Objective-To test the hypothesis that farmers are at high risk of hip osteoarthritis and to investigate possible causes for such a hazard.Design-Cross sectional survey. Setting-Five rural general practices. Subjects-167 male farmers aged 60-76 and 83 controls from mainly sedentary jobs. AU those without previous hip replacement underwent radiography of the hip.Main outcome measures-Hip replacement for osteoarthritis or radiological evidence of hip osteoarthritis.Results-Prevalence of hip osteoarthritis was higher in farmers than controls and especialiy in those who had farmed for over 10 years (odds ratio 9.3, 95% confidence interval 1-9 to 44-5). The excess could not be attributed to any one type of farming, and heavy lifting seems the likely explanation.

Study objective-The aim was to assess the geographical variation in low back pain and associated disability in Britain.Design Main results-The overall lifetime and one year period prevalences of low back pain were 58.3% and 36.1%. Rates in men and women were similar. Symptoms were more common in men with manual occupations than in those with non-manual jobs, but in women there was no clear trend in relation to social class. Geographical differences in prevalence were small, but the threshold for consulting general practitioners about symptoms varied markedly from place to place. After allowance for age, sex, social class, and severity of symptoms, subjects in the northern towns ofArbroath and Peterlee who had suffered from low back pain in the past year were three to four times as likely to have consulted their doctor about the problem as those living in the southern towns ofSt Austell and Dorking. Consultation rates in the Midlands were intermediate. Conclusions-Geographical variation inrates of general practice consultation for low back pain in Britain is due largely to differences in patient behaviour once symptoms have developed. The distribution of important causes of low back back pain across the country is probably fairly uniform.

Study objective-The aim was to assess the influence of childhood reading on the development of myopia after allowance for familial differences in susceptibility.Design-The study was a cross sectional survey. Setting-Four fishing harbours in HongKong in 1989.Subjects-Participants were 408 men and women aged 15-39 years old from 159 families.Main results-Histories of school attendance and reading habits in childhood were obtained at interview. Myopia was assessed by retinoscopy. Associations between myopia (defined as a refractive error of at least -1-0D in one or both eyes) and indices of reading in childhood were explored. Myopia was more common in subjects who had attended school (odds. ratio=1l7, 95% CI 1-0-3-0), with the highest risks in those who had started school at the earliest ages and who had spent the most time reading and writing while at primary school. Allowance for familial tendency to myopia produced no diminution in the risks associated with reading.Conclusions-These data support the hypothesis that reading in childhood is a cause of short sight.