McClure's (Reference McClure2000) review of trends in suicide in England and Wales, 1960-1997, provides a timely update of changes in patterns of suicide since Government health targets were set for its reduction (Secretary of State for Health, 1992, 1999). There are, however, two issues relating to his interpretation of trends that require clarification.

First, we were surprised with the suggestion that there have been reductions in suicide rates in all age/gender groups since 1990. Suicide rates in 25- to 34-year-old males have continued to increase in recent years (see Fig., updated to include data for 1998). In 1998, this was the group with the highest suicide rate of all age/gender categories. The trend in 15- to 24-year-old women is less clear-cut, but adding 1998 data indicates that there has been, if anything, an increase in the underlying rate. Thus, while overall suicide rates are declining, these reductions mask rising trends in younger age groups, the very groups which contribute most to premature life-years lost. From the public health perspective, age-specific suicide reduction targets might be more relevant than the overall targets currently set.

Fig. Age-standardised suicide rates for (a) males and (b) females aged 15-34, England and Wales 1979-98, all methods (ICD-10 E950-959) and hanging (E953.0). Trend lines for ‘all methods’ plots are three-year moving averages centred on 1999.

Second, McClure's figures are based on suicide rates (ICD-10 cause of death coded E950-959) alone. However, the Government's targets refer to suicide and undertermined deaths (E950-959 and E980-989, excluding E988.8). We confirmed that the age- and gender-specific trends described above are also seen when rates are calculated to include undetermined deaths. These analyses indicate a 6% increase in 15- to 24-year-old women and a 7% increase in 24- to 34-year-old men between 1990-1992 and 1996-1998.

We share the author's concern about the increasing popularity of hanging, a method whose availability cannot easily be controlled. In 15- to 34-year-old males, hanging (ICD E953.0) accounted for more deaths in 1998 than all other methods together. Since 1979, its use has increased three-fold in young (15- to 34-year-old) men and two-fold in young women. The increase has been especially pronounced in men aged 25-34, in whom the rates doubled in the past eight years (see Fig.). Similar increases have been observed in Australia (Reference Wilkinson and GunnellWilkinson & Gunnell, 2000). Whether such trends represent a shift away from methods that have become less lethal in recent years (such as poisoning and gassing) to more lethal methods is not clear (Reference Gunnell, Wehner and FrankelGunnell et al, 1999). Alternatively, rises may reflect a true deterioration in the mental health of young people.