Abstract

A 38-year prospective study of 992 New Zealand children on fluoride (F) and IQ, of whom an uncertain number (0–101) had not received, at age 5 years, additional F with F tablets (taken: 139, not taken: 763, unknown: 90), community water fluoridation (CWF) (lived in CWF area: 891, never lived in CWF area: 99, unknown: 2) or F toothpaste (always used: 634, sometimes used: 240, never used: 22, unknown: 96), has led to differing advice on avoiding Type I and Type II errors. Consideration of eight studies on F and IQ, and introducing a safety factor of 10 to allow for individual differences in water intake and sensitivity (between- and within-subject variations), suggests a maximum contaminant level goal (MCLG) for F in drinking water of 0.1 mg F/L. Whilst achieving this may be difficult in many areas, at the least no active steps should be taken to raise water F levels above 0.1 mg F/L. Some evidence suggests that the only assuredly safe level of F in drinking water is zero.