health problem.Brazil, 2014.
* weighted to represent the male and female population aged 20 to 59 living in state capitals and the Federal District.