Much of Sri Lanka’s Dry Zone is currently grappling with a drought caused by the delayed Monsoon. This is a double whammy for residents in several districts who have been engulfed by another ‘slow emergency’ for two decades: mass scale kidney failure affecting large numbers.

Diabetes or high blood pressure can lead to kidney failure. But beginning in the 1990s, thousands of people in the North Central Province (NCP) developed the condition without having either factor. Most were male farmers.

This puzzled doctors and other researchers who struggled to understand how and why. It was soon assigned an official name: Chronic Kidney Disease of unknown etiology (abbreviated as CKDu).

CKDu has evolved into a humanitarian tragedy on a mass scale. It is claiming more victims every passing year, spreading to more areas, and gradually overwhelming the healthcare system. Its causes are still unclear and hotly debated.

Anuradhapura and Polonnaruwa Districts are ‘Ground Zero’ of this mysterious ailment for which there is no known cure. In fact, it was first detected in the early 1990s by Dr Tilak Abeyeskera, Consultant Nephrologist attached at the Anuradhapura general hospital.

It has since spread to North Western, Uva, Eastern, Central and Northern provinces as well. The affected areas are now spread across approximately 17,000 sq km (a quarter of the island), which is home to around 2.5 million people.

Several thousand have already died; the exact number is not clear due to gaps in statistics and incomplete reporting. Hospital records show that over 15,000 people are kept alive with regular kidney dialysis.

Those who develop CKDu are mostly men between 30 and 60 years, working as paddy farmers or farm labourers. These factors are being investigated by health officials and many scientists searching for ways to contain and treat the ailment.

The latest attempt to make sense of the deepening mystery is by India’s Centre for Science and Environment (CSE), an independent research and advocacy organisation. They collaborated with the Ministry of Water Supply and Drainage, and the non-profit Centre for Environmental Justice (CEJ) in Colombo.

CSE’s scientists tested samples of drinking water, soil, rice plant and grain, pesticides and chemical fertilizer from many locations. Their study report, released in Anuradhapura on 16 August 2012, does not pinpoint one definitive factor or solve the jigsaw completely. Yet it clarifies some issues and takes the debate forward.

What causes it?

Helpfully, the CSE report collated many theories and speculations on possible or probable causes of CKDu.

Listed in no particular order, these include: excessive cadmium in the natural environment; high levels of fluoride in drinking water; using fluoride-rich water in low quality aluminium pots; “hard water” with higher than normal levels of minerals; and toxins generated by blue-green bacteria in the water.

Humans are exposed to multiple elements over time. Finding one rogue element is never easy. So factors such as illicit liquor and Ayurvedic medicine are also being studied. Meanwhile, a team at Peradeniya University is probing whether people in affected areas have a genetic predisposition.

In 2011, some researchers from Kelaniya and Rajarata universities argued that arsenic in pesticides and chemical fertilizers, when combined with calcium in hard water, causes CKDu. Their findings became controversial when one researcher claimed to have derived key insights through ‘divine intervention’.

So is it Nature, nurture, or a combination of the two, causes this misery? Nobody knows – as yet.

Interestingly, all the areas with high incidence of CKDu are located around reservoirs of the irrigation systems. In contrast, those communities who draw their drinking water from natural springs have few kidney failures among them. Is it something in the irrigated water – and if so, what?

Planet Earth is capable of springing some nasty surprises on us, even without any human provocation. As earth scientists remind us, this is an inevitable part of living on a restless planet.

In 2009, the World Health Organisation (WHO, a specialised agency of the UN system) and the Health Ministry’s epidemiological unit appointed 10 study groups to study this problem. Their findings have been submitted to the government but not yet released. Why?

We cannot afford bureaucratic apathy in a matter of such urgency and importance. The outcome of public science must be shared with the public and media in the public interest.

Beware of opportunists!

Delays in releasing research and analysis will only allow speculation and conspiracy theories to gain momentum. Selfish opportunists are already flocking to CKDu hit areas apparently seeking to implicate their pet hates. Sadly, some of these speculations are being peddled – and even cheered — by sections of our media without due diligence.

Such tilting at windmills is muddying the already suspect waters and can confuse policy makers. Senior scientists like Prof Oliver Ileperuma and Prof C B Dissanayake – at the forefront in related research — have stressed the need to separate facts from speculation and myths.

Another fervent plea for sanity appeared recently in the respected Ceylon Medical Journal (CMJ). Established in 1887, CMJ is published by the Sri Lanka Medical Association (SLMA), the national professional body of doctors. Writing in the December 2011 issue, three medical researchers — A R Wickremasinghe, R J Peiris-John and K P Wanigasuriya – called for dispassionate discussion of current knowledge and gaps.

Given the widespread discussion and debate in the media recently, they urged, “it is timely that the available, credible, scientific evidence on CKDu (published in peer reviewed journals) is collated and analysed, and the difficulties faced in establishing causality are discussed.” (Emphasis mine)

The authors added: “The cause of CKDu is likely to be multifactorial. At this point in time there is insufficient evidence to pinpoint a cause(s). Both the wellbeing of residents of the NCP and the enormous drain on health system resources and the economy demand that resolving the issue is a national priority.”

Medical doctors are on the frontline in treating affected people and counselling devastated families. But CKDu is much more than a mere medical or health emergency. Interdisciplinary studies are needed – involving both natural and social scientists – and with adequate coverage, intensity and scientific rigour.

Miracles and Sacred Cows

Scientific credibility requires that such studies are peer reviewed and published in national and international journals of high standing. Making unfounded claims at press conferences and throwing wild allegations in TV talk shows might create some ripples, but such grandstanding doesn’t help anyone.

The National Science Foundation (NSF), our main research funder and agenda setter, can provide directions. We also need NSF to ensure the findings are widely discussed in policy, professional and public forums. Secrecy is not an option.

There is no room for miracles or absolute truths in science. By its very definition, science is open to rigorous scrutiny, challenge and refinement. CKDu is no exception.

Last year’s ‘divine revelation’ story reminded me of a famous creation by Sidney Harris, the doyen of science cartoonists. Science demands its practitioners to be more explicit and accountable than resort to ‘miracles’. No one can claim to be a sacred cow.

There are also no shortcuts in science; the scientific method cannot be circumvented or accelerated at some officials’ whim or fancy. However, urgent societal problems like CKDu warrant fast-tracked study and action. These need more funds and brains.

But remember: haste makes waste. The CMJ paper called for vigilance that mere associations “should not be considered to be of causal importance without documented evidence of proof”.

The CMJ authors hit the nail on its head when they wrote: “We advocate caution by the scientific community and the media when using such assertions that impact on policy, the livelihoods of the farming community and the economy of the country without scientific validity and biological plausibility.”

Media’s Challenge

Science and media are inherently contested public spaces. They thrive only with open, informed and inclusive debate.

We in the media face many challenges in covering this complex and nuanced story. CKDu has all the elements that typically interest the media: rising death toll, widespread human suffering, as yet unclear origins and causes, scientific arguments, and — since of late – some conspiracy theories.

During its early years, CKDu was under-reported by our urban-centred media. Well, no longer. We now face the real danger of some media outlets engaging in fleeting, superficial and alarmist reporting.

CKDu is not a straightforward or simple story to report. It’s not like a tsunami or flood. It’s even slower than a drought (a gradual disaster that many journalists struggle to grasp), and not an infectious disease.

The impact is on human beings but no visible changes in the landscape.

CEJ’s Executive Director and chief scientist Hemantha Withanage has been travelling in CKDu affected areas, meeting hundreds of families and community workers. In some areas, every third house has lost at least one family member. Survivors are migrating to cities.

He relates the eerie and heart-wrenching scenario that repeats in thousands of households where men are living from one dialysis session to another, ideally every four days. Each session costs around LKR 12,000 (USD 92 approx) to the healthcare system, and many hospitals don’t have enough dialysis machines.

Secretary to the Ministry of Health told a media seminar in early 2011 that LKR 350 million (approx. USD 2.6 million) — 4.6% of the health budget – is spent every year on supporting CKDu patients.

With thousands affected, it is only natural for emotions like despair, suspicion and anger run high. Scientists and health officials need to be sensitive to such sentiments, without allowing scare-mongers and villain-hunters to hijack policies or research agendas.

CKDu is as much a national health emergency as dengue epidemic in our cities. To find relief for those already affected and protect millions more at risk, we must harness our nation’s brain power – giving them clear focus, adequate resources and academic freedom.

Never forget, however, that this isn’t just a research problem. Fellow human beings – not lab rats – are involved. They need interim help while long term solutions are being worked out.

Both CEJ and CSE have called for better medical facilities as well as clean drinking water to reduce people’s dependence on poor quality groundwater. These are development needs in their own right.

While scientists probe causes and solutions in earnest, the government and voluntary organisations can immediately start providing relief and livelihood support to those living with CKDu.

Finger-pointing, document-leaking and conspiracy-spinning might generate cheap thrills or media coverage for some with their own agendas. But these theatricals don’t help anyone hit by, or dealing with, the formidable problem.

Let’s not allow tragedy to be turned into farce. There are indeed better ways to care for the traumatised farmers and their families.


Science writer Nalaka Gunawardene asks more questions than he can answer, and writes weekly columns in Ravaya and Ceylon Today newspapers. All views are his own. A shorter version of this essay appeared in Ceylon Today of 19 August 2012. He blogs at

  • Ranil Senanayake

    Dear Nalaka,

    This WHO report reached me yesterday , perhaps it could help

    World Health Organization (WHO) report on Chronic Kidney Disease of unknownetiology (CKDu) in Sri Lanka- a summary :

    03. Progress report 13 Feb 2012

    In the urine analysis of 496 cases of CKDu 56% had a urine cadmium excretion over 1 ug/g creratinine. Data from recent studies show that changes of early kidney damage occurs at cadmium excretion levels of even 0.6-1.0 ug/g creratinine. About 63% of CKDu patients had urine arsenic levels above 21 ug/g Ceratinine. Urine arsenic levels above 21 ug/g creatinine have been shown to cause changes in kidney tissue that lead to chronic kidney disease.

    Approximately 88% of CKDu patients had urine arsenic >21 ug/g and/or urine cadmium >0.6 ug/g

    Arsenic was also analyzed in hair and nails of people living in NCP including CKDu patients. In about 90% arsenic levels in hair were higher than those observed in developed countries (0.02 ug/g).In about 94% arsenic levels in nails were higher than those observed in developed countries (>0.03 ug/g)

    Our analysis did not find high levels of cadmium exposure as reported in previous studies. The mean exposure for adults is at borderline of Recommended Total Weekly intake (TWI) of 2.5ug/kg body weight.

    Water from 98 water sources used by patients with CKDu was analyzed for hardness. 99% are hard to very hard. Hardness of water is known to affect heavy metal toxicity through antagonistic mechanisms and this may play a role in renal toxicity caused by heavy metals in the north central province.

    • kadphises

      “Arsenic was also analyzed in hair and nails of people living in NCP including CKDu patients. In about 90% arsenic levels in hair were higher than those observed in developed countries (0.02 ug/g).In about 94% arsenic levels in nails were higher than those observed in developed countries (>0.03 ug/g) ”

      I cant understand why the control population are citizens of developed countries. Shouldnt we be comparing those afflicted people from the NCP with those unaflicted people from the NCP using the same water sources? And also those afflicted with others in Sri Lanka where there is no CKDu?

      We should also compare symptoms with Bangladesh where groundwater Arsenic is a much bigger problem. As far as I know the groundwater Arsenic in Bangladesh results in cancer among the consumers and not Kidney Failure.

  • Janitha Liyanage

    CSE report is incomplete.They have just checked 3 samples of fertilizers and 5 samples of pesticides.Sample size is not enough even for AL students project.

  • Good poser, Nalaka.
    There definitely are many Qs than As as traversed by NG. With unrelated interventions from many interested parties, going on their own with unknown agendas, Qs wld only multiply. “Arsenic” in CKDu is one such unhealthy campaign, rather than an effort to unravel the unknown in this tragedy. My inquiries over the recent WHO tagged study made me understand that it has many pot holes to make it an acceptable scientific study. It had been presented to an invitee audience and one issue raised had been why no comparison was made with districts like Kalutara, Galle or even Hambantota where fertilizer use is minimal and where CKDu is not been reported as yet. Another science academic who was at this presentation told me, Arsenic wld have fatal effects on the Neuro system before it wld on the kidney. So far, complaints on Neuro related ailments have not been recorded in CKDu patients he said.
    Thus, there certainly are issues that need rational approach in finding answers than politico emotional berating.

  • gnewy

    As the author clearly points out CKDu could be causes of multiple factors ranging from chemicals in water the water source, illicit farming practices, so on and so forth. There could be other factors as well. Possibly, unregulated illicit medicine that is marketed in the country, that may be sold in those parts of the country where it’s easier to get away with such illicit activities. The author also rightly emphasis it is not a medical procedure that can find the solution but requires evidence from multiple sources.

    In my honest opinion, it is data that can provide most of the answers. This is the key element that is missing in the health sector. Our epidemiologist are comfortable spending time and money to simply report on a hand full of 25 infectious disease, called notifiable diseases. These diseases are seldom; no more than 66,000 patients per annum. The last (2007) Ministry of Health Annual Health Bulletin reports that their are over 48 million patient visitations at the Government hospitals. This means the Epidemiology Unit is only reporting monitoring less than 0.1% (zero point one percent) of the reported cases. Those are simply district aggregates and not granular with time and specific location information.

    How these numbers are achieved is questionable but the Medical Statistics Unit confirms that these numbers are obtained through paper registries, where nurses are asked to supply the numbers for the 660+ hospitals each year but there has not been any numbers since 2007. Absence of this rich set of data in digital form is unfortunate.

    Syndromic surveillance has proven to be effective in the early detection of potential outbreaks. Instead of waiting for the disease to be confirmed, which may take several days with laboratory tests, etc, a recording of the symptoms the patient complaints of at the Outpatient department is of greater utility in nipping escalating diseases in the bud. Moreover, the accumulation of geo-spatial information along with the syndromic data can aid medical statisticians with them drilling in to the data to find alarming patterns.

    If such data had been available then medical statisticians could have analyses the data looking at all different correlations to zone in on a meaningful data sample. Moreover, other data sets such as agriculture information and water supply information can be layered on top of the medical data (which we call multi-variate statistical analyses) to verify the hypotheses.

    This is exactly what the Real-Time Biosurveillance Pilot achieved; here’s a Lakbima news article that tells the story:

    It is a heart stopping struggle trying to scale such a project that has proven it’s capabilities to better serve the health information needs of the country. Not just with identifying infection diseases or reportable diseases but also contributing information on life-style chronic diseases. A key emphasis of this biosurveillance program was enabling syndromic surveillance or at least create a comprehensive rich data set of disease, symptoms, location, date, gender, and age specific clinical information. We don’t need patients’ names and private information. Moreover, the pilot tested biosurveillance program was estimated to be ~40% (forty percent) cheaper with much higher efficiency gains and incremental benefits, relative to the present day paper and laborious practices that have no utility. A national implementation of such a system would cost less than 1.0% (one percent) of the total annual national health expenditure.

    With the pilot data we found alarming trends of Respiratory Tract Infection, Urinary Tract Infection, Hypertension to be twice as more common among above 40 women than men in Kurunegala District. The system was capable of detecting the escalation of any disease with a day opposed to the present day practice that takes weeks or for several deaths reported to raise some eyebrows. All these adveres events affect household productivity and our quality adjusted life years. I suppose absence of such granular information might be better for the institutions because then there is less work and less blame to take.

    The most visible problem, in trying to introduce such innovations with proven utility, is breaking the fear within the institutional silos of the health ministry that like to safeguard their territory. Because they don’t want to share resources and don’t see eye-to-eye, a one stone kill all birds kind of solution that benefits all is hard to pass their acceptance. They all want to do their own thing.

    Doctors (not all) have superiority complexes that they know it all, including ICT. If anyone from outside tries to introduce something then that is not good enough. I recently, had a meeting with set of Epidemiologist, before they could even hear out the utility of the field tested biosurveillance solution they had already judged it as useless, yelling away like hooligans. It is very immature and unscientific behavior that I experienced; especially, coming from a scientific community. The author also points to other unscientific behavior, with the lacking of peer reviewed publications; moreover, sharing of the raw data for other researchers to verify the conclusions.

    Until such time government institutions open up their data, learn to share, drop their inferiority/silo-ish structure, and those responsible institutions are forced to perform based on outcomes, they will simply not solve CKDu or any other problem for that matter and would continue to dwell within their safeguarded cradles, lavishly spending Government money.

  • Citizen

    Since there are standards such as drinking water standards, surface water standards it is not difficult to rectify any situation using facilities availble in our scientific institutions. What is required is a strict regimen of testing and control of all sources of pollution. One has only to refer to US EPA to find out how they are tackling such problems in the US where there is widespread use of chemicals etc.

    What is lacking is a common sense approach to tackle the problem. Our Scientist are engaged in an academic debate whilst the poor farmers wither and die away. We dont have to reinvent the wheel when such know how is available. the problem appears to be the number of govt agencies involved and lack of a cohesive strategy.

  • Harendra de silva

    Just to set the record straight. Dr Thilak Abeysekera is a consultant Physician from Kandy, not a nephrologist fro apura as stated in the report. He did a lot of work in A’pura including clinics to investigate the cause

  • N. Ethirveerasingam

    Establishing a computerised database and sharing with medical researchers is not supported by funds from the Govt especially if the data is of violent deaths. In the North (This may be the case elsewhere too) there were no professionally developed software to keep medical records, up to 2011, which can be used for data based decision making. Though such software is in use in Education at the HQ level. Medical and some govt departments keep data for descriptive purpose in xls for Heads to make ppt presentations.

    Often decisions are made even where reliable and valid database is available without reference to the available data. Gut feelings and biased observations that suits the politicians are what is common.

    It is time the Medical Faculties and Teaching hospitals have well funded professional Research Units.

    As another observer mentioned above there are enough research methodology, statistical packages and competent professional medical researchers available in Sri Lanka to determine and predict the cause of CKD with an acceptable level of probability.

  • Priyantha

    Good synthesis. It is a problem of Sri Lanka. What I believe is all sectors should get together and find solutions rather than accusing each other.

  • Upali Samarajeewa

    Dear Nalaka
    A nicely written article which should make the responsible authorities more active. Most of us, scientists, were waiting long for the WHO report. Unfortunately the document in circulation (leaked !) does not carry a signature, nor it is in an official letterhead. More unfortunate part of the story is that the WHO rep at a meeting in prime ministers office in June 2011 said that WHO has analytical reports of urine from affected persons and they are waiting for results of controls (from NCP and other parts of country I believe). The document in circulation does not carry this information leaving room for each person to come out with his own interpretations rather than examining matter scientifically. We are so unfortunate that even the said WHO report cannot throw real light into the story.

  • Upali Amarasinghe

    There are hair-splitting arguments above. Unfortunately people are still dying of CKDu in the NC province. No long-term solution yet!