Article Text

Download PDFPDF

Original article
Mesoamerican nephropathy: geographical distribution and time trends of chronic kidney disease mortality between 1970 and 2012 in Costa Rica
  1. Catharina Wesseling1,
  2. Berna van Wendel de Joode2,
  3. Jennifer Crowe2,
  4. Ralf Rittner3,
  5. Negin A Sanati3,
  6. Christer Hogstedt1,
  7. Kristina Jakobsson3
  1. 1Unit of Occupational Medicine, Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden
  2. 2Program on Work, Environment and Health in Central America (SALTRA), Central American Institute for Studies on Toxic Substances (IRET), Universidad Nacional, Heredia, Costa Rica
  3. 3Division of Occupational and Environmental Medicine, Lund University, Lund, Sweden
  1. Correspondence to Professor Catharina Wesseling, Apdo 2291-1000, San José, Costa Rica; inekewesseling{at}gmail.com

Abstract

Objectives Mesoamerican nephropathy is an epidemic of chronic kidney disease (CKD) unrelated to traditional causes, mostly observed in sugarcane workers. We analysed CKD mortality in Costa Rica to explore when and where the epidemic emerged, sex and age patterns, and relationship with altitude, climate and sugarcane production.

Methods SMRs for CKD deaths (1970–2012) among population aged ≥20 were computed for 7 provinces and 81 counties over 4 time periods. Time trends were assessed with age-standardised mortality rates. We qualitatively examined relations between mortality and data on altitude, climate and sugarcane production.

Results During 1970–2012, age-adjusted mortality rates in the Guanacaste province increased among men from 4.4 to 38.5 per 100 000 vs 3.6–8.4 in the rest of Costa Rica, and among women from 2.3 to 10.7 per 100 000 vs 2.6–5.0 in the rest of Costa Rica. A significant moderate excess mortality was observed among men in Guanacaste already in the mid-1970s, steeply increasing thereafter; a similar female excess mortality appeared a decade later, remaining stable. Male age-specific rates were high in Guanacaste for age categories ≥30, and since the late 1990s also for age range 20–29. The male spatiotemporal patterns roughly followed sugarcane expansion in hot, dry lowlands with manual harvesting.

Conclusions Excess CKD mortality occurs primarily in Guanacaste lowlands and was already present 4 decades ago. The increasing rates among Guanacaste men in hot, dry lowland counties with sugarcane are consistent with an occupational component. Stable moderate increases among women, and among men in counties without sugarcane, suggest coexisting environmental risk factors.

View Full Text

Statistics from Altmetric.com

What this paper adds

  • Mesoamerican nephropathy (MeN) is a chronic kidney disease (CKD) epidemic unrelated to traditional causes, most prevalent among relatively young men, in particular sugarcane workers. This study provides the first systematic data for Costa Rica and is the first in Central America to explore the history of MeN as far back as 1970.

  • The epidemic, so far, is restricted to the hottest province, Guanacaste. Among men, excess CKD mortality was already noticeable in the mid-1970s and has markedly increased since, with a progressive shift towards younger age groups. Among women, excess CKD mortality emerged later in the mid-1980s, and has been moderate and stable over time, and without a shift towards younger age groups.

  • The spatiotemporal mortality pattern among men follows roughly the sugarcane expansion in hot and dry lowlands, and is compatible with an occupational component. The smaller and stable mortality excess among women in Guanacaste, as well as among men in the hot counties of Guanacaste without large-scale sugarcane production, suggest the coexistence of yet unidentified environmental risk factors, possibly interacting with occupational risk factors.

Introduction

Mesoamerican nephropathy (MeN) is an epidemic of chronic kidney disease of non-traditional (CKDnT) origin (CKDnT; also called chronic kidney disease of unknown (CKDu) origin) occurring along the Pacific coast of Mesoamerica.1–6 MeN was first observed in the 1990s among sugarcane workers in Nicaragua, El Salvador and Costa Rica,7–9 but probably extends also to Guatemala, Honduras and Panama10 and Mexico.2 ,3 The disease has overwhelmed local health systems6 and has caused premature deaths of at least 20 000 men.4 MeN has been characterised in Nicaragua and El Salvador.11–13 The disease occurs predominantly among young and middle-aged male agricultural workers, especially sugarcane workers, without conventional risk factors such as hypertension and diabetes.6 Clinically, the disease resembles a tubulointerstitial disease with low-grade proteinuria,11 ,12 ,14 but biopsy series from El Salvador showed a mixed pattern of tubular atrophy, interstitial fibrosis and extensive glomerulosclerosis.15 ,16

Knowledge about geographical boundaries and the approximate time period in which the epidemic emerged in Mesoamerica is anecdotal. It is thought that CKDnT has been increasing since the late 1990s, when early technical reports started addressing CKDnT in Nicaragua,9 and the number of patients with end-stage renal disease rapidly increased in the hospitals of El Salvador.17 Although the epidemic concentrates in the lowlands of the Pacific coast extending over at least El Salvador, Nicaragua and Costa Rica, not all communities in these lowlands have a high prevalence.11 Also, increased CKD prevalence has not been reported in locations at higher altitude, including a sugarcane growing village at 500 m above sea level (masl).12 One study reported high prevalence of CKD among men and women in two agricultural communities at 300 masl in El Salvador.18

Regarding Costa Rica, clinicians, more than a decade ago, noted an increased occurrence of CKD in the Guanacaste province,8 the North Pacific region of the country bordering with Nicaragua. Like in Nicaragua and El Salvador, the disease has been anecdotally linked to sugarcane work.19 Vital statistics are available in Costa Rica as far back as 1970, which provides an opportunity to explore mortality patterns of CKD decades back in time. This descriptive study analysed the geographical distribution in conjunction with time trends of mortality from CKD in Costa Rica, with emphasis on the Guanacaste province. We aimed at answering the following questions: (1) When and where did the epidemic emerge? (2) Do patterns differ among men and women? (3) Is there a shift in mortality towards younger age groups over time? (4) Does excess CKD mortality seem to be related to altitude, climate or sugarcane production?

Methods

The administrative geographical division of Costa Rica consists of seven provinces (San José, Alajuela, Cartago, Heredia, Guanacaste, Puntarenas and Limón), together with 81 counties, 11 of which are located in Guanacaste. Mortality statistics between 1970 and 2012 were extracted from the Central American Population Center (CCP) of the University of Costa Rica, which hosts the official vital statistics of Costa Rica (http://ccp.ucr.ac.cr/). During this 43-year period, three versions of the International Classification of Diseases (ICD) were used. We extracted the cases of death for all categories of CKD and unspecified renal failures (no categories of acute kidney failure were included): codes 582, 583, 584 of ICD8 during 1970–1979 (http://www.wolfbane.com/icd/icd8h.htm), codes 582, 583, 585, 586, 587 of ICD9 during 1980–1996 (http://www.wolfbane.com/icd/icd9h.htm) and N18, N19 of ICD10 during 1997–2012 (2nd edition, http://apps.who.int/classifications/apps/icd/icd10online2003/fr-icd.htm) (table 1). Cases of deaths were retrieved for the seven provinces and 81 counties, by sex and 10-year age strata.

Table 1

Cases of death from chronic kidney disease in the Guanacaste province and the rest of Costa Rica, by time periods of use of versions 8, 9 and 10 of the International Classification of Disease (ICD), 1970–2012, and person-years during the period

Population data were obtained from the same CCP source for the censuses conducted in 1973, 1984, 2000 and 2011 (table 1), also by sex and 10-year age strata for the seven provinces and 81 counties. Extrapolation of the census data to in-between years was based on general population estimates.20 ,21 We derived population numbers for the sex, age and county-specific strata by multiplying these estimates with the fractions of the corresponding strata in the nearest census.

We used the Rapid Inquiry Facility (RIF) to assess geographical differences across provinces and counties for four time periods, 1970–1982, 1983–1992, 1993–2002 and 2003–2012. RIF is embedded in a geographical information system (GIS) that requires ArcGIS and connects to an external database of geocoded health and population data, and is especially useful for spatial analyses around sources of presumed environmental hazards, and in small areas.22 ,23 With the RIF programme, we produced maps of indirectly standardised disease risks (ie, SMRs) with the Costa Rican population as reference, for men and women separately. This programme used empirical Bayes smoothing of the relative risks towards the global mean to account for sampling variability in the observed data but without accounting for spatial autocorrelation to avoid masking of the true risk distribution due to oversmoothing.

Time trends in the seven provinces were based on age-standardised mortality rates per 100 000 men and women over nine time periods (1970–1972 being a 3-year and all other 5-year periods), as well as rate ratios (RR) along with their 95% CIs for Guanacaste versus the rest of the country. In addition, age-adjusted time trends were computed for the counties within Guanacaste, over four time periods due to smaller number of deaths. For the direct age adjustment, the most recent WHO standard population was used.24 Trends were also computed for 10-year age-specific mortality rates,25 comparing Guanacaste with the rest of Costa Rica, for men over the nine periods and for women over four periods, due to the smaller number of deaths among females.

Maps were produced at the county level with averages for elevation, temperature and precipitation together with the location of the six sugarcane production areas of Costa Rica. Altitude data were obtained from the Digital Elevation Model (from El Atlas Digital de Costa Rica 2014), and climate data from the National Meteorological Institute (http://www.imn.ac.cr/). A temperature map was produced by interpolating the average temperature registered by the weather stations in the country between 1998 and 2002. A precipitation map was created by interpolating the average rainfall registered by the meteorological stations during 2008. The map with location of current sugarcane production areas was based on data of the Agro-Industrial Sugarcane League (LAICA for its acronym in Spanish).26 In addition, production data for the different sugarcane areas were collected from documents in the library of LAICA (http://www.laica.co.cr/biblioteca2/index.do), specifically hectares of cane over time, months of peak harvesting, degree of mechanisation, and labour organisation for harvesting. Relations between CKD mortality and altitude, climate and characteristics of sugarcane production were assessed qualitatively.

Results

Emergence and geographical spread of the epidemic

During the 43 years between 1970 and 2012, 3843 men and 2452 women died from any kind of chronic or unspecified kidney disease or failure in Costa Rica, 19% (23% and 12% of total male and female fatalities, respectively) in the Guanacaste province, with 7.5% of the person years in the follow-up period (see table 1).

Figure 1 contains maps with smoothed SMRs for the seven provinces and 81 counties of Costa Rica over the four time periods, for men and women 20 years and older. At a province level, excess mortality is only observed in the Guanacaste province for both sexes. Among men, a moderately increased SMR is observed in Guanacaste already in the 1970s, whereas for women a similar increase is observed a decade later. In the most recent decade 2003–2012, mortality among men in Guanacaste is highly increased, whereas among women it remains moderately increased as compared with the other provinces of Costa Rica, none of which showed an increased risk at any time.

Figure 1

Smoothed SMRs of chronic kidney disease (CKD) in male and female population age 20 and older, by seven provinces and 81 counties of Costa Rica, for four time periods between 1970 and 2012.

The geographic patterns by counties show that SMRs of three and higher only occurred in the Guanacaste province (see figure 1). Within Guanacaste, among men the highly increased mortality extended geographically from two counties in the 1970s to six in the most recent decade, specifically, the northwestern counties all bordering each other. Also for women in Guanacaste, increased SMRs are observed in an increasing number of counties over time, but the excess is moderate and the geographical pattern is somewhat different from the male pattern. Although a number of counties of other provinces also show increased SMRs, only in one county of the Puntarenas province in the southwest tip of Costa Rica, a moderate excess among men is observed consistently since the 1980s. Further, it is noticeable that in the most recent decade, a moderately increased mortality extends also to counties of other provinces bordering Guanacaste, in the northeast border towards the Alajuela province for both men and women, and in the southeast border to the Puntarenas province for women. The rates for counties and provinces for the four time periods are shown in online supplementary table S1.

Analyses restricted to population in the working age range 30–59 showed similar results (data not shown), but SMRs in Guanacaste were generally higher than in unrestricted analyses.

Time trends of mortality among men and women

Figure 2 shows age-adjusted mortality rates per 100 000 men and women, and RR for the Guanacaste province versus the rest of Costa Rica, by 5-year periods during 1970–2012, stratified by sex. CKD mortality increased over time markedly more in Guanacaste than in the rest of Costa Rica. In Guanacaste, mortality rates among men increased almost ninefold over the four decades, from 4.4 to 38.5 per 100 000, and among women more than fourfold, from 2.3 to 10.7 per 100 000, whereas in the rest of Costa Rica (excluding Guanacaste), rates approximately doubled for both sexes, from 3.6 to 8.4 for men and from 2.6 to 5.0 for women.

Figure 2

Age-adjusted chronic kidney disease (CKD) mortality rates per 100 000 men and women and rate ratios (RR) with 95% CIs for Guanacaste versus the rest of Costa Rica, by 5-year periods, 1970–2012.

Statistically significant excess mortality emerged in the Guanacaste Province in the mid-1970s for men (RR=1.7, 95% CI 1.1 to 2.7, during 1973–1977), and mortality is currently almost five times higher than in the rest of the country (RR=4.6, 95% CI 3.4 to 5.3 during 2008–2012). Similar significant excess mortality among Guanacaste women appeared a decade later (RR=1.7, 95% CI 1.1 to 2.7 during 1983–1987), and has remained approximately double the mortality rate of the rest of the country through today (RR=2.1, 95% CI 1.7 to 2.7 for the time period 2008–2012). The mortality rates for each of the six provinces that together compose the rest of Costa Rica were very similar (data not shown) and did not deviate significantly from the combined rate shown in figure 2.

Within the Guanacaste province, the age-adjusted CKD mortality trends show rather large differences between the counties among men (see online supplementary figure S1). The county with the fastest mortality growth and highest current rate is Cañas (73.7 per 100 000 men during 2003–2012): a 13-fold mortality rate as compared to the 1970s (5.6 per 100 000), and a sevenfold rate compared with Nandayure, the county with the lowest rate in Guanacaste (10.3 per 100 000). Female rates are clearly lower in the northwestern counties as compared with male rates, but in the southern counties, the sex differences are much smaller. From the 1990s onwards, mortality rates are higher than in the rest of Costa Rica combined, in all counties for males and in 9 of the 11 counties for females.

Age-specific mortality trends

Online supplementary figure S2 illustrates 10-year age-specific mortality rates among men and women in Guanacaste and the rest of the country. Except for boys aged 19 or younger, excess mortality is observed among men in Guanacaste as compared to the rest of Costa Rica for all age groups, and with increasing trends over time (section A). Excess CKD deaths in Guanacaste among men occurred already in the 1970s in the age categories between 30 and 69 years, and from the mid-1980s in the age categories of 70 years and older; a much smaller but increasing excess among young men aged 20–29 emerged from the late 1990s. Trends of age-specific mortality rates for women are less clear (section B). Nonetheless, from the mid-1980s onwards, mortality rates are slightly to moderately elevated for age groups between ages 20 and 69 among Guanacaste women as compared with women in the rest of the country.

Altitude, climate and sugarcane production

Figure 3 displays maps of elevation, temperature, precipitation and location of sugarcane production areas. Characteristics of the six sugarcane production areas of Costa Rica (so-called North Pacific, Central Pacific, North Zone, South Zone, Occidental Central and Oriental Central) are reported in table 2, and in more detail together with the data sources in online supplementary table S2. Between 1973 and 2013, the production in the North Pacific almost quadrupled the cane hectarage from 8500 to 32 850 Ha, an expansion over five times larger than in the Central Pacific area and the North and South Zones, whereas, the Occidental Central and Oriental Central production areas decreased considerably. The North and Central Pacific production areas are located at the lowest altitude; have the hottest, driest and sunniest climate (see figure 3 and see online supplementary table S2) and, in addition, the longest harvest time. The North and South Zones are also very hot, but much of the production takes place at somewhat higher altitudes with more rainfall and less solar irradiation, and the harvest time is considerably shorter as compared to the Pacific areas. The Occidental Central and Oriental Central sugarcane areas are located at higher altitudes with cooler and rainier climates. The Central Pacific has most mechanised harvesting (90% in 2006) followed by the North Pacific and the North Zone (approximately 35%).

Table 2

Production characteristics in the six sugarcane production areas of Costa Rica

Figure 3

(A) Sugarcane production areas of Costa Rica: (1) North Pacific, (2) Central Pacific, (3) North Zone, (4) South Zone, (5) Occidental Central, (6) Oriental Central; (B) Elevation; (C) Mean temperature (1998–2002); (D) Mean precipitation (2008).

Of the six production areas, only the North Pacific shows a clear excess in CKD mortality over time, which in the most recent time period is extending beyond Guanacaste towards the northeast (men and women) and towards the southeast (women) (see figure 1). Within the Guanacaste province, five of the six northwestern counties with the highest and most rapidly increasing mortality rates among men are important sugarcane producing counties; by contrast, one of the five most southern counties with considerably lower rates (but still elevated compared to the rest of Costa Rica) has appreciable sugarcane production (see online supplementary figure S1).

Discussion

The main findings of this study are that excess CKD mortality concentrates in the Guanacaste province and is observable among men from the mid-1970s with an ongoing, steeply increasing trend, whereas among women, a stable moderately increased mortality is observed since the mid-1980s. Clear excess mortality among men starts from age 30 to 39, but over the past 15 years there is a shift in mortality towards younger men aged 20–29. Within Guanacaste, CKD mortality is elevated, in particular in the counties with hot and dry lowlands plus extensive sugarcane production.

Comparison with other Mesoamerican countries

Recently the Pan American Health Organization (PAHO) called the world's attention to the epidemic of CKDnT in Mesoamerica by comparing national CKD mortality of five Mesoamerican countries with USA and Cuba.10 ,27 El Salvador and Nicaragua stand out with national age-adjusted male mortality rates increasing from about 45 per 100 000 men in 2000 to well over 60 in 2009, and considerably lower but increasing mortality rates among women in these same countries, on the order of 15–25 per 100 000 women over the decade.10 In the PAHO statistics, the national Costa Rican rates are below the other Mesoamerican countries and just above the USA, around 8–9 and 4–5 for men and women, respectively, during the 2000–2009 decade. These PAHO figures (very similar to our estimates for the entire male and female populations of Costa Rica) mask the epidemic in a small area like Guanacaste, with just over 7% of the country's population. Our results showing excess deaths in young men starting at age 20 are in agreement with the PAHO statistics for Nicaragua and El Salvador.10

Recurrent heat stress and dehydration in hot occupations, in particular among sugarcane workers, have been proposed as key aetiological factors of MeN1 ,28 based on epidemiologialc data,12 ,14 occupational hygiene data,29 and experimental data of repeated dehydration inducing CKD in mice.30 ,31 Other potential causes that have been proposed as key or contributing factors include arsenic, cadmium, pesticides, nephrotoxic medications, infectious agents and genetic susceptibility.1–5 Although this is a descriptive study, the data merit scrutiny in the light of their compatibility with various hypotheses, which should be further investigated with analytical study designs.

MeN, an occupational disease

Multiple results from this study are in line with what would be expected for an occupationally related disease: the epidemic appears first in men, increases much more rapidly among men, and shifts towards younger age groups over time among men only. For several reasons, heavy work on sugarcane fields stands out as a potential high-risk job. First, the disease was anecdotally noted by sugarcane workers themselves in Guanacaste as far back as in the early 1990s32 as well as by nephrologists somewhat later;8 ,19 second, sugarcane workers in the Pacific lowlands have been found to be at increased risk for kidney dysfunction in El Salvador12 and Nicaragua;11 ,14 and third, sugarcane may have become an increasing source of employment during the follow-up time, with production in Guanacaste increasing from less than 5000 Ha in 1973–18 000 Ha in 1991 and about 35 000 Ha in 2012 (see online supplementary table S2). Crowe et al29 found that sugarcane cutters in Guanacaste, Costa Rica, were at risk for heat stress during the majority of their shift. McClean et al14 observed a decline in kidney function over a one-harvest season harvest among Nicaraguan sugarcane workers in the job categories with the highest heat exposures.

Our data do not support that work in sugarcane per se underlies the epidemic, since only one of the six sugarcane areas coincides with increased mortality risk. In Guanacaste, however, the counties with hot and dry lowlands plus large sugarcane production areas with manual harvesting have very high risk for workers’ heat stress, whereas other sugarcane regions have one or more characteristics that could explain lower risk for heat stress (see online supplementary table S2). We believe that, most likely, the excess of CKD is related to a combination of geographical, climatic, technical and organisational factors during harvesting.

Kidney dysfunction has also been reported in other hot occupations in MeN-affected areas of Central America, specifically construction workers, miners and port workers in Nicaragua,14 and cotton workers12 and corn and subsistence farmers18 in El Salvador. We did not explore potential associations with other important crops or hot industries in Guanacaste. However, no other crop has experienced an expansion as large as sugarcane. Rice is a traditional crop with a stable production of about 25 000 Ha since the past several decades; production of melon, mango and orange combined, although increasing since the 1980s, adds to about 15 000 Ha, less than half the sugarcane hectarage (Agricultural Census of 2014, http://www.inec.go.cr/Web/Home/GeneradorPagina.aspx). Of note is also that in the most recent time period the high mortality seems to be spreading to neighbouring areas of Guanacaste where no sugarcane production takes place. Therefore, workers in other jobs with heat stress are likely to contribute to the risk for MeN in Guanacaste and beyond.

Overall, the epidemic in Guanacaste has not reached the level of severity currently observed in El Salvador and Nicaragua.10 ,27 The highest mortality rates computed for Guanacaste (38.5 and 10.7 per 100 000 men and women, respectively, for the period 2008–2012) are well below the national rates of El Salvador and Nicaragua in 2009, and in these countries mortality is even higher in MeN affected areas. For example, during 2009–2011 the mortality rates of departments in northwest Nicaragua were more than three times the national rate.33 No data exist on number of sugarcane workers in the MeN affected areas in the different countries to explain these differences, but based on the Food and Agricultural Organisation's national production statistics, it can be reasonably assumed that hectares of sugarcane production in the lowlands of El Salvador and Nicaragua, and hence number of workers, have exceeded those in Guanacaste historically (http://faostat.fao.org/site/567/DesktopDefault.aspx?PageID=567#ancor). In addition, an increasing proportion of sugarcane harvesters in Guanacaste are seasonal migrant workers from Nicaragua, most of whom probably do not count in the Costa Rican mortality statistics if they fall ill with CKD.

A possible environmental component

It is intriguing that women in Guanacaste have a higher CKD mortality than women elsewhere in Costa Rica. Because we have no individual data besides age, sex and residence, we do not have information about their occupational history. However, few women in Guanacaste work in agriculture (0.4% and 0.8% of total female population according to census data of 2000 and 2011, respectively) and, with all likelihood, not many of these in heavy sugarcane labour, which is most often performed by men.34 This may indicate that there exists another factor that affects both sexes. The finding that all counties of Guanacaste had mortality rates above average for the rest of Costa Rica for men and almost all counties for women, including those without sugarcane production, is also consistent with this hypothesis. Thus, possibly another factor, or factors, present in the general environment might interact with occupational risk factors, to produce an epidemic pattern among men.

A possible general environmental risk factor may be the quality of drinking water, which is a major community concern. The general public in Guanacaste is especially worried about contamination of their drinking water with arsenic or pesticides. Arsenic has been found in high levels in some of the affected areas.35 Arsenic has also been proposed as an aetiological factor of a similar CKD epidemic of unknown cause in Sri Lanka,36 although there is no consensus about it among scientists of this region. A comprehensive review found some evidence for an association with CKD mortality based on ecological studies, but judged the overall evidence of a causal association between arsenic and CKD as insufficient, and emphasised the need for studies with improved design.37 Pesticides are also an intensely debated potential aetiology.6 ,10 ,27 McClean et al14 did not find a decline of kidney function among pesticide applicators in sugarcane over the course of one harvest season. Other cross-sectional studies, however, report associations of MeN with exposure to pesticides among agricultural workers, albeit based on qualitative measures of exposure.38–40 Studies on pesticides with high-quality exposure assessment are warranted. In addition, exposures to other potential aetiological factors including the frequent use of non-steroidal anti-inflammatory drugs to alleviate muscle and articular pains associated with heavy labour, fertilisers, infectious agents and genetic susceptibility need to be assessed. Ideally, such studies should be conducted by using an ecosystem health approach.1

Limitations and strengths of the study

This study has several limitations. Uncertain quality of mortality statistics, in particular during the 1970s, is one concern. An evaluation of errors in vital statistics in Costa Rica, carried out after the census of 2000, reported an estimated 8% of under-registration of deaths nationally during the year 1970 with an additional 27% of deaths registered without death certificates; however, this unfavourable situation gradually improved to an estimated 0.3% under-registration, and less than 1% of deaths registered without certificates during 2000.41 Deficient quality of mortality statistics could explain, in part, that there were much fewer cases of CKD-related deaths registered in the early observation periods than today, in Costa Rica in general, and maybe especially for Guanacaste, an area of little development during the 1970s. However, there is a clear excess mortality among males in Guanacaste already during the 1970s, but not among females. It seems unlikely that registration in Guanacaste would have had better quality than in other regions of the country, including the capital with the main hospitals, and in addition, differentially for men. Therefore, these data support the hypothesis that the epidemic of MeN in Guanacaste started during the 1970s for men or, if there was an under-ascertainment, possibly earlier.

In Costa Rica, the diagnoses of the physicians on all death certificates are centrally recoded into ICD categories by trained personnel at the National Institute of Census and Statistics (INEC). The accuracy of codification at INEC has never been evaluated. However, considering the centralised nature of the codification process, any errors are unlikely to be related differentially to a specific geographical area. It is possible, though, that greater awareness among clinicians in Guanacaste has improved diagnostic quality differentially especially in the more recent years, but the pattern of CKD mortality in Guanacaste, with large differences for men and women since early in the epidemic, precludes that the increase is solely due to improved diagnoses noted on death certificates.

Besides categories of chronic kidney disease, we also included in the analyses categories of nephritis and nephrosis not specified as acute or chronic and unspecified kidney failure. These undefined nephropathies form a relatively small fraction (11%) of the total cases, and less in Guanacaste than in the rest of Costa Rica (4.0% vs 11.5% for men, and 8.5% vs 15% for women). Exclusion of these categories would produce a somewhat higher risk estimates for the Guanacaste population. Furthermore, all categories include cases that are not CKDnT, but known risk factors such as diabetes, hypertension and smoking have not increased more in Guanacaste than in the rest of Costa Rica.42 In addition, nephropathies from diabetes and hypertension have their own ICD codes and these codes were not included our analyses. Therefore, it seems reasonable to attribute the excess of cases in Guanacaste to MeN.

With regard to exploring a possible relationship between excess CKD mortality and sugarcane production, the interpretation is limited by our qualitative assessment of rather disperse historical data on technical and organisational issues related to harvesting in the different sugarcane producing areas as well as the absence of labour force data. Nevertheless, the excess of CKD in hot, dry lowland counties with sugarcane production in Guanacaste is consistent with the results of other studies that have reported increased risk for either heat stress or CKD among sugarcane workers.8 ,11 ,12 ,14 ,29

A strength of the study is that it provides the first systematic data for Costa Rica in relation to MeN, based on mortality statistics of more than four decades, with a large number of cases of deaths due to kidney failure, in particular for men. Although this is a descriptive epidemiological study and no causal conclusions can be drawn, it provides a novel insight into the origin of the epidemic and its development over time which may guide further aetiological research.

Conclusions

In conclusion, the patterns of geographical distributions and time trends of CKD mortality in Costa Rica indicate that excess mortality occurs primarily in the Guanacaste province and was already present four decades ago among men. The increasing rates among men in the hot and dry lowlands of Guanacaste with extensive sugarcane production are consistent with an occupational component in the epidemic of MeN, with heat stress and dehydration as possible stressors. The stable moderately increased rates among women, and also among men in counties without sugarcane production, suggest the coexistence of one or more as yet unidentified environmental risk factors. Epidemiological research targeting specific occupational and environmental risk factors and their interactions must follow this descriptive study.

Acknowledgments

The authors wish to acknowledge the valuable work of the Central American Population Center (CCP) of the University of Costa Rica. Through its website (http://ccp.ucr.ac.cr), vital statistics, census and other relevant data are publicly accessible.

References

View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • Contributors CW conceived the design of the study, all authors contributed to refinement of the study protocol. CW, BvWdJ collected the data. CW, BvWdJ, JC, RR, NAS and KJ contributed to the analyses of the data. CW wrote the manuscript and CH contributed to the writing. All authors critically revised and approved the manuscript.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.