Temporal trend of leprosy among the elderly in Brazil, 2001 – 2018

ABSTRACT Objective. To describe the temporal trends of leprosy indicators among the elderly in Brazil in 2001 – 2018. Methods. This was an ecological time-series study of new leprosy cases in the elderly reported to the Notifiable Diseases Information System. Prais-Winsten generalized linear regression was used to estimate temporal variations. Results. There were 687 317 new leprosy cases in Brazil from 2001 – 2018, of which 129 214 (18.8%) were among elderly people. Overall detection rates in the elderly and of new cases with grade-2 disability showed a falling trend, with an annual percent change of -4.6% (95%CI = -5.1 to -4.0) and -3.9% (95%CI = -4.6 to -3.2). New case and new multibacillary case proportions showed an increasing trend, with an annual percent change of 2.9% (95%CI = 2.6 to 3.3) and 1.4% (95%CI = 1.0 to 1.7), respectively. Detection rates for new leprosy cases in elderly people in Brazil are decreasing, but the proportion of new cases and multibacillary cases are trending upwards. Conclusions. New cases are shifting to older age groups, and demographic transition and immunosenescence are an influence. Inadequate reduction of grade-2 disability indicates a high risk of physical disability persists. Improved contact tracing and more effective action are needed in this age group.

Leprosy is a chronic infectious disease, manifested by dermatological and neurological signs and symptoms (1). According to the World Health Organization (WHO), Brazil has a high leprosy burden, given the magnitude and transcendence with which the disease affects the population (2,3). Magnitude means high occurrence of a disease in the population, while transcendence is the existence of severe clinical and epidemiological characteristics of social relevance (1).
In 2017, Brazil registered the second highest number of new cases of leprosy worldwide (n = 26 875) and 92.3% of the cases in the Region of the Americas (2). Of these, 6 598 new cases were detected among the elderly, at a rate of 25.4/100 000 inhabitants (twice the overall detection rate) and 24.5% of all the new cases (4). An analysis of 591 090 new leprosy cases diagnosed in Brazil in 2001 -2013 found an increase of 6% among the elderly (5). With regard to transcendence, Brazil in 2017 was second only to India in new cases with grade-2 disability (G2D; 2); among the elderly, the detection rate for these cases was 28.8/ 1 million (4).
Temporal trend analysis of indicators in elderly people with leprosy can provide a dynamic diagnosis of its occurrence, explaining its progression and the risk of physical disabilities. This data can inform hypothesis formulation and intervention planning (6).
Aging is accompanied by a reduced ability of the body to defend itself against internal and external imbalances and attacks (7). Thus, age may influence leprosy progression since the host's immune response to Mycobacterium leprae determines This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs 3.0 IGO License, which permits use, distribution, and reproduction in any medium, provided the original work is properly cited. No modifications or commercial use of this article are permitted. In any reproduction of this article there should not be any suggestion that PAHO or this article endorse any specific organization or products. The use of the PAHO logo is not permitted. This notice should be preserved along with the article's original URL. the course and clinical forms of the disease (8). Because most cases in this age group are multibacillary (5), delayed diagnosis can assist the chain of transmission and favor the emergence of physical disability (9). This situation, in turn, impacts the social security system in Brazil, which faces the difficult mission of caring for an aging population within the context of recent, complex social security reform (10,11).
Given the rising percentage of cases among the elderly (5) and their higher risk of physical disability (12), a better understanding of leprosy within this age group was needed. The objective of this study was to describe the temporal trends of leprosy indicators among the elderly in Brazil from 2001 -2018.

MATERIALS AND METHODS
This was an ecological time-series study. The units of analysis were Brazil, its five regions (North, Northeast, Southeast, South, and Midwest), and its 27 unidades federativas (26 states and 1 federal district; UF).

Data sources
Secondary data were obtained from the Notifiable Diseases Information System (SINAN), operated by the Ministry of Health's Surveillance Secretariat. Population estimates for 2001 -2018 were produced by the Brazilian Institute of Geography and Statistics and provided by the Brazilian National Health System Information Technology Department.

Study population
The study included all new leprosy cases in the elderly diagnosed and notified to the SINAN database from 2001 -2018. The elderly were defined according to the WHO definition and the National Policy on Health Care for the Elderly (13) as individuals 60 years of age or more. Cases excluded were any new cases that had been misdiagnosed (and removed from the database) and new cases with G2D notified in 2007 when a change was made to the, disability grade notification code (4). The latter are being evaluated by the Ministry of Health.

Indicators
Four indicators were identified and analyzed:

Data analysis
Prais-Winsten generalized linear regression was used to estimate temporal trends. This type of regression considers the dependence of a serial measurement on its own values at earlier times when estimating parameters, which are controlled by first-order autocorrelation. Annual percent change (APC) of the measurements was quantified and 95% confidence intervals (95%CI) were estimated using the following formula, where b1 is the slope of the line (14): Indicator trends are considered to be increasing when APC is positive, and decreasing when change is negative; both are statistically significant when APC variation does not include zero. Trend is considered to be stationary when APC is nil, i.e., there is no significant difference between its value and zero (14).
The analyses were performed with Stata ® Statistical Software: Release 15 (StataCorp LLC, College Station, Texas, United States).

Ethics
This study was approved by the Ethics Committee of the University of Brasília Faculty of Health Sciences, Certification of Submission for Ethical Appraisal 77799417.6.0000.0030, Opinion 2.411.253, issued on 2 December 2017. The secondary SINAN data did not identify individuals, so anonymity was preserved.
The proportion of new leprosy cases among the elderly showed an increasing trend, from 16.2% in 2001 to 24.3% in 2018, with an APC of 2.9% (95%CI: 2.6 to 3.3%). This trend was similar in all regions (P < 0.05), but the increase was significantly greater in the North (APC 4.2%) compared to the South (2.7%), Southeast (2.8%), and Northeast (2.9%). There was no difference between the Midwest trend (3.2%) and that of the other regions. The South, Southeast, and Northeast had the highest proportions of new elderly cases during the entire time-series, increasing from 24% to 33%, 20% to 30.4%, and 17.4% to 25.8%, respectively ( Table 2).
The proportion of new leprosy cases in the elderly grew significantly in all UF (P < 0.05), except in Santa Catarina where it was stationary. The highest APC was observed in Sergipe (4.6%), where the increase was statistically greater than in 14 other UF (Table 2 and Figure 1b).
The proportion of new multibacillary cases in the elderly also showed an increasing trend for Brazil, from 66% to 82.2% (APC 1.4%), and every region, without statistical differences among them (P < 0.05). During the time-series, more than 60% of new elderly cases were multibacillary in Brazil, as well as each region (Table 3).
The proportion of elderly multibacillary cases increased in 22 UF (P < 0.05). Mato Grosso (APC 3.3%) and Mato Grosso do Sul (3.0%) had significantly higher APCs than 11 UF, 8 of which are in the Northeast. This trend was stationary in five UF (Table 3 and Figure 1c).

DISCUSSION
This study revealed the magnitude and transcendence trend of leprosy among the elderly in Brazil, its regions, and the 27 UF. For Brazil, the detection rate trend and the trend of new cases with G2D decreased, while the proportions of new cases and new multibacillary cases in the elderly age group increased. The result was similar in all the country's regions and most of the UF.
The decreasing trend in detection rates among the elderly is similar to that of Brazil overall (5). Nery and colleagues (14) indicate that reductions in leprosy detection in endemic municipalities from 2004 -2011 reflect lower leprosy incidence associated with improved living conditions. This could be attributed, among other factors, to conditional cash transfers by means of the Bolsa Família, a program for families in poverty and extreme poverty. It has three main objectives: to transfer income, to improve access to education and health care, and to offer complementary social programs (15).
Also during the study period, the country registered reductions in the percentage of poor people, from 43.8% to 29.8%; the Gini index, 0.56 to 0.53; the illiteracy rate, 23.1% to 19.6%; and unemployment, 9.0% to 6.9% (15). The socioeconomic determinants of new leprosy case detection is supported by other studies identifying associations between high leprosy rates and poor socioeconomic conditions (16 -18).
Also important are the influence of leprosy surveillance, decentralization of leprosy control to the primary health care network (19), and the protective effect of BCG vaccination, which ranges from 26% -61% (20). Multidrug therapy (MDT), despite being widely questioned (21,22) regarding its ability to reduce bacillary exposure among the population, continues to be a pillar of the Global Leprosy Strategy (1, 3).    Although the detection rate in the elderly reflects the overall detection rate, since both are decreasing, it reaches much higher levels in the elderly, possibly because this age group is more susceptible. During the study period, the detection rates were about twice as high among the elderly than in the general population (5,23,24). The sharper reductions in the South and Southeast compared to the North, Northeast, and Midwest may be related to the higher endemicity levels in these regions where the influence of socioeconomic factors, BCG vaccination, and MDT may lag behind. Seven UF with stationary trends are located in the Northeast and Midwest, corroborating this hypothesis. Moreover, the Northeastern UF showed a statistically lower decline than Southern and Southeastern UF. Associations found between socioeconomic factors and leprosy occurrence reveal that people living in areas with worse poverty (i.e., the Midwest, North, and Northeast) have 5 -8 times the risk of leprosy (18).
Higher leprosy incidence in the elderly group may be linked to greater odds of reinfection in endemic areas. Moreover, its long incubation period could lead to late manifestation of signs and symptoms, increasing the possibility of leprosy emerging at older ages (25,26). Immunosenescence, comorbidities, and malnutrition could also favor the high magnitude of leprosy in this age group (23,25). Furthermore, the effect of demographic transition, with low mortality and fertility rates, negative growth, a high proportion of elderly people, and increased life expectancy could interfere with reducing detection rate because the denominator becomes larger over time (27).
In 7 UF, detection in the elderly was stationary ( Table 1), meaning that leprosy magnitude did not change substantially, despite the growing size of this age group (27). Overall detection in these UF in 2017 reveals different endemicity levels, with medium, high, and hyperendemic rates (28) and a decreasing trend in overall detection in some (24). It may be that despite the downward overall trend and different endemicity levels, greater susceptibility of the elderly favors the emergence of the disease in this group, even in areas with less exposure to leprosy bacilli. Further supporting this postulation, the proportion of new leprosy cases among the elderly increased throughout Brazil, its regions, and most of the UF despite reductions in overall detection rates during the study period. The rise could be due to the long incubation period of multibacillary leprosy-it can emerge years after the initial infection, which may have occurred when the individual was younger (26). The South and Southeast had the highest proportions of leprosy cases in the elderly to total cases, which corroborates Irgens' findings (26), given that endemicity in both regions is low (4). On the other hand, the largest increases in this proportion occurred in the North, indicating a possible change in the pattern of endemicity, i.e., switching to the elderly. Although the Northeast had a pattern similar to the South and Southeast, it continued to have lower proportions of cases in elderly. With the exception of Santa Catarina, there was an increase in the proportion of the elderly with leprosy in all the UF, although without significant differences among them.
The proportion of new multibacillary leprosy cases in elderly also showed an increasing trend for Brazil, its regions, and most of its UF. Nonetheless, rates were more stable, increasing only about 1% a year, with narrow confidence intervals and without differences among regions. The more homogenous pattern within Brazil's heterogeneous epidemiologic context of leprosy could be more closely linked to the population's increased life expectancy than to changes in endemicity. Individuals with a previous infection could have more time to develop the leprosy, which in this later stage of life, would be predominantly multibacillary.
Aging comes with an increased risk of developing biological, socioeconomic, and/or psychosocial vulnerabilities due to declining senescence. These vulnerabilities reduce capacity for self-protection (29), and any physical disability arising from leprosy would further threaten an elderly individual.
This study found a decreasing trend in detection rates for cases of leprosy with G2D in the elderly in Brazil, its regions, and 13 UF. However, in 2018 this indicator reached 28.3/ 1 million, whereas for Brazil's general population, it was 10/ 1 million (4). This illustrates the challenge that Brazil faces in reaching the target of < 1/1 million, established by the Global Leprosy Strategy 2016 -2020 (3). In addition, 14 UF had a stationary trend, therefore the rates continued to be high over the time series, possibly due to delayed diagnosis (3).
With regard to declines in detection and G2D rates, the latter indicator was more heterogeneous across the UF, with 13 having a stationary trend. The detection rate trend was stationary in 7 UF. Differences were also found among the regions. UFs with decreasing detection rates were in the Southeast, and those with stationary G2D rate trends were in the North, Northeast, South, and Midwest.
Leprosy can mimic dermatological, rheumatological, and neurological diseases. Diabetic neuropathy and rheumatoid arthritis, which are also prevalent in the elderly, may hinder differential diagnosis and favor delayed treatment (30). Moreover, elderly people with disabilities have greater difficulty accessing health services (31). A higher risk of physical disability in the elderly is possibly multifactorial and related to biological issues, comorbidities, and difficulties with diagnosis and access to health services. Thus, it is essential that the Health Care Model for the Elderly and the process of qualifying care include leprosy within the scope of multi-professional health team activities, including active case tracing, contact investigation, health education, and functional assessment (7).
Although Brazilian legislation on the care of the elderly is comprehensive and advanced, in practical terms it falls far short of overcoming the social vulnerabilities to which some are exposed (13). The results of this study corroborate this worrying situation and the leprosy mortality trends (32) that show the elderly to be at far greater risk than other age groups in the North and Northeast regions of Brazil (RR 7.9; P value < 0.0001) .
Leprosy control among the elderly is important for achieving the Global Leprosy Strategy 2016 -2020 targets (3). The increasing proportions of new cases and multibacillary cases among the elderly reinforce the need for specific active tracing and contact investigations in this age group (5)-especially the multibacillary leprosy cases, which are responsible for bacillus transmission (33). Also, elderly people with leprosy should be regarded as having both a communicable disease and a chronic condition, in accordance with the "Guidelines for SUS health care for the elderly" (7).
Limitations. This study had limitations inherent to ecological studies, which cannot establish a relationship between cause and effect. Of note is the complexity of leprosy evolution, with its long incubation periods, possibility of misclassification, and attributing G2D. There were also limitations caused by changes to SINAN, which may have impacted data quality, such as under-or over-reporting (17,24). Furthermore, a decline in detection rates can be influenced by operational issues (21), such as accurate disease diagnosis; community awareness (3,15); differences in coverage and access to comprehensive health care; and operational interventions in program management (34).
Despite these limitations, ecological studies enable hypotheses to be formed and can support decision making (17). Regarding data quality, detection rates are a function of true incidence and a health system's diagnostic capacity since surveillance systems can vary operationally (19). Notwithstanding, case detection trends can be considered close to incidence when the indicator of new cases with G2D remains stable over years (24,28,33).

Conclusions
Detection rates of new leprosy cases in elderly people in Brazil are decreasing, while the proportion of new cases and multibacillary cases are trending upwards. This context indicates that new case detection is shifting to older age groups, and factors related to demographic transition and immunosenescence are having an influence. The inadequate reduction of G2D rates indicates that a higher risk of physical disability persists among the elderly despite declining detection. Specific and active tracing and contact investigation are needed in this age group.