Bottled water is such a generic term. There are so many different kinds of bottled water. Some will be municipal supply water put through reverse osmosis or carbon filters or both. Some will be “spring” water. However, in Australia, at least, all need to adhere to the Trade Practices Act, 1968, specifically Part V, Division 2, Section 71: Implied undertakings as to quality or fitness. State legislation than mandates the use of the HACCP (Hazard Analysis and Critical Control Points) (ISO22000) system for quality control during manufacture.
So, it is fair to assume that bottled water is safe to drink, however, the first kind of bottled water described, the one that goes through the RO membranes, if that hasn’t been re-calcified, then there is a real and present possibility that it will corrode any calcium. That is, it will almost certainly have significantly (<-22) a negative CCPP (Calcium Carbonate Precipitation Potential). This means prolonged use of the specific kind of bottled water described will inevitably decay your teeth much quicker than normal. If you must use this water (some very sick patients who use dialysis machines need this very pure form of water) then I recommend fluoride tablets, if you can't add a little calcium salt to the water if at all possible, as this will increase the CCPP and make the water less corrosive. A simpler measure (but fair less accurate) of corrosivity is the Langelier Saturation Index, or LSI. This is more of an indicator, whereas CCPP is an analytical measurement. However, the ability to easily compute the LSI makes it a fair more attractive option to use as a “rule-of-thumb” for corrosivity. The formula for LSI is found on the internet in various forms, and this one is not special. The interpretation of LSI is similar to CCPP, however, smaller negative values have a stronger association with corrosion. I.E: Water with a CCPP of -21 may have an LSI of around -5. The relationship between CCPP and LSI is overtly complicated and this is an over simplified view, and LSI should never, NEVER, be used to calculate CCPP. There are specific models for CCPP (RTW4). However, LSI is a good rule-of-thumb. Negative LSI means corrosive water, positive LSI means water that will produce scale (not a problem for your teeth, but may be a problem for your appliances!). Fluoride in tap water mitigates most of the risk of decay from a low CCPP, and due to the nature of most municipal supplies (concrete (i.e. calcium) trunk mains), the CCPP is usually fairly close to zero anyway. Bottom Line: Tap water is a safer bet than bottled water when it comes to dental decay, however, any water over any other drink is recommended. The issues not discussed here, the fact that by drinking (some forms) of bottled water you may be missing out on some prescribed amount of fluoride, is not much of an issue, for if you brush and floss you will undoubtedly get more fluoride than you need to sustain the ion balance in your dentistry.