Sixty-eight engaged couples, 113 newly weds, and 222 couples married for over 5 years, were studied for a number of physical, psychological and socioeconomic traits. The measurements were repeated a year later. Correlations were highest for age (r = 0-454, 0-903 and 0-888 respectively). They were also high for height (but not weight), neuroticism (but not extroversion), occupation, education, religion and smoking. Similarities were already present about the time of marriage, with little convergence or divergence. Thus, men do not grow to resemble their wives, nor vice versa.

Assortative marriage for smoking habits was studied in 68 engaged couples, 112 newly wed couples, and 223 couples married for six years or more. All three groups showed a fairly strong correlation between couples for smoking, r = 0.332 +/- 0.047 (mean + SD). Smoking habits were highly intercorrelated with social class and education but the correlation for smoking remained highly significant, even when correction was made for these factors. Because this correlation was present even in those who have not yet married, smoking habits may well be an important factor in choosing a marriage partner

(J Med Screen 2001;8:183-186)

We suggest that the current notification system is not working in respect of gastroenteritis and food poisoning, and should be either substantially revised or abandoned.

The effectiveness of breast screening in women aged 50-64 has been well established by randomised trials, and recent analyses have shown that screening in women aged 65-69 is probably as effective in terms of reducing mortality from breast cancer as screening in women aged 50-64 (Chen et al, 1995).The report of the Forrest committee (Forrest, 1986) in 1986, which led to the implementation of the NHS Breast Screening Programme in the United Kingdom, recommended that women aged 50-64 be invited for screening every 3 years, with women aged 65 and over being able to self-refer. The main reason for this distinction was concern over possible lower acceptance rates among older women, together with possible reduced costeffectiveness due to decreased life expectancy in older women.Population screening in a number of countries includes women up to age 69 or 70 (Shapiro et al, 1998). Some demonstration studies of inviting older women also showed that their uptake was only slightly lower than that for 50-64 year olds from the same locality, implying that reasonable uptakes could be achieved across the NHS as a whole (Hobbs et al, 1990;Hendry and Entwhistle, 1996;Horton-Taylor et al, 1996). The number of selfreferrals in women aged 65 and over had increased to 65 032 by 1998/9, 44 811 of these being aged 65-69.Demonstration studies at 3 sites were established by the Department of Health to determine the implications of extending the NHS Breast Screening Programme to women aged 65-69 by including this age-group in the routine invitation system. Findings from one of the sites after one year have previously been reported (Rubin et al, 1998). This paper presents the results of the evaluation of the full 3 years of all 3 sites. Decisions on the extension of the programme have been made on the basis of the results of these studies. METHODSThe demonstration studies were scheduled to run for 3 years at each of 3 sites, with appointments for women aged 65-69 intermingled with those of women aged 50-64 and self-referrals of older women. As for women aged 50-64, those aged 65-69 were sent by post an invitation to attend at a specific date/time for mammographic screening. Those with an abnormality detected on the mammogram were sent an appointment for further assessment, which could include clinical examination, further mammographic views and ultrasound.East Sussex, Brighton and Hove began screening for the study in May 1996, Nottingham in February 1997 and Leeds and Wakefield in April 1997. This timing means that women aged 65-67 had mostly been invited 3 years previously in the most recent screening round, but those aged 68-69 had mostly missed a round and would have been last invited 6 years previously.The principal screening process and outcome measures were already routinely recorded at each demonstration site as part of the NHSBSP; these measures were analysed by age and time since last screen. Comparisons with equivalent data for the whole of England for the age-group 50-64 have been made to determine the representativeness of t...