In this prospective cohort study investigating AGI among the Norwegian population for a period of 12 months, a total of 9,946 persons participated, for an overall response rate of 11.5%. The cohort participants represented both large and small drinking water supplies, sex (male/female), age, education level and geographical region in Norway.
We found a relatively low number of AGI per 100 person-months (approximately 5), and a very low number of severe AGI per 100 person-months (< 1). The highest number of AGI per 100 person-months was found among those aged 0–5 years (9.5), followed by those aged 19–49 years (7.6). Overall, the results from the adjusted model show very little to no effect on AGI or severe AGI and water consumption (glasses of water consumed). There were no clinically significant associations between the consumption of tap water and AGI or severe AGI in the models adjusted for possible confounders, with the expectation of a small effect of age on AGI. The risk of AGI was higher among small children (0–5 years; 5 percentage points higher risk of AGI compared to those 19–49 years old) and lower among the eldest participants (65–80 years; 4 percentage points less than those 19–49 years old). AGI varied by season, but other possible confounding variables (sex, education level and size of the drinking water supply) were either not statistically or clinically significant. The results are somewhat lower than those of a previous study in Norway (29); however, the study was conducted more than 20 years ago with other methods for data collection. In addition, several precautionary actions in the drinking water sector have been implemented since the studies were conducted, such as enhanced treatment processes, among others, from a publicly funded program; general improvements in best practices; and updates and revisions of regulations in line with the EU Directive on drinking water. On the other hand, the results of the present study are in line with those of a study in the municipality of Ale, Sweden (35). The present study included children for whom the Ale-study did not. Children are more susceptible to gastrointestinal infections, which might explain the small association observed.
As an observational study, causality cannot be drawn, and caution must be used when interpreting the results. In the present study, participants were more likely to be female, older, have higher education, and come from the eastern region compared to the general Norwegian population. These differences were accounted for in the adjusted regression analyses. The study did not include the etiology causing the disease. Although adjusted for confounders, viral infection during the winter and bacterial infection during the summer may have affected the outcome, as the 1-year follow-up of the participants was conducted during different time periods across the seasons. Despite the findings of low numbers of AGI among the participants during the study, based on the characteristics of the cohort and adjustment for confounders, we assume that the external validity is high, meaning that the outcome may be generalizable to the population of Norway. However, we might not have captured the extent to which the patients were exposed to contaminated drinking water. Contamination events could occur hypothetically at any location in the distribution system, at any time, if three key susceptibility conditions must be met for an accidental intrusion to occur in a distribution system: adverse pressure gradient, intrusion pathway, and contaminant source (41).
Recall bias, for example, by the participants’ tendency to overestimate their own positive behaviour in retrospect (e.g., drinking a “high and healthy” amount of water) or interest in the topic being studied (e.g., having a motivation to be a part of the study due to a high frequency of disease), may have affected the results. The duration of follow-up was one month in our study. In the Ale-study, a difference between 14-day and 28-day recalls was observed, where shorter recalls were associated with a 20% greater incidence (35). The study duration was relatively long (12 months), and this, compared to a crisis such as a waterborne outbreak with massive media coverage, may have led the participants to lose interest and leave the study. We were unable to conduct an analysis of the nonresponders. Although mobile phones are a highly common tool among the Norwegian population, because of the easy access to questionnaires via SMS, the response rate was quite low. This has become a common feature among such a data collection method because it is influenced by the massive increase in marked and customer surveys and may also affect the constraints of fulfilling stricter requirements of personal data protection acts.
In this relatively large cohort study undertaken in Norway, we could not detect any clinically significant association between the consumption of drinking water and AGI. The very low number of AGI cases associated with the consumption of drinking water may indicate that efforts to safeguard drinking water in Norway, such as regulations, technical improvements, and publicly funded programs, are effective in providing safe drinking water to the public. It can also be assumed that contamination events, either detected by routine monitoring schemes or critical events such as main breaks or similar events, in the distribution system have led to corrective action by the water supplier, such as issuing boil water advisories to the customers of the affected supply area (42–45). The low incidence of cases underscores the importance of control measures in the drinking water sector, as these measures seem effective. Considering the vulnerability of the drinking water distribution system in Norway, it is imperative to continue investing and maintaining the distribution system to avoid future waterborne outbreaks caused by contamination entering the system after water treatment processes (25).
These results show a very small or no association between water consumption and AGI in Norway between 2018 and 2020. During these years, Norway had a stable and relatively robust water supply system, except for a severe waterborne outbreak linked to a contaminated reservoir after heavy rainfall in 2019, which caused approximately 1,500 cases of Campylobacteriosis (25). However, concerns about the risk of waterborne outbreaks are emerging due to an increase in the hygienic load related to the import of new or re-emerging pathogens from the effects of climate change, people travelling abroad, pressure from the expansion of dwelling areas, and activities near raw water sources (26, 27, 46). With increasing severe weather due to climate change, the quality of water and safe operation of the water supply system in Norway may decrease, limiting the generalizability of these results to the future (47), although documented health effects on waterborne diseases linked to climate change in Norway are scarce (48). However, some of these challenges are common among drinking water supply systems in similar contexts. In a review of waterborne outbreaks in Europe, North America and New Zealand, among 66 identified outbreaks, the causes were the contamination of raw water from surface waters (13/66) and groundwater (11/66), treatment deficiencies in the water treatment plant (18/66) and more than one-third from distribution system failures (26/66) (49). In terms of outbreaks, it is estimated that in North America, drinking water distribution systems could account for approximately 30% of waterborne outbreaks (50). The effects of changing climatic factors are expected to act as stressors to aging and vulnerable drinking water supply systems and health consequences (19, 20). Concern about the ability of small water supply systems to manage a water crisis for effective public health protection is also a concern (22), in which the majority of drinking water supplies in Norway are smaller. Furthermore, other countries may have other challenges, regulations and characteristics related to their water supply systems, limiting the generalizability of these results to other countries.