ObjectivesTo describe the characteristics of children and adults with incident type 1 diabetes in contemporary, multiethnic UK, focusing on differences between the islet autoantibody negative and positive.DesignObservational cohort study.Setting146 mainly secondary care centres across England and Wales.Participants3312 people aged ≥5 years were recruited within 6 months of a clinical diagnosis of type 1 diabetes via the National Institute for Health Research Clinical Research Network. 3021 were of white European ethnicity and 291 (9%) were non-white. There was a small male predominance (57%). Young people <17 years comprised 59%.Main outcome measuresAutoantibody status and characteristics at presentation.ResultsThe majority presented with classical osmotic symptoms, weight loss and fatigue. Ketoacidosis was common (42%), especially in adults, and irrespective of ethnicity. 35% were overweight or obese. Of the 1778 participants who donated a blood sample, 85% were positive for one or more autoantibodies against glutamate decarboxylase, islet antigen-2 and zinc transporter 8. Presenting symptoms were similar in the autoantibody-positive and autoantibody-negative participants, as was the frequency of ketoacidosis (43%vs40%, P=0.3). Autoantibody positivity was less common with increasing age (P=0.0001), in males compared with females (82%vs90%, P<0.0001) and in people of non-white compared with white ethnicity (73%vs86%, P<0.0001). Body mass index was higher in autoantibody-negative adults than autoantibody-positive adults (median, IQR 25.5, 23.1–29.2vs23.9, 21.4–26.7 kg/m2; P=0.0001). Autoantibody-negative participants were more likely to have a parent with diabetes (28%vs16%, P<0.0001) and less likely to have another autoimmune disease (4%vs8%, P=0.01).ConclusionsMost people assigned a diagnosis of type 1 diabetes presented with classical clinical features and islet autoantibodies. Although indistinguishable at an individual level, autoantibody-negative participants as a group demonstrated features more typically associated with other diabetes subtypes.Trial registration numberISRCTN66496918; Pre-results.

The authors represent a research consortium 1 which has adopted a task performance based approach for nighttime driving to establish a system for photometry in the mesopic region. This article analyses the experimental investigations described in earlier articles on visual performance in the mesopic domain using reaction time, detection threshold, and discrimination threshold techniques. These results are used to develop a system for mesopic photometry, which balances the quality of the fit to the experimental data with the ease of practical implementation by the lighting industry. A more complex model is also described, which takes account of the chromatic visual response channels and thus provides a better fit to some of the experimental results (particularly those involving monochromatic stimuli), but describes the totality of the data less well and is furthermore less suitable for practical photometric measurements.

Achromatic thresholds are measured at three sites to investigate the effects of light spectrum on achromatic detection thresholds in the mesopic region. The results of measurements using quasi-monochromatic targets reveal spectral sensitivity functions with two or three peaks, a so-called 'chromatic effect' which is assumed to be due to the influence of the colour-opponent channels. This chromatic contribution seem to be less significant at lower luminances but it is accentuated for peripheral observation. Results obtained with broadband stimuli show that the contrast threshold, defined in terms of V(), is markedly higher for red targets than for other colours, particularly at lower light levels. For these broadband targets, contrast threshold values calculated using the new spectral sensitivity curves instead of V() are in better agreement with the experimental results, particularly for the peripheral detection task.

At present, suitable methods to evaluate the visual effectiveness of lighting products in the mesopic region are not available. The majority of spectral luminous efficiency functions obtained to date in the mesopic range have been acquired by heterochromatic brightness matching. However, the most recent studies in the mesopic field have adopted a task performance-based approach. This paper summarizes the major mesopic models proposed so far, presenting in detail the experimental conditions of these studies. The authors represent a research consortium which has adopted the task performance-based approach for night-time driving in which mesopic visual performance has been divided into three subtasks. Data for each sub-task will be generated by using a set of common parameter values and 120 observers. The approach and methods used by the consortium are presented.

Reaction times are recorded to chromatic, mesopic stimuli to investigate mesopic reaction time spectral sensitivity. Measurements are made using three laboratory setups and a driving simulator. The chromatic stimuli have spectral distributions that range from quasi-monochromatic to broadband. Reaction time spectral sensitivity for small (0.298) foveal stimuli changes little with luminance and is described adequately by V (l). Reaction time spectral sensitivity for peripheral stimuli exhibits a Purkinje shift, and approaches V 0 (l) at 0.01 cd m À2 . In the periphery, the results for medium-sized stimuli (28) reflect the activity of colour-opponent neurons in addition to rod and cone-based achromatic activity, resulting in three-peaked spectral sensitivity functions.

Little is known about how color signals and cone- and rod-based luminance signals contribute to perceived contrast in the mesopic range. In this study the perceived contrast of colored, mesopic stimuli was matched with that of spatially equivalent achromatic stimuli. The objective was to develop a metric for perceived contrast in the mesopic range in terms of an equivalent achromatic luminance contrast, referred to here as effective contrast. Stimulus photopic luminance contrast, scotopic luminance contrast, and chromatic difference from the background all contributed to effective contrast over the mid-mesopic range, but their contributions were not independent and varied markedly with background luminance. Surprisingly, color made a significant contribution to effective contrast from 10 to approximately 0.003 cd m(-2). A model describing this relationship is introduced (R2 = 0.89) and compared with predictions of mesopic luminance contrast obtained from a number of models proposed as systems of mesopic photometry.

The spectral sensitivity of the eye was investigated using reaction times to broadband chromatic stimuli over a range of background luminances. Relative sensitivity was determined from the nonlinear reaction time curve by converting reaction times to a linear measure that was independent of spectral sensitivity. Two models for mesopic spectral sensitivity were compared. The first was a linear combination of V(lambda) and V'(lambda), and the second included input from the L-M colour-opponent mechanism and the S-cones. The second model produced a significantly better fit to the data. The chromatic mechanisms appear to contribute to reaction time when there is an appreciable chromatic signal but luminance contrast is low.

The authors represent a research consortium which has adopted a task performance based approach for nighttime driving to obtain mesopic spectral sensitivity functions. This study describes investigations based on discrimination threshold techniques and compares the results with those obtained from detection threshold measurements. It forms a continuation of `Mesopic models — from brightness matching to visual performance in night-time driving: a review'. 1 The achromatic increment contrast threshold for discrimination of a quasi-static target represents the ability to correctly identify a target which has low contrast with the background and is therefore a highly relevant visual task in the context of nighttime driving. The psychophysical experiments reported here are carried out in the laboratory and in dynamic nighttime driving tests. The spectral sensitivity functions resulting from the discrimination threshold contrast investigations show a significant contribution from the colour-opponent (or chromatic) channels of the visual system. Similar features have been observed in results from investigations using detection threshold techniques.