Search

BackgroundStreet imagery is a promising and growing big data source providing current and historical images in more than 100 countries. Studies have reported using this data to audit road infrastructure and other built environment features. Here we explore a novel application, using Google Street View (GSV) to predict travel patterns at the city level.MethodsWe sampled 34 cities in Great Britain. In each city, we accessed 2000 GSV images from 1000 random locations. We selected archived images from time periods overlapping with the 2011 Census and the 2011-2013 Active People Survey (APS). We manually annotated the images into seven categories of road users. We developed regression models with the counts of images of road users as predictors. The outcomes included Census-reported commute shares of four modes (combined walking plus public transport, cycling, motorcycle, and car), as well as APS-reported past-month participation in walking and cycling.ResultsWe found high correlations between GSV counts of cyclists ('GSV-cyclists') and cycle commute mode share (r = 0.92)/past-month cycling (r = 0.90). Likewise, GSV-pedestrians was moderately correlated with past-month walking for transport (r = 0.46), GSV-motorcycles was moderately correlated with commute share of motorcycles (r = 0.44), and GSV-buses was highly correlated with commute share of walking plus public transport (r = 0.81). GSV-car was not correlated with car commute mode share (r = -0.12). However, in multivariable regression models, all outcomes were predicted well, except past-month walking. The prediction performance was measured using cross-validation analyses. GSV-buses and GSV-cyclists are the strongest predictors for most outcomes.ConclusionsGSV images are a promising new big data source to predict urban mobility patterns. Predictive power was the greatest for those modes that varied the most (cycle and bus). With its ability to identify mode of travel and capture street activity often excluded in routinely carried out surveys, GSV has the potential to be complementary to new and traditional data. With half the world's population covered by street imagery, and with up to 10 years historical data available in GSV, further testing across multiple settings is warranted both for cross-sectional and longitudinal assessments.

To investigate whether and how social class and social mobility in grandparents and parents predict alcohol-related disorders (ARDs) in males and females aged 12+ years, and whether intergenerational social prediction of ARDs varies across time periods.The study sample included four successive generations (G) of Swedish families from the Uppsala Birth Cohort Multigenerational Study: G0 born 1851-1912; G1 born 1915-1929; G2 born 1940-1964 and G3 born 1965-1989. Two study populations were created, each consisting of grandparents, parents and offspring: population I 'G0-G1-G2' (offspring n = 18 430) and population II 'G1-G2-G3' (offspring n = 26 469). Registers and archives provided data on ancestors' socio-demographic factors and ARD history, together with offspring ARD development between 1964-2008. Cox regression models examined the hazard of offspring ARD development according to grandparental social class and grandparental-to-parental social trajectories, controlling for offspring birth year, grandmother's and mother's marital status and parental ARDs.Disadvantaged grandparental social class predicted increased ARD risk in offspring in population I, although the effect attenuated and became non-significant in males after adjusting for parental characteristics (adjusted hazard ratio (HR) = 1.80 (95%CI; 1.07, 3.03) in females, HR = 1.32 (95%CI; 0.93, 1.89) in males). In population II, no increase in ARD risk by grandparental social was evident. In both populations, males were at the highest ARD risk if both parents and grandparents belonged to disadvantaged social class (population I: HR = 1.82 (95%CI; 1.22-2.72); population II: HR = 1.68 (95%CI; 1.02-2.76)).Intergenerational social patterning of ARDs appears to be time-contextual and gender-specific. The role of grandparental social class in developing ARDs in grandchildren seems to decline over time, while persistent grandparental-to-parental social disadvantage remains associated with higher ARD risk in males. When targeting higher risk groups, continuity of familial social disadvantage, particularly among males, should be considered.

To investigate which facets of parent and grandparent socio-economic position (SEP) are associated with eating disorders (ED), and how this varies by ED subtype and over time.Total-population cohort study of 1,040,165 females and 1,098,188 males born 1973-1998 in Sweden, and followed for inpatient or outpatient ED diagnoses until 2010. Proportional hazards models estimated associations with parental education, income and social class, and with grandparental education and income.15,747 females and 1051 males in our sample received an ED diagnosis, with rates increasing in both sexes over time. ED incidence in females was independently predicted by greater educational level among the father, mother and maternal grandparents, but parent social class and parental income showed little or no independent effect. The associations with education were equally strong for anorexia nervosa, bulimia nervosa and ED not-otherwise-specified, and had increased over time. Among males, an apparently similar pattern was seen with respect to anorexia nervosa, but non-anorexia ED showed no association with parental education and an inverse association with parental income.Family history of education predicts ED in gender- and disorder-specific ways, and in females the effect is observed across multiple generations. Particularly given that these effects may have grown stronger in more recent cohorts, these findings highlight the need for further research to clarify the underlying mechanisms and identify promising targets for prevention. Speculatively, one such mechanism may involve greater internal and external demands for academic success in highly educated families.

Increasing walking and cycling, and reducing motorised transport, are health and environmental priorities. This paper examines levels and trends in the use of different commute modes in England and Wales, both overall and with respect to small-area deprivation. It also investigates whether commute modal share can serve as a proxy for travel behaviour more generally.23.7 million adult commuters reported their usual main mode of travelling to work in the 2011 census in England and Wales; similar data were available for 1971-2001. Indices of Multiple Deprivation were used to characterise socio-economic patterning. The National Travel Survey (2002-2010) was used to examine correlations between commute modal share and modal share of total travel time. These correlations were calculated across 150 non-overlapping populations defined by region, year band and income.Among commuters in 2011, 67.1% used private motorised transport as their usual main commute mode (-1.8 percentage-point change since 2001); 17.8% used public transport (+1.8% change); 10.9% walked (-0.1% change); and 3.1% cycled (+0.1% change). Walking and, to a marginal extent, cycling were more common among those from deprived areas, but these gradients had flattened over the previous decade to the point of having essentially disappeared for cycling. In the National Travel Survey, commute modal share and total modal share were reasonably highly correlated for private motorised transport (r = 0.94), public transport (r = 0.96), walking (r = 0.88 excluding London) and cycling (r = 0.77).England and Wales remain car-dependent, but the trends are slightly more encouraging. Unlike many health behaviours, it is more common for socio-economically disadvantaged groups to commute using physically active modes. This association is, however, weakening and may soon reverse for cycling. At a population level, commute modal share provides a reasonable proxy for broader travel patterns, enhancing the value of the census in characterising background trends and evaluating interventions.

Growing up with many siblings, at least in the context of modern post-industrial low fertility, low mortality societies, is predictive of relatively poor performance on school tests in childhood, lower levels of educational attainment, and lower income throughout adulthood. Recent studies further indicate these relationships hold across generations, so that the descendants of those who grow up with many siblings are also at an apparent socioeconomic disadvantage. In this paper we add to this literature by considering whether such relationships interact with the sex and relative age of siblings. To do this we utilise a unique Swedish multigenerational birth cohort study that provides sibling configuration data on over 10,000 individuals born in 1915-1929, plus all their direct genetic descendants to the present day. Adjusting for parental and birth characteristics, we find that the 'socioeconomic cost' of growing up in a large family is independent of both the sex of siblings and the sex of the individual. However, growing up with several older as opposed to several younger siblings is predictive of relatively poor performance on school tests and a lower likelihood of progression to tertiary education. This later-born disadvantage also holds across generations, with the children of those with many older siblings achieving lower levels of educational attainment. Despite these differences, we find that while individual and descendant income is negatively related to the number of siblings, it is not influenced by the relative age of siblings. Thus, our findings imply that the educational disadvantage of later-born children, demonstrated here and in numerous other studies, does not necessarily translate into reduced earnings in adulthood. We discuss potential explanations for this pattern of results, and consider some important directions for future research into sibling configuration and wellbeing in modern societies.