From motorised to active travel: using GPS data to explore potential physical activity gains among adolescents.

BACKGROUND: A high proportion of adolescents worldwide are not doing enough physical activity for health benefits. Replacing short motorised trips with walking or cycling has the potential to increase physical activity at the population level. This study aimed to estimate the proportion of short distance motorised trips that could be replaced with walking or cycling, and the potential physical activity gains by sociodemographic and trip characteristics. METHODS: Data were from a subsample of the NEighbourhood Activity in Youth (NEArbY) study conducted among adolescents in Melbourne. A total of 217 adolescents with at least one motorised trip completed a survey and wore a Global Positioning Systems (GPS) device for eight consecutive days. Classification of travel modes were based on speed. GPS data points were geocoded in ArcGIS. Motorised trips within walkable (1.3 km) and cyclable (4.2 km) distances were identified (threshold based on 80th percentile of walking and cycling trip distances among Victorian adolescents), and the additional physical activity minutes that could be accrued by replacing walkable or cyclable motorised trip to active trips were quantified. Multilevel linear regression was used to assess differences in physical activity minutes gain by sociodemographic and trip characteristics. RESULTS: A total of 4,116 motorised trips were made. Of these, 17% were walkable and 61% were cyclable. Replacing motorised trips by walking and cycling resulted in estimated gains of six minutes and 15 min of physical activity per day, respectively. CONCLUSION: The sizable proportion of replaceable trips and potential physical activity gains from this shift calls for attention to improve safe and connected infrastructure to support active travel.

Do residents with a 20-min neighbourhood walk more? Findings from ProjectPLAN.

BACKGROUND: The 20-min neighbourhood (20 MN) concept aims to provide people the ability to meet their daily needs within a 20-min non-motorised trip from home. Evidence as to whether the 20 MN encourages more walking for transport or recreation is currently absent. METHODS: This cross-sectional study used self-reported data from the Places and Locations for Activity and Nutrition study (ProjectPLAN) targeting adults (n = 843) residing in Melbourne or Adelaide, Australia. Multiple services and amenities were used to represent access to five service domains (healthy food, community resources, recreational resources, public open space, public transport). Address points meeting the access criteria for each of the five domains were defined as having a 20 MN. Non-20 MNs were defined as having five or fewer individual services and amenities. This study examined if those residing in a 20 MN compared with a non-20MN undertook more walking for transport or for recreation. The analysis considered separately each of the cities to support the estimation of effects specific to each local context. RESULTS: Respondents residing in a 20 MN relative to a non-20MN had higher odds of walking for transport in Melbourne (OR = 4.24, 95% CI = 2.38, 7.56), whilst in Adelaide there was no evidence of a difference (OR = 1.31, 95% CI = 0.80, 2.13). In Melbourne, the mean time spent walking for transport was greater for 20 MNs (82.5 min/week, 95% CI = 65.3, 99.7) compared to non-20MNs (41.2 min/week, 95% CI = 32.7, 49.7). Whilst minutes spent walking for recreation was higher than minutes spent walking for transport, no differences were found between neighbourhood types and walking for recreation in either city. CONCLUSION: 20 MNs appeared to promote walking for transport in the higher density setting of Melbourne, but no association was observed in the lower density city of Adelaide. Further investigation is required to determine other factors beyond service provision that can promote walking for transport in Adelaide (e.g. pedestrian safety). Recreational walking did not differ across neighbourhood types highlighting that service provision and thus the 20 MN is not related to walking for exercise/recreation purposes.

Outdoor public recreation spaces and social connectedness among adolescents.

BACKGROUND: Outdoor public recreation spaces are important settings for leisure and physical activity. Adolescents' use of these spaces may contribute to social connectedness via social interaction with peers and the community in these settings. However, research on this topic is limited. This exploratory study examined associations of frequency of visitation and physical activity in outdoor public recreation spaces with social connectedness among adolescents in Melbourne, Australia. METHODS: Adolescents self-reported their frequency of visitation to parks, trails, beach/lake, and sports facilities; frequency of physical activity in a park, local street or path, and their street; and social connectedness. Separate analyses were conducted for visitation (n = 349, 15.4 ± 1.6 years, 58% female) and physical activity (n = 441, 15.4 ± 1.6 years, 59% female) using multilevel linear regression models. RESULTS: No significant associations were observed for frequency of visitation to a park (B = 0.86, 95% CI = - 0.26, 1.99), trails (B = 0.41, 95% CI = - 0.61, 1.44), beach/lake (B = - 0.44, 95% CI = - 1.46, 0.57), or sports facilities (B = 0.64, 95% CI = - 0.43, 1.70), nor for frequency of physical activity in their street (B = - 0.07, 95% CI = - 0.46, 0.31), local street/path (B = - 0.05, 95% CI = - 0.43, 0.33) or in a park (B = 0.23, 95% CI = - 0.14, 0.60) with adolescents' social connectedness. CONCLUSIONS: The findings did not support the hypothesis that visiting and being active in outdoor public recreation spaces are associated with adolescents' social connectedness. Future research should consider the duration and context of outdoor public recreation space use (e.g., sports, socialising, relaxing alone) and whether different types and/or a combination of public spaces are more/less conducive to social connectedness.

Neighbourhood food typologies, fast food outlet visitation and snack food purchasing among adolescents in Melbourne, Australia.

OBJECTIVE: Despite the increased attention on neighbourhood food environments and dietary behaviours, studies focusing on adolescents are limited. This study aims to characterise typologies of food environments surrounding adolescents and their associations with fast food outlet visitation and snack food purchasing to/from school. DESIGN: The number of food outlets (supermarket; green grocers; butcher/seafood/deli; bakeries; convenience stores; fast food/takeaways; café and restaurants) within a 1 km buffer from home was determined using a Geographic Information System. Adolescents' self-reported frequency of fast food outlet visitation and snack food purchasing to/from school. Latent Profile Analysis was conducted to identify typologies of the food environment. Cross-sectional multilevel logistic regression analyses were conducted to examine the relationships between food typologies, fast food outlet visitations and snack food purchasing to/from school. SETTING: Melbourne, Australia. PARTICIPANTS: Totally, 410 adolescents (mean age= 15·5 (sd = 1·5) years). RESULTS: Four distinct typologies of food outlets were identified: (1) limited variety/low number; (2) some variety/low number; (3) high variety/medium number and (4) high variety/high number. Adolescents living in Typologies 1 and 2 had three times higher odds of visiting fast food outlets ≥1 per week (Typology 1: OR = 3·71, 95 % CI 1·23, 11·19; Typology 2: OR = 3·65, 95 % CI 1·21, 10·99) than those living in Typology 4. No evidence of association was found between typologies of the food environments and snack food purchasing behaviour to/from school among adolescents. CONCLUSION: Local government could emphasise an overall balance of food outlets when designing neighbourhoods to reduce propensity for fast food outlet visitation among adolescents.

Understanding children's preference for park features that encourage physical activity: an adaptive choice based conjoint analysis.

BACKGROUND: Parks are a key setting for physical activity for children. However, little is known about which park features children prefer and which features are most likely to encourage them to be active in parks. This study examined the relative importance of park features among children for influencing their choice of park for engaging in park-based physical activity. METHODS: Children (n = 252; 8-12 years, 42% male) attending three primary schools in Melbourne, Australia completed a survey at school. They were required to complete a series of Adaptive Choice-Based Conjoint analysis tasks, with responses used to identify the part-worth utilities and relative importance scores of selected park features using Hierarchical Bayes analyses within Sawtooth Software. RESULTS: For the overall sample and both boys and girls, the most important driver of choice for a park that would encourage them to be active was presence of a flying fox (overall conjoint analysis relative importance score: 15.8%; 95%CI = 14.5, 17.1), followed by a playground (13.5%; 95%CI = 11.9, 15.2). For the overall sample, trees for climbing had the third highest importance score (10.2%; 95%CI = 8.9, 11.6); however, swings had 3rd highest importance for girls (11.1, 95%CI = 9.3, 12.9) and an obstacle course/parkour area had the 3rd highest importance score for boys (10.7, 95%CI = 9.0, 12.4). For features with two levels, part-worth utility scores showed that the presence of a feature was always preferred over the absence of a feature. For features with multiple levels, long flying foxes, large adventure playgrounds, lots of trees for climbing, large round swings, large climbing equipment, and large grassy open space were the preferred levels. CONCLUSION: To ensure parks appeal as a setting that encourages children to engage in physical activity, park planners and local authorities and organisations involved in park design should prioritise the inclusion of a long flying fox, large adventure playgrounds, lots of trees for climbing, large round swings and obstacle courses/parkour areas.

Important park features for encouraging park visitation, physical activity and social interaction among adolescents: A conjoint analysis.

Little is known about what park features adolescents perceive as most important for encouraging them to visit and be active and social in parks. This study examined the relative importance of park features for encouraging these behaviours among adolescents and explored differences according to gender and frequency of visitation. Adolescents were recruited from secondary schools in diverse areas of Melbourne, Australia. In Phase 1 (2018-2019), participants [n = 222; 14.0 years (SD 1.1); 46% female] rated photographs of 40 park features in an online survey, indicating the degree to which each feature encouraged them to visit, and be active and social in the park. In Phase 2 (2019-2020), the top ten-rated features from Phase 1 for each outcome (overall and by gender and frequency of visitation) were presented in Adaptive Choice-Based Conjoint tasks to another sample of participants [n = 244; 14.7 years (SD 1.3); 46% female]. The most important feature for encouraging visitation was large swings; for physical activity was sports courts; and for social interaction was a café. Grassy open space was important for all behaviours, and preferred park features varied among sub-groups. The results highlight which park features park planners and designers should prioritise to support active and social park use among adolescents.

Residential vs school neighborhoods: Associations with physical activity among adolescents.

Adolescents have the potential to be active in multiple places across weekdays and weekends. While the built environment features around home are known to be important for adolescents' physical activity, specific built environment features that facilitate physical activity outside the home neighbourhood are not well explored. Given that adolescents spend most days of the week at school, the built environment surrounding school may be particularly important for supporting their physical activity. This study examined cross-sectional associations of objective built environments around home and school at multiple spatial scales with accelerometer assessed moderate-to-vigorous physical activity (MVPA) across the week. Moderating effects of adolescents' age were further explored. Data from up to 417 adolescents aged 12-20 years from 18 schools in Melbourne, Australia were used. Features of the built environment around school were positively associated with MVPA on weekdays; and a wide range of built environment features around home were found to be important for MVPA during weekends. Age was found to moderate the relationship between some built environment features around both home and school and MVPA across the week. Understanding the contribution of the built environment features around both home and school with MVPA could inform policy recommendations and public health interventions for each setting.

Environmental Mismatch: Do Associations between the Built Environment and Physical Activity among Youth Depend on Concordance with Perceptions?

Without accurate awareness of features within the built environment, the availability of a supportive built environment alone may not be sufficient to influence physical activity levels. We examined the moderating effects of concordance/discordance between selected objective and perceived built environment features in the relationship between objective built environment features and physical activity. Cross-sectional data from 465 youth aged 12-20 years from 18 schools in Melbourne, Australia were used. The relationship between trails and physical activity differed by concordance/discordance. There were positive relationships among those with concordant perceptions, but no significant differences for those with discordant perceptions. At least for trails, environmental interventions designed to enhance physical activity may be less effective if efforts are not made to enhance individuals' awareness of their environment.

Built environment and physical activity among adolescents: the moderating effects of neighborhood safety and social support.

BACKGROUND: Increasing emphasis has been placed on improving physical activity levels through multilevel interventions. This study aims to examine moderating effects of neighborhood safety (crime and traffic) and social support (from parent and sibling/peer) for physical activity in the relationship between the built environment and moderate-to-vigorous physical activity (MVPA) outside school hours among adolescents in Melbourne. METHODS: Data were from the NEighbourhood Activity in Youth study conducted among adolescents in Melbourne, Australia (n = 358, 15.3 (SD = 1.5) years). MVPA outside school hours was assessed by accelerometer. Built environment features within 1 km and 2 km residential buffers including recreation facilities, park area, and walkability and its components were assessed using Geographic Information Systems. Neighborhood safety, social support for physical activity and sociodemographic information were self-reported by adolescents. Multilevel linear regression models were used to estimate associations. RESULTS: Support for physical activity from sibling/peer positively moderated the relationship between recreation facilities (1 km), residential density (2 km) and MVPA. Recreation facility (count within 2 km), walkability (1 km and 2 km) and residential density (1 km) had significant positive associations with MVPA outside school hours. CONCLUSION: The built environment appeared to have stronger facilitating effects on MVPA among adolescents who had favourable support for physical activity from their sibling or peer. Multilevel interventions that target the built environment and social factors are needed to promote MVPA outside school hours among adolescents.

Do Differences in Social Environments Explain Gender Differences in Recreational Walking across Neighbourhoods?

Within a city, gender differences in walking for recreation (WfR) vary significantly across neighbourhoods, although the reasons remain unknown. This cross-sectional study investigated the contribution of the social environment (SE) to explaining such variation, using 2009 data from the How Areas in Brisbane Influence healTh and AcTivity (HABITAT) study, including 7866 residents aged 42-67 years within 200 neighbourhoods in Brisbane, Australia (72.6% response rate). The analytical sample comprised 200 neighbourhoods and 6643 participants (mean 33 per neighbourhood, range 8-99, 95% CI 30.6-35.8). Self-reported weekly minutes of WfR were categorised into 0 and 1-840 mins. The SE was conceptualised through neighbourhood-level perceptions of social cohesion, incivilities and safety from crime. Analyses included multilevel binomial logistic regression with gender as main predictor, adjusting for age, socioeconomic position, residential self-selection and neighbourhood disadvantage. On average, women walked more for recreation than men prior to adjustment for covariates. Gender differences in WfR varied significantly across neighbourhoods, and the magnitude of the variation for women was twice that of men. The SE did not explain neighbourhood differences in the gender-WfR relationship, nor the between-neighbourhood variation in WfR for men or women. Neighbourhood-level factors seem to influence the WfR of men and women differently, with women being more sensitive to their environment, although Brisbane's SE did not seem such a factor.

Neighbourhood built environment and physical function among mid-to-older aged adults: A systematic review.

This systematic review included 23 quantitative studies that estimated associations between aspects of the neighbourhood built environment and physical function among adults aged ≥45 years. Findings were analysed according to nine aspects of the neighbourhood built environment: walkability, residential density, street connectivity, land use mix, public transport, pedestrian infrastructure, aesthetics, safety and traffic. Evidence was found for a positive association of pedestrian infrastructure and aesthetics with physical function, while weaker evidence was found for land use mix, and safety from crime and traffic. There was an insufficient number of studies for walkability, residential density, street connectivity and access to public transport.

The potential for walkability to narrow neighbourhood socioeconomic inequalities in physical function: A case study of middle-aged to older adults in Brisbane, Australia.

Residents of disadvantaged neighbourhoods have poorer physical function than their advantaged counterparts, although the reasons for this remain largely unknown. We examined the moderating effects of walkability in the relationship between neighbourhood disadvantage and physical function using 2013 cross-sectional data from 5115 individuals aged 46-72 living in 200 neighbourhoods in Brisbane, Australia. The relationship between neighbourhood disadvantage and physical function differed by levels of walkability: positive associations as levels of walkability increased for those living in more disadvantaged neighbourhoods, and no difference for those living in more advantaged neighbourhoods. Further work is required to better understand the underlying mechanisms.

Neighborhood Disadvantage and Physical Function: The Contributions of Neighborhood-Level Perceptions of Safety From Crime and Walking for Recreation.

BACKGROUND: Residents of more socioeconomically disadvantaged neighborhoods are more likely to report poorer physical function, although the reasons for this remain unknown. It is possible that neighborhood-level perceptions of safety from crime contribute to this relationship through its association with walking for recreation. METHODS: Data were obtained from the fourth wave (collected in 2013) of the HABITAT (How Areas in Brisbane Influence HealTh and AcTivity) multilevel longitudinal study of middle- to older-aged adults (46-74 y) residing in 200 neighborhoods in Brisbane, Australia. The data were analyzed separately for men (n = 2190) and women (n = 2977) using multilevel models. RESULTS: Residents of the most disadvantaged neighborhoods had poorer physical function, perceived their neighborhoods to be less safe from crime, and do less walking for recreation. These factors accounted for differences in physical function between disadvantaged and advantaged neighborhoods (24% for men and 25% for women). CONCLUSION: This study highlights the importance of contextual characteristics, through their associations with behaviors, that can have in explaining the relationship between neighborhood disadvantage and physical function. Interventions aimed at improving neighborhood safety integrated with supportive environments for physical activity may have positive impact on physical function among all socioeconomic groups.

Changes in waist circumference independent of weight: Implications for population level monitoring of obesity.

Population monitoring of obesity is most commonly conducted using body mass index (BMI). We test the hypothesis that because of increases in waist circumference (WC) independent of increases in weight, BMI alone detects an increasingly smaller proportion of the population with obesity. METHODS: Australian adults with measured height, weight, and WC were selected from three nationally representative cross-sectional surveys (1989, 1999-2000, 2011-12; n=8313, 5903 & 3904). Participants were defined as having obesity using classifications for an obese BMI (≥30kg·m-2) and substantially-increased-risk WC (≥88cm [women], ≥102cm [men]). Age-standardised prevalence of obesity according to BMI and/or WC, and the proportion of these detected by BMI and by WC were compared across surveys. FINDINGS: Between 1989 and 2011-12, weight and WC increased by 5.4kg and 10.7cm (women), and by 7.0kg and 7.3cm (men). For women and men, 63% and 38% of increases in WC were independent of increases in weight. Over this period, the prevalence of obesity according to BMI and/or WC increased by 25.3 percentage-points for women (18.9% to 44.3%) and 21.1 percentage-points for men (17.1% to 38.2%). The proportion of these detected by BMI decreased for women by 20 percentage-points (77% to 57%) with no change for men. The proportion of these detected by WC increased for women and men by 10 percentage-points (87% to 97%) and 6 percentage-points (85% to 91%) respectively. CONCLUSION: BMI alone is detecting a decreasing proportion of those considered obese by BMI and/or WC. Renewed discussion regarding how we monitor obesity at the population level is required.

Associations between physical activity and the neighbourhood social environment: baseline results from the HABITAT multilevel study.

Limitations have arisen when measuring associations between the neighbourhood social environment and physical activity, including same-source bias, and the reliability of aggregated neighbourhood-level social environment measures. This study examines cross-sectional associations between the neighbourhood social environment (perceptions of incivilities, crime, and social cohesion) and self-reported physical activity, while accounting for same-source bias and reliability of neighbourhood-level exposure measures, using data from a large population-based clustered sample. This investigation included 11,035 residents aged 40-65years from 200 neighbourhoods in Brisbane, Australia, in 2007. Respondents self-reported their physical activity and perceptions of the social environment (neighbourhood incivilities, crime and safety, and social cohesion). Models were adjusted for individual-level education, occupation, and household income, and neighbourhood disadvantage. Exposure measures were generated via split clusters and an empirical Bayes estimation procedure. Data were analysed in 2016 using multilevel multinomial logistic regression. Residents of neighbourhoods with the highest incivilities and crime, and lowest social cohesion were reference categories. Individuals were more likely to be in the higher physical activity categories if they were in neighbourhoods with the lowest incivilities and the lowest crime. No associations were found between social cohesion and physical activity. This study provides a basis from which to gain a clearer understanding of the relationship between the neighbourhood social environment and individual physical activity. Further work is required to explore the pathways between perceptions of the neighbourhood social environment and physical activity.

The impact of a tax on sugar-sweetened beverages according to socio-economic position: a systematic review of the evidence.

OBJECTIVE: A tax on sugar-sweetened beverages (SSB) has been proposed to address population weight gain but the effect across socio-economic position (SEP) is unclear. The current study aimed to clarify the differential impact(s) of SSB taxes on beverage purchases and consumption, weight outcomes and the amount paid in SSB taxes according to SEP. DESIGN: Databases (OVID and EMBASE) and grey literature were systematically searched in June 2015 to identify studies that examined effects of an SSB price increase on beverage purchases or consumption, weight outcomes or the amount paid in tax across SEP, within high-income countries. RESULTS: Of the eleven included articles, three study types were identified: (i) those that examined the association between variation in SSB taxes and SSB consumption and/or body weight (n 3); (ii) price elasticity estimation of SSB demand (n 1); and (iii) modelling of hypothetical SSB taxes by combining price elasticity estimates with population SEP-specific beverage consumption, energy intake or body weight (n 7). Few studies statistically tested differences in outcomes between SEP groups. Nevertheless, of the seven studies that reported on changes in weight outcomes for the total population following an increase in SSB price, all reported either similar reductions in weight across SEP groups or greater reductions for lower compared with higher SEP groups. All studies that examined the average household amount paid in tax (n 5) reported that an SSB tax would be regressive, but with small differences between higher- and lower-income households (0·10-1·0 % and 0·03 %-0·60 % of annual household income paid in SSB tax for low- and high-income households, respectively). CONCLUSIONS: Based on the available evidence, a tax on SSB will deliver similar population weight benefits across socio-economic strata or greater benefits for lower SEP groups. An SSB tax is shown to be consistently financially regressive, but to a small degree.

Neighborhood disadvantage, individual-level socioeconomic position and physical function: A cross-sectional multilevel analysis.

INTRODUCTION: Understanding associations between physical function and neighborhood disadvantage may provide insights into which interventions might best contribute to reducing socioeconomic inequalities in health. This study examines associations between neighborhood-disadvantage, individual-level socioeconomic position (SEP) and physical function from a multilevel perspective. METHODS: Data were obtained from the HABITAT multilevel longitudinal (2007-13) study of middle-aged adults, using data from the fourth wave (2013). This investigation included 6004 residents (age 46-71years) of 535 neighborhoods in Brisbane, Australia. Physical function was measured using the PF-10 (0-100), with higher scores indicating better function. The data were analyzed using multilevel linear regression and were extended to test for cross-level interactions by including interaction terms for different combinations of SEP (education, occupation, household income) and neighborhood disadvantage on physical function. RESULTS: Residents of the most disadvantaged neighborhoods reported significantly lower physical function (men: ß -11.36 95% CI -13.74, -8.99; women: ß -11.41 95% CI -13.60, -9.22). These associations remained after adjustment for individual-level SEP. Individuals with no post-school education, those permanently unable to work, and members of the lowest household income had significantly poorer physical function. Cross-level interactions suggested that the relationship between household income and physical function is different across levels of neighborhood disadvantage for men; and for education and occupation for women. CONCLUSION: Living in a disadvantaged neighborhood was negatively associated with physical function after adjustment for individual-level SEP. These results may assist in the development of policy-relevant targeted interventions to delay the rate of physical function decline at a community-level.

Bisphenol A and the risk of cardiometabolic disorders: a systematic review with meta-analysis of the epidemiological evidence.

Bisphenol A (BPA) is suspected to be associated with several chronic metabolic diseases. The aim of the present study was to review the epidemiological literature on the relation between BPA exposure and the risk of cardiometabolic disorders. PubMed and Embase databases were searched up to August 2014 by two independent investigators using standardized subject terms. We included observational studies (cohort, case-control and cross-sectional studies) carried out in children or adults, measuring urinary BPA (uBPA), including at least 100 participants and published in English. The health outcomes of interest were diabetes, hyperglycemia, measures of anthropometry, cardiovascular disease (CVD) and hypertension. Data were extracted and meta-analyzed when feasible, using a random-effects model. Thirty-three studies with sample size ranging from 239 to 4811 met the inclusion criteria, including five with a prospective design. Twelve studies reported on diabetes or hyperglycemia, 16 on anthropometry, 6 on CVD and 3 on hypertension. Evidence for a positive association between uBPA concentrations and diabetes, overweight, obesity, elevated waist circumference (WC), CVD and hypertension was found in 7/8, 2/7, 6/7, 5/5, 4/5 and 2/3 of the cross-sectional studies, respectively. We were able to conduct outcome-specific meta-analyses including 12 studies. When comparing the highest vs. the lowest uBPA concentrations, the pooled ORs were 1.47 (95% CI: 1.21-1.80) for diabetes, 1.21 (95% CI: 0.98-1.50) for overweight, 1.67 (95% CI: 1.41-1.98) for obesity, 1.48 (95% CI: 1.25-1.76) for elevated WC, and 1.41 (95% CI: 1.12-1.79) for hypertension. Moreover, among the five prospective studies, 3 reported significant findings, relating BPA exposure to incident diabetes, incident coronary artery disease, and weight gain. To conclude, there is evidence from the large body of cross-sectional studies that individuals with higher uBPA concentrations are more likely to suffer from diabetes, general/abdominal obesity and hypertension than those with lower uBPA concentrations. Given the potential importance for public health, prospective cohort studies with proper adjustment for dietary characteristics and identification of critical windows of exposure are urgently needed to further improve knowledge about potential causal links between BPA exposure and the development of chronic disease.

The validity of self-reported cancer in an Australian population study.

OBJECTIVE: The aim of this study is to determine the validity of self-reported cancer data by comparing it to the Australian Cancer Database (ACD). METHODS: Self-reported data were obtained from the Australian Diabetes, Obesity and Lifestyle (AusDiab) study, which were then linked to the ACD up until 31 December 2010. Positive predictive value, negative predictive value, sensitivity and specificity were calculated. Cohen's kappa coefficient (ĸ) was also calculated to assess the agreement between self-reported cancer and the ACD. Logistic regression was used to examine the determinants associated with false negative and false positive reporting. RESULTS: The overall sensitivity of self-report cancer was 71.1%, and sensitivities showed great variation by cancer site. Higher sensitivities were observed for breast (90.7%), bowel (77.8%) and prostate (77.1%) cancers, whereas the lowest sensitivity was observed for melanoma of the skin (36.9%). Similarly, the kappa coefficient analysis showed substantial agreement for self-reported breast cancer (ĸ=0.79) and moderate agreement for melanoma (ĸ=0.45) against the ACD. Years since cancer diagnosis and older age were associated with false negative reporting and older age was associated with false positive reporting. CONCLUSIONS AND IMPLICATIONS: The use of self-reported cancer to collect cancer outcomes has varying reliability, depending on cancer type and population. The findings presented here may assist medical researchers in making informed decisions when conducting research using self-reported cancer data in Australia where the acquisition of registry data is not feasible.