The cross-sectional sample and data collection in Laya is successfully completed (despite additional challenges)!

We successfully completed the cross-sectional sample and data collection in Laya! We collected a total of 227 stool samples and conducted an equal amount of related dietary recalls and integrated surveys between 25 March and 5 May 2025. As part of this number, we also have the first-time sampling of 49 longitudinal participants, who we hope to follow at 9 more time-points within a year, to study the seasonal variation in nutrient intake and the gut microbiome. The longitudinal part is expected to be both the most challenging and most groundbreaking aspect of the EATWELL project. Our ambition was to have 50 longitudinal participants and a total of 250 cross-sectional samples and surveys, so even though we have not reached these numbers, we still have achieved a very high success rate for any microbiome study, and this despite several additional challenges, but thanks to all the support from our project partners, local officials, and the community members we should be able to generate solid analyses from these samples and data.

Once settled down after our journey to Laya (which was a whole logistical undertaking as described in the previous Laya blog), we immediately started assembling the stool sampling kits for the longitudinal and cross-sectional study, continued with the training of our research assistants (RA’s), and held community meetings in each of the chiwogs (villages) of Laya for recruiting participants. Meanwhile we had to attend to some additional technical and logistical hiccups, which delayed the start of the collection of stool samples and the related 24-hours dietary recalls and integrated nutrition/microbiome surveys by two weeks after our arrival.

The organization of the community meetings incurred some minor delays as many of the semi-nomadic villagers had opted to stay in Gasa on the way back to Laya from their winter migration. This was in part due to the earlier extensive roadblock between Gasa and Taktsi Makang in the beginning of March due to heavy snowfall. Some, therefore, opted to stay in Gasa to attend the upcoming religious events, such as the Gasa Tshechu and Moenlam Chenmo. Likewise, most of the Tshogpa’s or village heads hadn’t returned yet and normally they would organize such community meetings. Hence, with their kind assistance (from a distance) and with the help of our host, we managed to organize smaller community meetings for recruitment, as our time was very limited.

As we continued to train our research assistants (RA’s) in the cover story of EATWELL, the informed consent forms, the 24-hours recalls, the stool sample protocols and the integrated nutrition/microbiome survey, taking nearly two months of nearly full-time training, we also gave some of the RA’s the possibility to conduct the meetings. Dr Neyzang Wangmo (KGUMSB) completed the details and, with her known enthusiastic style, motivated people to join as a participant.

Given the delayed start and with some already preparing for the summer migration to the higher summer herder camps, our window of opportunity for conducting the cross-sectional part of our study was becoming even shorter than what would already have been the case. It is indeed not evident to conduct a cross-sectional study in a highly mobile community. Yet, despite these challenges, recruitment has gone relatively well, thanks to the support of the Thrizin of Gasa and Tshogpa’s as well as our host.

The recruitment also included the village of Lungo and we took about a three-hours hike to reach the community meeting immediately upon our arrival in the early evening. Thereafter, we stayed with the Tshogpa to discuss the practicalities of the research there. From this visit it became clear that there would be additional logistical challenges to solve here before we could start handling the stool samples and maintaining the cold chain. His help has proved to be very important later on. Next morning we hiked back to our mainstay in Laya.

Yet, before we could start with the actual sample and data collection, we had to solve some unexpected technical challenges with the -80 freezer and the generator in Laya, which contributed to the additional delays.

Upon arrival, we had first dropped the freezer at the hospital in Laya, which is about a 15-min walk from where the research assistants (RA’s) and I stayed. Yet, encountering several power cuts, it soon was clear that we had to organize logistics in a different way. The distance of the hospital to our place was inconvenient for turning on the freezer or generator regularly after each power cut, but also for going there to collect the cooling element and bag each time before each sample collection and then returning there to deposit it in the freezer. Also, we would not notice a brief power cut when sleeping or when out in the field, so this would mean that we would anyhow have to regularly check the freezer since it would not switch on automatically. Such regular checks then would be needed to avoid any serious temperature rise, which could lead to overgrowth of bacteria and hamper the stool sample analyses. After trying to find automated solutions for switching on the generator, it became clear that getting the freezer to work again would always require human intervention. Hence, we had to find other solutions and therefore needed to first conduct tests to see how fast the temperature would rise until it reached such detrimental temperature levels. We tested it when the freezer was empty and noticed the temperature rose very fast, which meant that in practice we would have to check the freezer every 2 hours and this 24/7. This would especially be problematic for the longitudinal part where we collect samples wherever our participants are, and they may be several day-hikes away, so there would be many times no-one to look after the freezer.

To make all this somewhat more manageable, we already decided to move and store the freezer and generator in an old neighboring house to where the RA’s stay, so that we could intervene rapidly in case of a power cut. Yet, this would not solve the problem of having to check every 2 hours whether electricity was there. We discussed with Dr. Raul from the Raes Lab at the VIB-KULeuven on how we could delay the warming up of the freezer in case of an unnoticed power cut (especially in the night) and he suggested filling the freezer with bottles of saltwater and cooling elements which take a longer time to thaw. We decided to test this out before starting the actual sample collection.

However, when we filled the freezer, we soon noticed that it took a long time for the freezer to cool down and that it was not able to reach -80 as per our planned protocol. After a round of trouble shooting, we concluded that the electricity supply was not strong enough and so we had to hire an electrician to increase the voltage so that the freezer could reach the required temperature, but it did not work sufficiently well. Hence, we had to hire the electrician again to further increase voltage and the freezer then reached the desired -80 degrees, but it still took quite some time.

Then we tested how long the freezer would take to warm up, when disconnected from power. Even in the cold environment of Laya, we saw that the temperature rose rapidly. We tried another test over a longer time and then saw that after a night off, the temperature rose to -30 degrees.

Hence, after all the electricity work and tests, we had no other option but to adjust our protocol and make it a little less strict regarding the temperature maintenance. This was then compensated by programming the data loggers, which record the temperature at given intervals, such that they record the values at short intervals, so that any brief temperature rise can be considered and compensated for in the later biostatistical analysis. Anyhow, the temperature had to be kept as low as possible, but there are clearly limits to what it is possible to do in these situations. Given our tests we were quite confident that we could avoid reaching important thresholds for the potential overgrowth of certain microbes. With the adapted protocol, which seemed and proved to be feasible for the cross-sectional part of the study, we could finally start from 7 April.

Since we initially had more volunteers than we could involve in our study, I selected based on gender, age and main activity, and assigned unique codes to the participants. The 10 RA’s were each assigned their participants, and they rechecked their willingness to participate, after which they made appointments to explain the informed consent form as it contains aspects of complex European regulations which do not really make sense to ordinary people in the high Himalayas. Being fully aware of this complexity, we spent two full weeks training our RA’s in how to explain it in an accessible way before obtaining the signed or fingerprinted consent. The explanation and obtaining of the informed consent had to be done separately from the actual sample and data collection as it takes time. In addition, the use of the sample collection kit had to be demonstrated and given to the participant before the stool could be collected and the integrated nutrition/microbiome survey could be conducted. As anticipated, it would become a challenge to ‘plan’ such sample and data collection at a time that suited both participant and RA, as continuously shifting plans mark everyday life in Laya, and besides, bowel movements can of course not be well-planned in advance.

As we first started the sample and data collection in the four clustered villages in Laya, we also prepared the logistics in Lungo, arranging lodging, buying an additional regular freezer and chiller box in Thimphu, getting it transported to Lungo, and arranging for someone to carry the samples from Lungo to Laya daily. In Lungo, we only would store the samples in the regular freezer as a back-up for samples collected late in the evening (given the unpredictability of bowel movements of participants) until they would be transported to the -80 freezer in Laya during the day. Once our host returned to Lungo from the religious events in Gasa, our RA’s could leave by 17 April to start work there.

While collecting informed consent forms, it gradually became clear that nearly one third of the telephone numbers given were wrong or were not contactable for weeks. Additionally, some who expressed their interest were not available after all or rejected their participation. This meant we were eventually getting short of the desired target of 250 people for both cross-sectional and longitudinal participants, and we had to make additional recruitment efforts, but as the cordyceps collection was about to start early May, we could not extend our collection period as then everybody would be collecting cordyceps and many others would start going to the summer herding camps with the yaks. We addressed additional potential participants as we met them in the different public spaces and at other meetings. In this way, we could increase participation to the result of having 227 participants after all.

Normally, people would have to call immediately to the RA’s after they have collected their stool in the kits, and then the RA’s should leave immediately to collect the cooling element from the freezer, put it in the blue cooling bag, and then find the location of the respondent, which hasn’t always been easy. Then they would collect the samples, put them in the blue bag with the cooling element, and then conduct the 24-hour dietary recall and the survey, after which they would bring back the sample to the -80 freezer. The collection of samples and the execution of the surveys needed to be done simultaneously as we seek to correlate the microbiome with food and diet as well as some other activities that may be affecting the gut microbiome. As participants had to call the RA’s when they were ready, the procedure sometimes led to overlaps when respondents called around the same time and then another RA had to take over.

As the samples were dropped in the -80 freezer, we soon noticed that its space was filling up rapidly. Based on the amount of samples collected I correctly estimated that we would have to take four trips to Thimphu—arranging each time porters to carry the -80 freezers down to place it in a car with 12 V connection and minimum 20 Ampère to keep the freezer running during the 8-hours journey—to deliver the samples at the Royal Centers for Disease Control in Thimphu to return the freezer asap to Laya the next day. As such, we would lose a total of 8 days in sample collection within a tight timeframe of three weeks, and so we had to find a solution to keep the sample collection going. After going down with the samples in the freezer carried on the back of a strong person or porter, we immediately hired him to carry our freezer, which was initially destined for Lungo, to Laya instead. This freezer could be used for storing samples while the -80 freezer was transported. That same day, we bought another freezer for Lungo after arriving in Thimphu and which another hired porter brought up to Lungo the next day. These 2-day journeys were intense each time, as we would have to do additional shopping after a total of 10 hours travelling, for instance to fill the food supply for our team in Laya. Upon return, this meant hiring a few horses to carry the supply up in the evening.

Indeed, the organization of food supply and the preparation of meals and the washing of the dishes was another challenging logistical element that is central in making the cross-sectional study successful. I write about this here as there would be no space in scientific journals to write about such practical aspects, but which may affect scientific results. Coming with 13 people to our host, we could not expect them to cook for us three times a day for nearly six weeks. Hence, we brought our own food supply and made turns for who should cook and do the dishes when. Cooking rice three times a day has taken considerable time from the RA’s and sometimes clashed with unexpected calls from participants having their sample ready, which then required immediate adaptation so that the respective RA could go and do their job. At times when all RA’s were busy, this meant big delays or even skipping meals, so that the samples would not be lost (as they ideally needed to be in the -80 freezer within 2 hours after exiting the body). A cook would have come in handy, but would have been quite expensive, and the budget of EATWELL is too limited for such additional expenses. And we had already too many unforeseen additional expenses given the logistical challenges and which then transformed also into budgetary hiccups.

Towards the last week of our work in Laya, we started facing additional challenges in our recruitment of participants as people started preparing to leave for the cordyceps collection and for the summer herding camps. Some people who agreed to participate were no longer contactable or available and so, we decided to make extra efforts in recruiting additional participants. This helped and the last days became very intense with many recruitment efforts, informed consent meetings, and sample and data collection often clashing with meal preparations and thus creating delays in having our meals. At times, we received help from the host family, which has been overall very helpful in making this part work.

In that same week, we also noticed that some of the tubes for the longitudinal sample collection had started leaking. This was likely due to the altitude, as we were at around 3850 metres, whereby the pressure internally outweighed the pressure in the environment, thereby probably causing the tubes prepared at sea level in Belgium to leak in Laya. We checked all the tubes before using them in the last push to recruit more people and collect their samples along with the recalls and surveys. Several RA’s made extraordinary efforts which contributed considerably to our final success, and this in shifting weather circumstances.

Just 2 days before leaving, we were also lucky to witness the Bonkor ritual, which only takes place every 2 years. It is a multi-layered event with many aspects to it, which requires more analysis, but overall, it is said to be conducted for the well-being of the entire community and for the protection of the harvest. It is also a great food sharing event where meals are offered and eaten in each of the villages (except Lungo due to its distance but there is an offering to its direction) tying the entire community together. I plan to write more on this event elsewhere as I have observed it two times now. It was a new experience for the RA’s who also got some opportunity to socialize with some of the participants.

As our departure date came close, the preparation for our return started, and required us again to book porters for carrying down the last batch of the samples in the -80 freezer and the freezer in Lungo, 15 horses, a pick-up truck for the freezer transport, 2 additional bolero pick-up trucks to transport the RA’s and their luggage from Taktsi Makang to Gasa, and a bus to transport these same RA’s and their stuff from Gasa to Thimphu. Hence, the journey was the other way around as described in the previous blog, but a little less extreme.

When the final daily reporting was finished the evening before leaving, I could share with the RA’s the good news of the successful completion of the first cross-sectional part and the first sample collection of the longitudinal part in Laya. We had a great final meal before leaving, but we could not get complacent yet. The next morning, the power supply was weak, and this had caused the -80 freezer to switch off, which meant we had to cool it down by using the generator before carrying it to Taktsi Makang to avoid any further temperature rise. When the temperature reached -80 and the horses were fed, we said temporarily goodbye to our hosts (at least for the longitudinal RA’s) and went down with the -80 freezer and the samples. After arranging the different bills, we could finally leave Taktsi Makang, after which the RA’s in the 2 bolero pick up trucks could transfer to their bus in Gasa. We only arrived in the evening in Thimphu after the delays and this meant we could not deliver the samples in time to RCDC, which was done the morning after.

It indeed feels satisfying that, despite all challenges, we achieved this important milestone in the project, and this thanks to the efforts of so many people. Now, the work has started for the preparations for the cross-sectional sample and data collection in Buli, Zhemgang, and then among those who have migrated from Laya and Buli to urban food environments to discern long-term evolutions in the nutrient intake and gut microbiome. More about this in following blogs…