In study GBD-2017, depressive disorders were among the four major causes of disability in Brazil, following low back pain, headache and anxiety disorders, and accounted for a high number of years lived with disability, especially in women, working-age adults and residents of the Southern region of the country. Despite the decrease in the age-standardized rate of burden associated with depressive disorders in the last 27 years, there was a considerable increase in the number of YLD, which means an enlargement in demand for services and requires greater efforts in public health to provide adequate assistance to the Brazilian population.
There were significant differences in the prevalence of depression among some Brazilian FUs. Due to its large territorial extension, the investigation of prevalence of depression in all regions of the country is fundamental, since regional differences can be associated with variations in the rates of these disorders [31]. PNS-2013 found a significantly higher point prevalence of individuals at greater risk of depression in the Southern region (4.8%, 95% CI: 4.1–5.4) and lower in the Northern region (2.9%, 95% CI: 2.4–3.3) using the Patient Health Questionnaire-9 (PHQ-9) [8]. When considering self-report of prior medical diagnosis of depression, the prevalence was 12.6% (95% CI: 11.2–13.9) in the South and 3.1% (95% CI: 2.7–3.5) in the North of the country [32]. This result is corroborated by other studies that used the same measuring instrument and found a prevalence of depressive symptoms of 20.4% (95% CI: 18.9–21.8) in Rio Grande do Sul [33] and 7% (95% CI: 6–8) in Amazonas [34].
In the study GBD-2017, in addition to the high prevalence found in the Southern states, the states of Roraima (3.67), in the Northern region, and Pernambuco (3.52) and Alagoas (3.48), in the Northeastern, stood out. Comparability of these data is a complex task, since the estimates of GBD study are based not only on raw data of the studies available, but also on adjustments based on covariables and other procedures [35]. Furthermore, there is a scarcity of studies for the Northern and Northeastern regions of the country [9].
It is worth mentioning that the states with the highest prevalence of depression also present with the highest mortality rates from suicide, according to data from studies showing depressive disorders are among the major risk factors for suicide deaths [4]. Rio Grande do Sul has the highest rate in the country: 10.5 per 100,000 (95% UI: 6.9–13). Roraima ranked 2nd, with a suicide rate of 9.4 per 100,000 (95% UI: 7.2–11.2) [36], indicating there is still much to investigate on the subject. It also draws attention the fact that Pernambuco and Alagoas were the states with the highest mortality rates due to violence in 2015 [37, 38]; and violence is an important risk factor for onset and aggravation of depression [10].
Between 1990 and 2017, there was a considerable increase in the number of YLD for depression in Brazil. Therefore, greater attention must be given to mental health, since there are more people living with depression and the tendency is for it to increase as the population ages. Especially in developing countries such as Brazil, the increased life expectancy due to improved reproductive health, nutrition and control of infectious diseases in childhood, results in more people living until adulthood, the mean age of population increases, and the burden of disease shifts to noncommunicable and chronic diseases and disabilities, such as depression [13].
In Brazil, the epidemiological transition does not occur homogeneously among its regions; i.e., the less developed regions of the North and Northeast, with lower SDI, present a slower transition than the Southern and Southeastern regions [38]. Thus, the YLDs of the Southern states resemble those of high SDI countries, while the YLDs of most Northern states are closer to countries with medium SDI values. Although many FUs significantly improved as to fertility rates, income per capita, and mean years of education (SDI), there was an increase in the number of YLDs from depressive disorders in recent years, indicating a challenge for Brazilian mental health care.
It is known that, the higher the SDI, the lower the mortality rates (YLL) for communicable, maternal, neonatal and nutritional diseases [39]. Therefore, in high-income countries, such as Canada, Australia and England, the most prevalent and disabling diseases, with lower mortality, such as depressive disorders, stand out and account for the highest positions in the DALY classification, whereas in low- and middle-income countries, such as Brazil, Mexico, Colombia and Argentina, diseases with higher mortality rates still prevail [13].
Thus, the burden of disability (YLD) of depressive disorders does not seem to vary according to development of countries as measured by SDI, since MDD ranked among the ten major causes of YLD in 191 out of 195 countries analyzed by GBD-2016 [40]. In the GBD-2017 study, the YLD of these disorders in Brazil and in other countries, such as Mexico and Argentina, considered middle-income countries, was similar to the burden of high-income, politically and economically stable countries, like the USA, Canada, England and Australia.
The lack of a relation between the YLD of depressive disorders and SDI may be associated to a limitation of GBD estimates, which refers to the scarcity of epidemiological data, especially in low- and middle- income countries and in places with subnational estimates, such as Brazil, hindering confidence in variations of prevalence and burden.
Moreover, the lack of a standard that allows predicting the burden of depressive disorders as a function of SDI may suggest that the burden of these disorders depends on factors other than those measured through SDI. It is, therefore, necessary to better understand the relation of SDI with the epidemiological factors of these disorders, at the individual level, before any interpretation.
It is a fact that social inequalities in income and education levels, included in SDI, are risk factors for depression, as revealed by a meta-analysis involving 56 studies from different countries [41]. In the National Health Survey (PNS-2013), depression was also associated with low levels of education [8], although its relation with income and fertility rates was not evaluated.
SDI allows monitoring not only the development of countries/regions over time, but also calculating the expected estimates for each region, given its level of development [13]. In Brazil, the disability generated by depressive disorders was within expectation, given the SDI. In most of the Southern FUs, the observed YLD rate was higher than expected, while in most FUs in the North and Northeast, the opposite happened. Thus, although Southern states have better access to treatment [42], the impact of YLD on depressive disorders was greater in this region, as found in countries with high SDI [13]. In Brazil, 78.8% of individuals with depressive symptoms receive no type of treatment for this problem; in that, the Northern region has the largest proportion of untreated individuals (more than 90%), and the Southern region, the lowest proportion (67.5%) [42].
The results of this study corroborate findings from different regions of the world of depression affecting predominantly women [43–48], for reasons related to both biological and social factors [48, 49]. Regarding age, it is worrying that the burden of disability of these disorders is greater precisely in the working-age population, since depression is related to an important loss of productive potential, causing these people to be away from work [50]. In the USA, the prevalence of depression is increasing more rapidly among younger people, which may, over time, reduce the prevalence gradient differences between age groups [51]. These data show the urgency of investing mainly in preventive actions, early detection and improving quality of services available for treatment of depression, focusing on the risk factors and predictors that may influence the prevalence and burden of these disorders [52].
Among the strengths of GBD study is the addition of covariates that best predict prevalence, the expansion of epidemiological data on mental disorders, and the improvement of subnational estimates. Regarding Brazil, as well as other countries, it is important to evaluate the need to include covariate on child sexual abuse and intimate partner violence in the depression estimates model of GBD, considering that interpersonal violence is one of the major causes of burden of disease in the country [37, 38].
There is also a need for ongoing studies with sustainable population data, which allow assessing the prevalence of depression in FUs, as well as identifying demographic subgroups that require more interventions. Trends of past-year depression from 2005 to 2015 in the US study [51] indicated the overall prevalence of depression increased significantly over this period, mainly due to stress, related to lack of employment and low income. Brazil is currently experiencing one of the biggest economic crisis in its history, with 13,7 million unemployed [53], which is likely to affect the scenario of estimates of the burden of depressive disorders in the coming years.
In terms of GBD limitations to estimate the burden of mental disorders, including depressive disorders, it must be emphasized that the low coverage of epidemiological data on mental health, as previously mentioned, especially in less developed regions, such as Brazil [38, 44], makes the real contribution of these disorders to the global burden of disease still underestimated. In primary care and other general medical services, it is estimated that 30 to 50% of cases are undiagnosed [54], meaning the challenge of ensuring an increase in the population's healthy life expectancy is greater than expected.
In addition, the distribution of severity levels of MDD and dysthymia was derived from a limited number of sources of data from high-income countries, which limited the overall representativeness of distribution of disorder severity. There is, therefore, a need for further studies with comparable methods in the distribution of MDD and dysthymia severity, and their variation among countries and levels of access to care [4].