A simple imputation method for longitudinal studies with non-ignorable non-responses

Missing data are a common problem in longitudinal studies in the health sciences. Motivated by data from the Muscatine Coronary Risk Factor (MCRF) study, a longitudinal study of obesity, we propose a simple imputation method for handling non-ignorable non-responses (i.e., when non-response is related to the specific values that should have been obtained) in longitudinal studies with either discrete or continuous outcomes. In the proposed approach, two regression models are specified; one for the marginal mean of the response, the other for the conditional mean of the response given non-response patterns. Statistical inference for the model parameters is based on the generalized estimating equations (GEE) approach. An appealing feature of the proposed method is that it can be readily implemented using existing, widely-available statistical software. The method is illustrated using longitudinal data on obesity from the MCRF study.     

The effect of low survey response rates on estimates of alcohol consumption in a general population survey

Background

Response rates for surveys of alcohol use are declining for all modes of administration (postal, telephone, face-to-face). Low response rates may result in estimates that are biased by selective non-response. We examined non-response bias in the NZ GENACIS survey, a postal survey of a random electoral roll sample, with a response rate of 49.5% (n = 1924). Our aim was to estimate the magnitude of non-response bias in estimating the prevalence of current drinking and heavy episodic (binge) drinking.

Methods

We used the “continuum of resistance” model to guide the investigation. In this model the likelihood of response by sample members is related to the amount of effort required from the researchers to elicit a response. First, the demographic characteristics of respondents and non-respondents were compared. Second, respondents who returned their questionnaire before the first reminder (early), before the second reminder (intermediate) or after the second reminder (late) were compared by demographic characteristics, 12-month prevalence of drinking and prevalence of binge drinking.

Results

Demographic characteristics and prevalence of binge drinking were significantly different between late respondents and early/intermediate respondents, with the demographics of early and intermediate respondents being similar to people who refused to participate while late respondents were similar to all other non-respondents. Assuming non-respondents who did not actively refuse to participate had the same drinking patterns as late respondents, the prevalence of binge drinking amongst current drinkers was underestimated. Adjusting the prevalence of binge drinkers amongst current drinkers using population weights showed that this method of adjustment still resulted in an underestimate of the prevalence.

Conclusions

The findings suggest non-respondents who did not actively refuse to participate are likely to have similar or more extreme drinking behaviours than late respondents, and that surveys of health compromising behaviours such as alcohol use are likely to underestimate the prevalence of these behaviours.     

Genetic model for longitudinal studies of aging, health, and longevity and its potential application to incomplete data

Many longitudinal studies of aging collect genetic information only for a sub-sample of participants of the study. These data also do not include recent findings, new ideas and methodological concepts developed by distinct groups of researchers. The formal statistical analyses of genetic data ignore this additional information and therefore cannot utilize the entire research potential of the data. In this paper, we present a stochastic model for studying such longitudinal data in joint analyses of genetic and non-genetic sub-samples. The model incorporates several major concepts of aging known to date and usually studied independently. These include age-specific physiological norms, allostasis and allostatic load, stochasticity, and decline in stress resistance and adaptive capacity with age. The approach allows for studying all these concepts in their mutual connection, even if respective mechanisms are not directly measured in data (which is typical for longitudinal data available to date). The model takes into account dependence of longitudinal indices and hazard rates on genetic markers and permits evaluation of all these characteristics for carriers of different alleles (genotypes) to address questions concerning genetic influence on aging-related characteristics. The method is based on extracting genetic information from the entire sample of longitudinal data consisting of genetic and non-genetic sub-samples. Thus it results in a substantial increase in the accuracy of statistical estimates of genetic parameters compared to methods that use only information from a genetic sub-sample. Such an increase is achieved without collecting additional genetic data. Simulation studies illustrate the increase in the accuracy in different scenarios for datasets structurally similar to the Framingham Heart Study. Possible applications of the model and its further generalizations are discussed.     

Biomarkers of aging in women and the rate of longitudinal changes

The purposes of this study were (1) to estimate biological age score (BAS) in Japanese healthy women based on the 4-7 years longitudinal data for physiological, hematological and biochemical examinations and (2) to examine the rate of aging changes in adult women based on the estimated BAS. The samples consisted of cross-sectional (n=981) and longitudinal (n=110) groups. Out of 31 variables examined, five variables (forced expiratory volume in 1.0 s, systolic blood pressure, mean corpuscular hemoglobin, glucose, albumin/globulin ratio) that met the following criteria: 1) significant cross-sectional correlation with age; 2) significant longitudinal change in the same direction as the cross-sectional correlation; and (3) assessment of redundancy, were selected as candidate biomarkers of aging. This variable set was then submitted into a principal component analysis, and the first principal component obtained from this analysis was used as an equation for assessing one's BAS. Individual BAS showed a high longitudinal stability of age-related changes, suggesting high predictive validity of our newly developed aging measurement equation. However, changes in the aging rate based on the estimated BAS were not constant. The mean slopes of the regression lines of BAS for the three age groups (age<45, 45相似文献   

Precision and Disclosure in Text and Voice Interviews on Smartphones

As people increasingly communicate via asynchronous non-spoken modes on mobile devices, particularly text messaging (e.g., SMS), longstanding assumptions and practices of social measurement via telephone survey interviewing are being challenged. In the study reported here, 634 people who had agreed to participate in an interview on their iPhone were randomly assigned to answer 32 questions from US social surveys via text messaging or speech, administered either by a human interviewer or by an automated interviewing system. 10 interviewers from the University of Michigan Survey Research Center administered voice and text interviews; automated systems launched parallel text and voice interviews at the same time as the human interviews were launched. The key question was how the interview mode affected the quality of the response data, in particular the precision of numerical answers (how many were not rounded), variation in answers to multiple questions with the same response scale (differentiation), and disclosure of socially undesirable information. Texting led to higher quality data—fewer rounded numerical answers, more differentiated answers to a battery of questions, and more disclosure of sensitive information—than voice interviews, both with human and automated interviewers. Text respondents also reported a strong preference for future interviews by text. The findings suggest that people interviewed on mobile devices at a time and place that is convenient for them, even when they are multitasking, can give more trustworthy and accurate answers than those in more traditional spoken interviews. The findings also suggest that answers from text interviews, when aggregated across a sample, can tell a different story about a population than answers from voice interviews, potentially altering the policy implications from a survey.     

Happy aged people are all alike, while every unhappy aged person is unhappy in its own way

Aging of the world's population represents one of the most remarkable success stories of medicine and of humankind, but it is also a source of various challenges. The aim of the collaborative cross-cultural European study of adult well being (ESAW) is to frame the concept of aging successfully within a causal model that embraces physical health and functional status, cognitive efficacy, material security, social support resources, and life activity. Within the framework of this project, we show here that the degree of heterogeneity among people who view aging in a positive light is significantly lower than the degree of heterogeneity of those who hold a negative perception of aging. We base this conclusion on our analysis of a survey involving 12,478 people aged 50 to 90 from six West European countries. We treat the survey database as a bipartite network in which individual respondents are linked to the actual answers they provide. Taking this perspective allows us to construct a projected network of respondents in which each link indicates a statistically validated similarity of answers profile between the connected respondents, and to identify clusters of individuals independently of demographics. We show that mental and physical well-being are key factors determining a positive perception of aging. We further observe that psychological aspects, like self-esteem and resilience, and the nationality of respondents are relevant aspects to discriminate among participants who indicate positive perception of aging.     

Nutrition as a determinant of functional autonomy and quality of life in aging: a research program

With the aging of the Canadian population, functional autonomy and quality of life among seniors are now important public health issues. We hypothesized that nutrition is an important determinant of the quality of aging because of its potential to modulate the transitions from vulnerability to frailty and dependence. Over the past 15 years, our research program addressed the prevalence, the determinants, and the consequences of undernutrition among seniors, especially the free-living frail elderly. Very low energy and nutrient intakes were observed as well as a high prevalence of involuntary weight loss. These chronic conditions were associated with early institutionalization and increased mortality rates. Intervention strategies were then developed and evaluated, including the Nutrition Screening Program and the Nutrition Support Program. The effectiveness of these programs was shown with respect to improvement of nutritional status. However, this improvement was not sufficient to produce significant changes in functional autonomy or quality of life. Methodological issues related to the conduct of intervention studies in this specific population were addressed. A conceptual framework of nutritional intervention is currently being validated. A large longitudinal study that is being undertaken will further contribute to our understanding of the aging process as determined by a modifiable factor such as nutrition.     

Conservation of physiological dysregulation signatures of aging across primates

Two major goals in the current biology of aging are to identify general mechanisms underlying the aging process and to explain species differences in aging. Recent research in humans suggests that one important driver of aging is dysregulation, the progressive loss of homeostasis in complex biological networks. Yet, there is a lack of comparative data for this hypothesis, and we do not know whether dysregulation is widely associated with aging or how well signals of homeostasis are conserved. To address this knowledge gap, we use unusually detailed longitudinal biomarker data from 10 species of nonhuman primates housed in research centers and data from two human populations to test the hypotheses that (a) greater dysregulation is associated with aging across primates and (b) physiological states characterizing homeostasis are conserved across primates to degrees associated with phylogenetic proximity. To evaluate dysregulation, we employed a multivariate distance measure, calculated from sets of biomarkers, that is associated with aging and mortality in human populations. Dysregulation scores positively correlated with age and risk of mortality in most nonhuman primates studied, and signals of homeostatic state were significantly conserved across species, declining with phylogenetic distance. Our study provides the first broad demonstration of physiological dysregulation associated with aging and mortality risk in multiple nonhuman primates. Our results also imply that emergent signals of homeostasis are evolutionarily conserved, although with notable variation among species, and suggest promising directions for future comparative studies on dysregulation and the aging process.     

The use of interviews with surrogate respondents in a case-control study of oral cancer

The purpose of this study was to assess the possible bias that may occur in case-control studies when exposure information is not collected from all potentially eligible cases. The data used in this study were collected in the metropolitan area of Atlanta as part of a multicenter, population-based, case-control study of oropharyngeal cancer. In-person interviews were conducted with 112 cases (67.9 percent) and information on an additional 23 ill or deceased cases (13.9 percent) was collected through surrogate respondents. The cases about whom information was collected from surrogate respondents had more advanced disease at the time of diagnosis and were more likely to be black and less educated than cases who were interviewed in person. Cigarette smoking and consumption of hard liquor were more common among the cases about whom information was collected through surrogates. Therefore, failure to include such information would have resulted in underestimates of the strength of association between these exposures and the risk of oropharyngeal cancer.     

Processes Linking Religious Involvement and Telomere Length

Although numerous studies suggest that religious involvement is associated with better health and longer life expectancies, it is unclear whether these general patterns extend to cellular aging. The mechanisms linking indicators of religious involvement with indicators of cellular aging are also undefined. We employed longitudinal data from the 2004 and 2008 Health and Retirement Study, a national probability sample of Americans aged 50 and older, to test whether average telomere length varied according to level of religious attendance. We also tested several potential mechanisms. Our results showed that respondents who attended religious services more frequently in 2004 also exhibited fewer stressful events, lower rates of smoking, fewer symptoms of depression, and lower levels of C-reactive protein in 2008. Respondents who increased their level of attendance from 2004 to 2008 also exhibited lower rates of smoking in 2008. Although religious attendance was not directly associated with telomere length, our mediation analyses revealed significant indirect effects through depression and smoking, but not stressful events or C-reactive protein. We conclude that religious attendance may promote telomere length indirectly by reducing symptoms of depression and the risk of smoking. There was no evidence to support stressful events or C-reactive protein as mechanisms of religious attendance.     

Environmental Management for Malaria Control: Knowledge and Practices in Mvomero,Tanzania

Environmental conditions play an important role in the transmission of malaria; therefore, regulating these conditions can help to reduce disease burden. Environmental management practices for disease control can be implemented at the community level to complement other malaria control methods. This study assesses current knowledge and practices related to mosquito ecology and environmental management for malaria control in a rural, agricultural region of Tanzania. Household surveys were conducted with 408 randomly selected respondents from 10 villages and qualitative data were collected through focus group discussions and in-depth interviews. Results show that respondents are well aware of the links between mosquitoes, the environment, and malaria. Most respondents stated that cleaning the environment around the home, clearing vegetation around the home, or draining stagnant water can reduce mosquito populations, and 63% of respondents reported performing at least one of these techniques to protect themselves from malaria. It is clear that many respondents believe that these environmental management practices are effective malaria control methods, but the actual efficacy of these techniques for controlling populations of vectors or reducing malaria prevalence in the varying ecological habitats in Mvomero is unknown. Further research should be conducted to determine the effects of different environmental management practices on both mosquito populations and malaria transmission in this region, and increased participation in effective techniques should be promoted.     

北京某综合医院门诊患者购药行为影响因素调查

?????? 目的 了解患者购药行为及其影响因素。方法 对501名门诊患者进行问卷调查,了解其就诊购药行为,对调查数据进行描述性分析。结果 就诊后97.08%的调查对象选择在医院药房购药,73.33%的患者没有外出购药的意识,认为医院药品质量比零售药店药品好的有43.54%,认为药店药品更便宜的占55.83%。结论 患者的购药习惯、认识水平,医院和药品厂商对处方流动的人为限制,以及现有医疗保险制度不利于处方流动。     

Multi-day Longitudinal Assessment of Physical Activity and Sleep Behavior Among Healthy Young and Older Adults Using Wearable Sensors

ObjectivesThe number of elderly people is growing rapidly and aging is found to affect activities of daily living. Older adults are found to perform less physical activity when compared to younger ones. In the perspective of movement behavior, it is not well understood how are elderly different from younger ones. It is not known whether they produce only low frequency movement accelerations or the overall number of movements produced are reduced in elderly. It is also not known how elderly and younger ones perform movement transitions throughout the duration of a day and during night-time sleep.Material and methodsIn this study, 10 healthy young and 10 healthy old participants wore inertial measurement unit at their lower back for 3-days. The 24 hours of day were divided into four 6 hour time zones and transitions made by young and elderly were investigated. All participants performed their regular daily activities unhindered and longitudinal multi-day signals for acceleration and angular velocity were analyzed. Time-frequency analysis was performed using wavelet transform and frequency content of each movement performed was computed.ResultsWe found that both young and older adults performed significantly more low amplitude movements than medium and high amplitude movements. Healthy young adults produced significantly more movements at 1.1 Hz than older adults. Healthy young adults were also found to have produced significantly smaller number of transitions in the mid-phases of sleep. They were also found to produce significantly larger accelerations during night-time sleep transitions than their older counterparts.ConclusionThe advantages of collecting longitudinal data about human movement and sleep transition data can lead us to important clinical diagnosis. The information from longitudinal assessment can help develop lifestyle interventions for disease prevention, monitoring of chronic diseases to prevent or slow disease progression among elderly people.     

Inbred mouse strains reveal biomarkers that are pro‐longevity,antilongevity or role switching

Traditionally, biomarkers of aging are classified as either pro‐longevity or antilongevity. Using longitudinal data sets from the large‐scale inbred mouse strain study at the Jackson Laboratory Nathan Shock Center, we describe a protocol to identify two kinds of biomarkers: those with prognostic implication for lifespan and those with longitudinal evidence. Our protocol also identifies biomarkers for which, at first sight, there is conflicting evidence. Conflict resolution is possible by postulating a role switch. In these cases, high biomarker values are, for example, antilongevity in early life and pro‐longevity in later life. Role‐switching biomarkers correspond to features that must, for example, be minimized early, but maximized later, for optimal longevity. The clear‐cut pro‐longevity biomarkers we found reflect anti‐inflammatory, anti‐immunosenescent or anti‐anaemic mechanisms, whereas clear‐cut antilongevity biomarkers reflect inflammatory mechanisms. Many highly significant blood biomarkers relate to immune system features, indicating a shift from adaptive to innate processes, whereas most role‐switching biomarkers relate to blood serum features and whole‐body phenotypes. Our biomarker classification approach is applicable to any combination of longitudinal studies with life expectancy data, and it provides insights beyond a simplified scheme of biomarkers for long or short lifespan.     

Psychological functioning and aging. Learning from longitudinal studies

Population-based longitudinal studies are very useful tools for investigating the characteristics of the aging process among different geographical and socio-cultural contexts. On the other hand, these kind of projects look at a wide range of key issues of aging. In this review, most recent psychological findings in this field are analysed. The increasing interest in longitudinal studies for psychological variables is shown, as well as explicit links with other areas of study. Finally, main areas of improvement are identified according to future analysis of psychological variables, regarding both conceptual and methodological issues.     

Women’s Perceptions and Misperceptions of Male Circumcision: A Mixed Methods Study in Zambia

Women’s perceptions of male circumcision (MC) have implications for behavioral risk compensation, demand, and the impact of MC programs on women’s health. This mixed methods study combines data from the first two rounds of a longitudinal study (n = 934) and in-depth interviews with a subsample of respondents (n = 45) between rounds. Most women correctly reported that MC reduces men’s risk of HIV (64% R1, 82% R2). However, 30% of women at R1, and significantly more (41%) at R2, incorrectly believed MC is fully protective for men against HIV. Women also greatly overestimated the protection MC offers against STIs. The proportion of women who believed MC reduces a woman’s HIV risk if she has sex with a man who is circumcised increased significantly (50% to 70%). Qualitative data elaborate women’s misperception regarding MC. Programs should address women’s informational needs and continue to emphasize that condoms remain critical, regardless of male partner’s circumcision status.     

Gift card incentives and non-response bias in a survey of vaccine providers: the role of geographic and demographic factors

This study investigates the effects of non-response bias in a 2010 postal survey assessing experiences with H1N1 influenza vaccine administration among a diverse sample of providers (N = 765) in Washington state. Though we garnered a high response rate (80.9%) by using evidence-based survey design elements, including intensive follow-up and a gift card incentive from Target, non-response bias could exist if there were differences between respondents and non-respondents. We investigated differences between the two groups for seven variables: road distance to the nearest Target store, practice type, previous administration of vaccines, region, urbanicity, size of practice, and Vaccines for Children (VFC) program enrollment. We also examined the effect of non-response bias on survey estimates. Statistically significant differences between respondents and non-respondents were found for four variables: miles to the nearest Target store, type of medical practice, whether the practice routinely administered additional vaccines besides H1N1, and urbanicity. Practices were more likely to respond if they were from a small town or rural area (OR = 7.68, 95% CI = 1.44−40.88), were a non-traditional vaccine provider type (OR = 2.08, 95% CI = 1.06−4.08) or a pediatric provider type (OR = 4.03, 95% CI = 1.36−11.96), or administered additional vaccines besides H1N1 (OR = 1.80, 95% CI = 1.03−3.15). Of particular interest, for each ten mile increase in road distance from the nearest Target store, the likelihood of provider response decreased (OR = 0.73, 95% CI = 0.60−0.89). Of those variables associated with response, only small town or rural practice location was associated with a survey estimate of interest, suggesting that non-response bias had a minimal effect on survey estimates. These findings show that gift card incentives alongside survey design elements and follow-up can achieve high response rates. However, there is evidence that practices farther from the nearest place to redeem gift cards may be less likely to respond to the survey.     

Predictors of non-response to a questionnaire survey of a volunteer twin panel: findings from the Australian 1989 twin cohort.

Questionnaire surveys, while more economical, typically achieve poorer response rates than interview surveys. We used data from a national volunteer cohort of young adult twins, who were scheduled for assessment by questionnaire in 1989 and by interview in 1996-2000, to identify predictors of questionnaire non-response. Out of a total of 8536 twins, 5058 completed the questionnaire survey (59% response rate), and 6255 completed a telephone interview survey conducted a decade later (73% response rate). Multinomial logit models were fitted to the interview data to identify socioeconomic, psychiatric and health behavior correlates of non-response in the earlier questionnaire survey. Male gender, education below University level, and being a dizygotic rather than monozygotic twin, all predicted reduced likelihood of participating in the questionnaire survey. Associations between questionnaire response status and psychiatric history and health behavior variables were modest, with history of alcohol dependence and childhood conduct disorder predicting decreased probability of returning a questionnaire, and history of smoking and heavy drinking more weakly associated with non-response. Body-mass index showed no association with questionnaire non-response. Despite a poor response rate to the self-report questionnaire survey, we found only limited sampling biases for most variables. While not appropriate for studies where socioeconomic variables are critical, it appears that survey by questionnaire, with questionnaire administration by telephone to non-responders, will represent a viable strategy for gene-mapping studies requiring that large numbers of relatives be screened.