The CKDu picture in Sri Lanka is certainly confusing

The CKDu picture in Sri Lanka is certainly confusing

Amarasiri de Silva, PhD

In a bid to seek assistance from friendly countries to combat kidney disease, the President of Sri Lanka stated in a recent meeting that about 40, 000 people were suffering from kidney disease of unknown etiology (CKDu) and that approximately 1000 persons succumbed to the disease annually (Island July 29). He further stated that even though steps were taken to improve the facilities in hospitals that attend to such patients, the number of patients was still on the rise. The inability of the government to tame the disease is partly due to the poor attention paid by the scientific community to understand and explain the socio-cultural aspects of the disease. 

The failure to take account of sociocultural factors in research so far carried out on CKDu in Sri Lanka is responsible for the present confusion and lack of understanding about the disease. Biomedical and epidemiological scientists have led us astray by their emphasis on the search on etiologies, and so the whole phenomenon remains an unduly difficult one to comprehend. The scientists themselves have accepted this confusion and continue to assign a ‘u’ at the end of CKD to denote the unknown nature of the etiology of the disease.

While focusing on environmental and other factors such as water pollution, the use of fertilizers and pesticides in agriculture, snake bites, the use of aluminum utensils, and heredity in their analysis of CKDu, the researchers have forgotten the human face of the disease. They often saw the persons affected by such environmental nephrotoxins as patients – CKDu patients in their 4th and 5th stages. Often, they merely analyzed urine and blood samples, and carried out kidney autopsies. The most important aspect of these patients – their personhood, their humanness, their social relationships and culture – were completely forgotten in carrying out this scientific research. These studies have failed to see who these ‘patients’ were, what activities, and behaviors they engaged in, the social groups they belonged to, the predicament that made them members of certain social groups that were affected by the disease. Why it is that predominantly men in these groups get affected by this disease? Why do many women in these groups not progress towards end-stage CKDu? In order to understand how CKDu is structured in agricultural societies in South Asia and Mesoamerica, it is important that we provide answers to these behavioral questions.

Although scientists vary in their opinions on the etiology of the CKDu, they all agree that the occurrence of the disease is a result of the interface between humans and their living and working environment. This interface is often seen in people who depend on the environment for their livelihoods. The link between environment and livelihoods can be understood only if we study how people make use of the environment. In other words, what are the behaviors that people are engaged in, in order to exploit the environment for gain?  In the areas where CKDu is prevalent, what distinguishes those who are affected from those who are not?

Following well-structured scientific procedures with high technology, rigorous methodologies and laboratory experiments aimed at isolating etiologies is pertinent if CKDu is found in all segments of society. However, the story of CKDu in Sri Lanka as well as other affected areas in the world is that it is found in pockets or ‘islands’. This shows that CKDu occurs in specific social settings.  The simple question we have failed to ask so far is why this disease occurs in certain social groups and not in others. The scientists have never asked who these CKDu patients are; where they come from; what they do for a living; and whether those affected share common social and cultural characteristics. If we were able to answer these simple questions, we would also be able to address many of the challenging issues surrounding CKDu.  So far, these questions remain unanswered.  

The article published in BMC Nephrology by Jayatilake et al in 2013 (Chronic kidney disease of uncertain aetiology: prevalence and causative factors in a developing country) is the published form of the widely discussed WHO study on CKDu in Sri Lanka. Although involved a huge investment in research, this investigation did not provide a comprehensive understanding of CKDu, partly due to its inability to answer these sociocultural questions and the failure to comprehend the behavioral component of the disease. It is important to consider why it failed to incorporate the sociocultural and behavioral components into the analysis.

A cross-sectional study: Although Jayatilake’s article is based on a cross-sectional study that randomly selected 100 households from twenty-two villages in three districts, it failed to discuss how the study brings into focus the social and cultural makeup of the village communities and agricultural resettlements in the affected districts. No description is given of the type of villages chosen, or how they represent different socio-cultural zones of the affected districts. An analysis of community types, in both case and control groups, is necessary, as CKDu is socially and culturally situated in certain segments of the affected districts. One could argue that, since ‘patients’ were randomly selected from the affected districts, they would inevitably include people from all walks of society. But the problem is that the sample was not broken down into socio-cultural groups and, therefore, the most important variables of such groups were not included in the analysis.

Prevalence: The overall prevalence of CKDu was 15.3% with a higher prevalence in females (16.8%) than males (13.3%) (p < 0.05). More severe grades of CKDu were seen more frequently in males (grade 3: males vs. females =19.9 vs. 5.3%, grade 4: males vs. females = 16.1 vs. 3.8%).  In both sexes, prevalence was higher with increasing age (p < 0.05). This shows that less severe grades were more prevalent in females. If so, with age increase will there be more females with severe stages of CKDu? The article states that male sex has been reported to be a risk factor for progression to end-stage renal disease. This is interesting but the article does not elaborate on this very important behavioral factor associated with male farmers. If this aspect were dealt with, it would have brought in important socio-cultural factors associated with CKDu.

The magnitude of the affected population as discussed in this publication becomes questionable. For example, based on 15.1% prevalence rate, how do we square the figure of 129,291 people affected in Anuradhapura alone (the prevalence rate of CKDu in Anuradhapura District as mentioned in the article was 15.1%), as implied by the WHO study, with the much lower numbers reported for hospital attendance? Are we really to believe that, in a country such as Sri Lanka where seeking medical treatment is a well-established practice, so many people affected by the disease refrain from going to hospital? This would suggest that, taking Anuradhapura alone with a total population of 856,232 (Census and Statistics 2011), the CKDu positive population increased from around 21,000 in 2011 (based on data provided by Chandrajith (2011) where prevalence of 18 years and over was 2-3%) to over 129,291 in 2012, a 516% increase over one year.

The WHO data indicate that the prevalence of the disease is higher among females, but this is not corroborated by any other research carried out so far. Demographic information collected by Jayasekara et al (2013) from Northern Region clinics, covering 11,323 patients, shows that the male: female ratio among those suffering from the disease was 2.4:1. Other published research also shows that males form the majority of the affected population. A review of published findings from El Salvador, Nicaragua, Costa Rica, Sri Lanka, Egypt and India (Almaguer et al 2014) shows that prevalence was generally higher in male farmworkers aged 20–50 years.

The samples were drawn from three locations for the cross sectional study. They were the CKDu confirmed patients from CKDu affected locations, non CKDu people from the affected locations and a third sample from Hambantota, which is a non CKDu affected area. However, the research design did not allow a sample of CKDu patients from a non-affected area. The 4^th category is important as the disease is now spreading into newer locations beyond areas that are already identified as high incidence locations.

CKDu and Occupation: When separate logistic regressions were run for each potential exposure, only occupation type (being a farmer increased odds by 19.5%) and type of agriculture (paddy cultivation compared to cultivation of vegetables and other crops decreased odds by 26.8%) were shown as significant. Most important to see how means of production of people in these farming communities are related to CKDu. In an agrarian society such as rural Sri Lanka, the means of production are the soil (land) and the plough (labor). The fact that whether these farmers own land, if so how much or whether they were owner farmers or they were wage laborers was not mentioned. However, the fact that those who cultivated chena land (highland for vegetables – slash and burn cultivation -- being a chena cultivation farmer increased the OR by 19.5%, while engaging in paddy cultivation compared to cultivation of vegetables and other crops [chena cultivation] decreased the OR by 26.8%) had the higher rate of CKDu. This shows that the disease was higher in percentages among poor slash and burn farmers who do not own or own only a little paddy land and were poor laborers. This is a very important aspect, and if analyzed would have given much needed socio-cultural information on the social standing of the affected people (more than just patients) and their socio-cultural location in the communities.

The article concludes that ‘the risk was reduced in individuals who were male or who engaged in paddy cultivation (OR = 0.745, 95% CI = 0.562 to 0.988 and OR = 0.732, 95% CI = 0.542 to 0.988 respectively, P < 0.05). This is misleading for the very fact that paddy cultivation is regarded as one simple variable in the analysis. Paddy agriculture is a complex variable, which has many different forms, which cuts across land ownership, type of tenure and wage labor. Paddy agriculture constitutes many types such as owner farmers, wage laborers, tenants, lessees, partnership caretakers and mortgagers etc. If the variable paddy agriculture were fragmented into simple variables of different types of paddy agricultural work, picture would have been different and more realistic. However, a majority of studies show that there is a link between paddy agriculture and CKDu. Most critical statements made in the article says ‘…In the present study, older age, being female and being a chena cultivation farmer increased the risk of CKDu’ (p9). In simple language, it says that the females who are older and engage in chena cultivation display a high risk for CKDu. It would have been good if age related odd ratios were provided in Table 2 to better understand the argument. In Table 2, none of the OR were significant with years of engagement in agriculture is interesting, and one would therefore, tend question the validity of the above statement that being an older age female chena farmer is a risk for CKDu.  Table 1 on page 3 says the mean Age (years) of CKDu patients was 39.1 (SD14.2). In 2013 Jayasekara et al, in a study of 796 and 67 patients from Padaviya and Nikawewa respectively, conclude that the mean age of CKDu patients was 54.7 years. Two years after studying three cohorts of population in the North Central Province, Jayasekara et al (2015) were able to establish reliability of the mean age statistic of CKDu patients as 54.7 years. It is problematic for the reader to understand why the WHO study deviated from other studies with respect to the mean age of CKDu patients and the identification of gender of the majority of patients.

Ground water and CKDu: The article says ‘CKDu occurs in areas where groundwater is the main source of drinking water’, but failed to account for this factor. The only thing that they say is that it may be because of high concentration of fluoride in ground water. They borrow findings from other researchers to support their hypothesis, but their study did not invest in testing out this hypothesis. This explanation is a camouflage used by the article to hide their inability to prove cadmium availability in excess of the recommended amounts in well water. The faults in sampling of wells for water testing has created this situation in this study. The study by Jayasumana et al (2015) examined water quality of abandoned old wells in the endemic areas and found that they had a very high cadmium content in water.

The lessons we learn from this critique of the WHO study are many. First, it is important to explore the social and cultural characteristics of the affected population. Especially when a disease entity is affecting a particular region and a social segment of the society one of the first steps is to find out who these people (as opposed to patients) are, why they are prone to this particular disease. Second, we must recognize the importance of both social and cultural dimensions in delineating etiologies, and, therefore, research of this nature must be multi-disciplinary. There should be complete exchange of data and findings among different researchers from wide array of disciplines. Assigning a sociocultural component of the study to an isolated group, as in the case of the WHO study, would not bring the socio-cultural aspects of the disease into central focus.  The article we reviewed showed that the sociological study of the WHO project did not lead to deep understanding in the final published document of the project, apparently because of absence of exchange of fundamental ideas and data between the groups. Finally, pursuing etiologies prior to a sophisticated social and cultural analysis (or risk factor analysis as epidemiologists say) of the regional problem is premature in disease entities such as the CKDu which are regionally and socially confined diseases. In order to better serve the purpose of evidenced based policy and programming in diseases that are localized in certain communities, it is necessary that the researchers should have a better informed community focus in their research.