There has been a huge increase in sales of bottled waters during recent years, especially in Europe and North America, and the market is estimated to be growing at a rate of 25% every year.

Consumers perceive bottled water to be a pure, healthy and natural product, unaffected by pollution and concerns about safety. Tap water, in contrast, is held in much lesser regard, and the lack of confidence that consumers have in domestic water supplies is one of the main driving forces in the switch to bottled drinking water. This attitude may be seen as irrational, but is nonetheless real, and is helping to create a thriving growth area for the beverage industry.

Most consumers would be shocked by the suggestion that their freshly purchased bottle of 'pure water' is home to millions of bacteria, but this is often the case. In fact, the microbiology of bottled water is unlike that of any other food, and is still not fully understood.

There are three main categories of bottled water, natural mineral water, spring water, and other bottled drinking water. The differences between these are usually defined in legislation, and have a significant effect on their microbiology.

Natural mineral water must come from a protected underground source and must be naturally free of pollutants and harmful microorganisms. It must meet specified microbiological standards at source and immediately after bottling, but cannot be treated to attain these standards.

Spring water must also come from an underground source, and must comply with the same standards, but can be treated to meet them. Ozonation and treatment with UV light are two processes used to do this, and carbonation is also a very effective bactericidal treatment.

Other bottled drinking water covers almost everything else, and could simply be tap water. Again microbiological standards must be met, but treatment is allowed. A number of flavoured waters have also come on to the market in recent years, but these are generally considered as soft drinks, and are very different from plain waters in terms of microbiology. Most published research on microorganisms in bottled waters refers to natural mineral water, and what follows reflects this.

Sources of microbial contamination

Even deep aquifers contain indigenous microbial populations, and when water is drawn off from an underground source it will contain microorganisms (autochthanous flora). During the bottling operation, other microorganisms may be introduced as contaminants from equipment and the environment (allochthanous flora). These are the two principal sources of microorganisms in bottled waters, and their characteristics may be quite different.

The autochthanous flora are generally harmless organisms, mainly Gram negative bacteria including species of Pseudomonas, Flavobacterium, Cytophaga, and Moraxella, which normally inhabit water. They are often in a 'starved' state, and can be present as 'ultramicrocells', very small bacterial cells that may pass through filters. This is an adaptation to low nutrient levels. However, inadequately protected sources may sometimes be contaminated by run-off from the surface and then contain other species, including a variety of pathogens, especially if faecal contamination is present.

The allochthanous flora may include a very varied range of species. Organisms such as Pseudomonas, Flavobacterium and Aeromonas spp. may colonise the pipes of the bottling plant, and may build up as biofilms, which act as sources of contamination. Other bacteria, such as staphylococci and coliforms, may enter the water from the general environment or as a result of poor hygiene by personnel. Plastic bottles and caps usually have very low levels of bacteria, but inadequately cleaned, recycled glass bottles may be badly contaminated.

After bottling

In Europe, natural mineral waters should contain not more than 20 bacterial cells per ml at source, not more than 100 per ml 12 hours after bottling, and pathogens should be absent. It should not be a problem to maintain this standard in carbonated water throughout the shelf life, but bacterial populations can change significantly in still bottled waters during storage, by a process that is still not fully understood.

After bottling, the autochthanous flora begin to multiply, and may eventually reach levels of 10⁴-10⁵ per ml, or even more. At this point, the numbers stop increasing, and stay more or less constant for long periods. A recent UK survey of eight brands of bottled water sampled from retailers showed bacterial levels of up to 10⁴ per ml, which remained largely unchanged after six months storage.

The limiting factor on growth is the low nutrient level in bottled waters. It seems that numbers are controlled by a continuous recycling of these nutrients. The bacterial population increases until the available nutrients are exhausted, and then cells begin to die. As they die, the cells lyse and release nutrients back to the water, thus allowing further growth. This cycle allows a constant population to be maintained. More growth seems to occur in plastic bottles than in glass ones, although it is not known exactly why this should be. It may be inhibition of growth by traces of cleaning chemicals on the surface of glass, stimulation of growth by organic nutrients leaching from plastic, or the greater suitability of plastic surfaces for bacterial colonisation.

The allochthanous flora, by contrast, tend to die out during storage. Many reports have been published on the survival of various bacteria (mostly pathogens) in bottled waters, and although results are very variable, a significant decline in numbers is usually apparent after a few weeks. One exception to this pattern is Pseudomonas aeruginosa, which was reported to persist for up to five years in one study.

Public health concerns

Although there have been several well publicised incidents of chemical contamination in bottled water, notably benzene in Perrier water and elevated bromate levels in Coca-Cola's disastrous 'Dasani' bottled tap water venture, microbiological problems have been few and far between.

An outbreak of cholera in Portugal in 1974 was associated with bottled drinking water, the source of which was probably a nearby river. However, since 1980, when new European regulations were implemented, there appear to be no recorded cases of disease associated with consumption of bottled water in Europe. There have been some recalls because of coliform contamination, including major incidents in the UK and Norway, but I am not aware of any documented cases of illness to date. This is a good safety record, but it could lead to complacency, especially at a time of rapidly increasing production.

The first thing to be said is that the autochthanous flora are not generally a cause for concern. These organisms are not adapted to living in the bodies of animals, and are believed to be harmless, even at the relatively high levels sometimes found in bottled water. The exception to this might be for some severely immunocompromised individuals, such as AIDS patients.

Of more concern are some of the emerging pathogens, such as Aeromonas, which is adapted to survival in water and which may cause gastroenteritis, and more particularly E. coli O157. This potentially serious pathogen is generally more resistant than other E. coli, and a recent study suggests it may be able to survive in bottled water for up to 63 days. Since only a few cells of E. coli O157 are enough to cause illness, protection of underground sources of natural mineral water from contamination is critical. There have already been several outbreaks associated with contaminated wells in North America.

Ps. Aeruginosa is also a concern, since it may be able to grow in bottled water. Although not a food poisoning organism, Ps. Aeruginosa is an opportunist pathogen and can cause a variety of infections. It may colonise bottling lines and therefore effective cleaning and hygiene are essential precautions.

There have also been more controversial suggestions that bottled water may be the cause of some Campylobacter infections, but solid evidence for this is hard to find. Campylobacter is reported to survive well in water under some circumstances, but I have not seen reports of its isolation from commercial bottled waters.

A Swiss research report in 2002 found evidence of norovirus contamination in a startling 38% of European bottled water brands tested, suggesting widespread faecal contamination. However, the public health significance of this finding was by no means clear and the methodology used was heavily criticised.

Conclusion

So far, consumers have good reason to be confident about the microbiological safety of bottled water, even though they may be largely unaware of the complex microbial communities living happily in it.

This is just as well, since consumer confidence in bottled water is probably its main selling point, but without strict control of microbiological safety and quality, confidence could easily be lost. Just witness the effect of the famous benzene incident on Perrier sales.

As the sector continues to grow and the pressure on production operations increases, it could be argued that the chances of a bottled water-related outbreak of illness increase. Producers need to remain vigilant if this is to be avoided.