Attrition In Longitudinal Studies Research Articles

Abstract P3-12-12: ATTITUDE: Understanding and reducing ATTrition in longiTUDinal studiEs of cancer survivors

Abstract Background: Reduction in response rates is common in longitudinal studies, generating significant attrition bias. ATTITUDE is a comprehensive project built on the CANcer TOxicity (CANTO) cohort, using a mixed-methodology, multi-step approach aimed at reducing attrition: Step 1- understand patterns of attrition; Step 2- understand patients’ experiences in providing data; and Step 3- design and pilot a tailored intervention to reduce attrition. Here we present the results of Step 1. Methods: CANTO (NCT01993498) is a longitudinal cohort study that includes patients with stage I-III breast cancer from 26 cancer centers. CANTO collects detailed clinical and patient-reported outcome (PRO) evaluations at diagnosis of breast cancer, and at year-1, year-2 and year-4 post-diagnosis. Concurrent enrollment in other investigational studies is allowed. Our primary outcome of interest was “non-response”, defined as not responding to pre-planned CANTO evaluations, prematurely exiting the study, withdrawing consent or being lost to follow-up. Patients with a valid reason for study termination, including fatal or recurrence events, were excluded. Multivariable logistic regression with bootstrapped Augmented Backwards Elimination retained factors associated with non-response. Results: Among 5903 patients enrolled in CANTO from 2012-2015 with mature follow-up at year-4 post-diagnosis, mean age was 56.7 years (SD 11.4), 51.5% had stage I breast cancer, and 27.2% were concomitantly enrolled in other investigational studies. The rate of non-response to clinical and PRO evaluations was 0% and 5.7% at diagnosis, 12.3% and 21.2% at year-2, and 20.1% and 43.1% at year-4, respectively. Factors associated with non-response at year-4 included study entry period, CANTO center size, smoke, alcohol consumption behavior, and symptom burden at inclusion. Patients included in larger patient volume centers, with higher socio-economic status at diagnosis and those that co-enrolled in other studies had reduced odds of non-response (Table). Conclusions: In this large cohort of patients with early-stage breast cancer, attrition at year-4 post-diagnosis was substantial, particularly for PRO evaluations, and it was affected by center- and patient-related factors. Results of the Step 1 of ATTITUDE will help us identify patients with higher risk of non-response and inform targeted interventions to minimize attrition, including better risk stratification and personalization of the follow-up during participation in longitudinal research studies. These interventions will be developed and tested during subsequent steps of ATTITUDE. Table. Factors associated with non-response at year-4 by type of evaluation; Adjusted* Odds Ratio (95% CI)Clinical evaluationsPRO evaluationsStudy entry period2012-2013Ref.Ref.2014-20150.74 (0.62 - 0.87)1.77 (1.57 - 2.02)Center size (patients enrolled in CANTO)Small (<100)Ref.Ref.Medium (100-1000)0.54 (0.41 - 0.73)0.27 (0.21 - 0.35)Large (≥1000)0.44 (0.32 - 0.61)0.33 (0.25 - 0.44)Age at diagnosis5-year increase1.08 (1.03 - 1.13)0.98 (0.94 - 1.01)EducationPrimary schoolRef.Not retainedHigh school0.69 (0.55 - 0.88)College or higher degree0.65 (0.49 - 0.86)Monthly household income (Euro)<1500Ref.Ref.≥1500 and <30000.89 (0.71 - 1.13)0.85 (0.71 - 1.03)≥30000.75 (0.58 - 0.97)0.73 (0.60 - 0.88)Tobacco behaviorNever smokerRef.Ref.Current smoker1.72 (1.37 - 2.14)1.39 (1.16 - 1.66)Alcohol consumption frequencyLess than dailyRef.Ref.Daily1.21 (0.96 - 1.52)1.23 (1.02 - 1.47)Fatigue**10-point increase1.05 (1.02 - 1.08)1.04 (1.01 - 1.08)Pain**10-point increaseNot retained1.02 (0.99 - 1.08)Co-enrollment in other studies,NoRef.Ref.Yes0.67 (0.55 - 0.81)0.79 (0.68 - 0.91)*by marital status, physical activity, comorbidities, psychological distress (Anxiety and Depression; HADS), breast cancer stage, and treatment-related factors**EORTC QLQ-C30, assessed at study entry (diagnosis) Citation Format: Pietro Lapidari, Julie Havas, Daniele Presti, Davide Soldato, Elise Martin, Barbara Pistilli, Aurelie Bertaut, Diane Boinon, Guillemette Jacob, Anne-Laure Martin, Sibille Everhard, Alexandra Jacquet, Gwenn Menvielle, Matteo Lambertini, Laurence Vanlemmens, Stefan Michiels, Fabrice André, Ines Vaz-Luis, Antonio Di Meglio. ATTITUDE: Understanding and reducing ATTrition in longiTUDinal studiEs of cancer survivors [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P3-12-12.

Validation and generalizability of machine learning prediction models on attrition in longitudinal studies

Attrition in longitudinal studies is a major threat to the representativeness of the data and the generalizability of the findings. Typical approaches to address systematic nonresponse are either expensive and unsatisfactory (e.g., oversampling) or rely on the unrealistic assumption of data missing at random (e.g., multiple imputation). Thus, models that effectively predict who most likely drops out in subsequent occasions might offer the opportunity to take countermeasures (e.g., incentives). With the current study, we introduce a longitudinal model validation approach and examine whether attrition in two nationally representative longitudinal panel studies can be predicted accurately. We compare the performance of a basic logistic regression model with a more flexible, data-driven machine learning algorithm—gradient boosting machines. Our results show almost no difference in accuracies for both modeling approaches, which contradicts claims of similar studies on survey attrition. Prediction models could not be generalized across surveys and were less accurate when tested at a later survey wave. We discuss the implications of these findings for survey retention, the use of complex machine learning algorithms, and give some recommendations to deal with study attrition.

Attrition in longitudinal studies among patients with schizophrenia and other psychoses; findings from the STRATA collaboration

A major problem with longitudinal studies is the bias generated due to attrition, particularly apparent amongst patients suffering from psychotic disorders. Factors associated with study-participation were investigated as part of a larger research collaboration (STRATA). Out of 479 eligible participants, only 50 (10,4%) were successfully followed up. The present study investigated whether study participation differed depending on baseline characteristics. Results indicated that individuals who did not participate were more likely to report an alcohol use disorder while those who did respond were more likely to have been in full-time education for longer and be of white ethnicity. Participation did not differ depending on diagnosis, symptoms, GAF, age of onset or depression.

Predictors of students' participation in a learning environment survey with annual follow-ups.

Longitudinal research is one effective way to gauge changes in a student cohort over time, however attrition in these studies is typically high, which can result in study bias. This study explored learning environment factors, approaches to studying, and academic performance as predictors of occupational therapy students' consistent participation in data collection conducted over three years of their professional program. A longitudinal study of Norwegian occupational therapy students (analyzed n = 240) was conducted. Logistic regression analysis was used to explore occupational therapy students' perceptions of the learning environment, their approaches to studying, and exam grades as they related to the likelihood of consistent participation at three annual surveys. Annual response rates varied between 55.1%, and 65.6%, and consistent participation was observed among 49.2%. The fully adjusted regression models showed that higher strategic approach scores increased the odds of consistent participation (adjusted OR: 1.04, p < 0.01), whereas higher surface approach scores decreased the odds of consistent participation (adjusted OR: 0.95, p < 0.05). Neither sociodemographic factors, learning environment factors nor academic performance predicted participation over time. Researchers can anticipate relatively high levels of attrition in longitudinal studies of occupational therapy students, but attrition seems to be largely proportional between groups. However, completers in longitudinal studies may be somewhat more well-organized and academically oriented than drop-outs.

Selective attrition in longitudinal studies: effective processes for Facebook tracing

ABSTRACT Participant attrition is a significant challenge for longitudinal studies, particularly those involving disadvantaged populations. We attempted to re-engage n = 990 families in a five-year follow-up of a randomised controlled trial of an early childhood parenting intervention. Tracing was attempted for n = 90/990 (9.1%) uncontactable participants. Evidence for the identification of a matching Facebook profile was classified as strong, moderate or weak. A private message was sent to those with strong evidence (n = 63/90, 70.0%), of which 32 responded, 22 agreed to participate and 19 provided data. Compared to participants engaged using traditional methods, those re-engaged through Facebook were more likely at initial recruitment to have a younger child (p =.02), to be a single parent (p =.04), less educated (p <.001), lower income (p =.01) and in an unemployed household (p <.001). Findings suggest that social media can be an effective engagement tool, helping to minimise overall and selective attrition in longitudinal studies, particularly for hard-to-reach populations.

Factors Associated with Attrition of Adult Participants in a Longitudinal Database: A National Institute on Disability, Independent Living, and Rehabilitation Research Burn Model System Study.

Participant attrition in longitudinal studies can lead to substantial bias in study results, especially when attrition is nonrandom. A previous study of the Burn Model System (BMS) database prior to 2002 identified participant and study-related factors related to attrition. The purpose of the current study was to examine changes in attrition rates in the BMS longitudinal database since 2002 and to revisit factors associated with attrition. Individuals 18 years and older enrolled in the BMS database between 2002 and 2018 were included in this study. Stepwise logistic regression models identified factors significantly associated with attrition at 6, 12, and 24 months postburn injury. The percentage of individuals lost to follow-up was 26% at 6 months, 33% at 12 months, and 42% at 24 months. Factors associated with increased risk of loss to follow-up across two or more time points include male sex, lower TBSA burn size, being unemployed at the time of burn, shorter duration of acute hospital stay, younger age, not having private health insurance or workers' compensation, and a history of drug abuse. Retention levels in the BMS have improved by at least 10% at all time points since 2002. The BMS and other longitudinal burn research projects can use these results to identify individuals at high risk for attrition who may require additional retention efforts. Results also indicate potential sources of bias in research projects utilizing the BMS database.

Selective attrition does not affect cross-sectional estimates of associations with emotional and behavioral problems in a longitudinal study with onset in adolescence

The present examination raises the question whether attrition in a longitudinal study leads to biased findings. The Zurich Adolescent Psychology and Psychopathology Study (ZAPPS) originated in 1994 by following a sample of 1239 adolescents with follow-ups after three, six, and 22 years. The study collected data on life-events, self-esteem, and coping behavior as independent and mental problems as dependent variables. The baseline sample was partitioned according to follow-up status for the three subsequent waves of assessment. Baseline measures of associations between independent and dependent variables were compared for those retained in the study (group A) and those lost to follow-up (group B) at each phase of cross-sectional data collection by use of multiple linear regression analyses. There were significant differences for some baseline independent variables between the two groups with small effect sizes. Males and migrants dropped out more frequently. The main findings indicated that the strength of the associations between independent and dependent variables at baseline in the total sample and in each of the two groups as defined by sample sizes at the three follow-ups was close to equal. Thus, one may conclude from the current study that attrition in longitudinal studies rarely affects cross-sectional estimates of associations.

Factors influencing attrition in 35 Alzheimer's Disease Centers across the USA: a longitudinal examination of the National Alzheimer's Coordinating Center's Uniform Data Set.

A lack of understanding of the causes of attrition in longitudinal studies of older adults may lead to higher attrition rates and bias longitudinal study results. In longitudinal epidemiological studies of Alzheimer's disease and related dementias, high rates of attrition may cause a systematic underestimation of dementia prevalence and skew the characterization of the disease. This can compromise the generalizability of the study results and any inferences based on the surviving sample may grossly misrepresent the importance of the risk factors for dementia. The National Institute on Aging outlined a National Strategy for Recruitment and Participation in Alzheimer's Disease Clinical Research to address this problem, providing evidence of the magnitude of this problem. To explore predictors of attrition, this study examined the National Alzheimer's Coordinating Center (NACC) Uniform Data Set, a repository of observations of older adults spanning 11years, using survival analysis. Four samples were examined: the full sample (n = 30,433), the alive subsample excluding those who died (n = 24,231), the MRI sample [participants with complete MRI data (n = 1104)], and the alive MRI subsample [participants with MRI data excluding those who died (n = 947)]. Worsening cognitive impairment, neuropsychiatric symptoms, and difficulty with functional activities predicted attrition, as did lower hippocampal volume in the MRI subsample. Questionable co-participant reliability and an informant other than a spouse also increased risk of attrition. Special considerations exist in recruiting and retaining older adults in longitudinal studies, and results of baseline psychological, functional, and cognitive functioning should be used to identify targeted retention strategies.

Sustaining Engagement in Longitudinal Research With Vulnerable Families: A Mixed-Methods Study of Attrition.

The aim of this mixed-methods study was to investigate attrition at the age 10-year follow-up in a study of vulnerable children and their families living with low income following a two-generation preschool program in Calgary, Alberta, Canada. Quantitative factors associated with attrition included: (a) food bank use; (b) unstable housing; (c) child welfare involvement; (d) unpartnered status; and (e) caregiver noncompletion of high school. Qualitative themes related to attrition included: (a) income and employment; (b) health; (c) unstable housing; (d) change of guardianship; (e) domestic violence; (f) work and time management challenges; and (g) negative caregiver-child relationships. Triangulation of quantitative and qualitative results occurred using Maslow's Hierarchy of Needs; families with unmet physiological, safety, belongingness and love needs, and esteem needs were more likely to attrite. Attrition in longitudinal studies with vulnerable families is complex, affected by frequently changing life circumstances, and struggles to access necessities of life. Strategies for retaining vulnerable families in longitudinal research are offered.

Testing the validity of telephone interviews to assess chronic pain in children and adolescents: A randomized cross-over trial.

Telephone surveys are intended to reduce attrition in longitudinal studies. For paediatric chronic pain patients, the comparability of pain-related information gathered using telephone interviews and postal surveys remain unknown. Furthermore, it remains unknown how social desirability may influence answers. To compare data from telephone interviews and postal surveys, a randomized cross-over design with two measure points 2weeks apart and four conditions (combinations of telephone interviews (T) and postal surveys (P): P-T, T-P, P-P, T-T) was conducted in a sample of N=323 paediatric chronic pain patients. In the inter-group comparison, pain-related information did not differ between telephone interviews and postal surveys except for the information on pain location (back and extremities). Agreement measures of the intra-group comparisons suggest substantial to excellent agreements for all items and did not differ between the groups. The internal consistency of a disability scale was excellent for both assessment modes; the number of missing values did not differ. Participation rate was higher for telephone interviews compared to the postal surveys. Across both time points, attrition was lowest for the groups without a switch in assessment mode compared to the groups with a switch in assessment mode. Except for pain-related school absence, no effect of social desirability occurred. Telephone interviews are a useful method to achieve a high response rate. Pain locations should be asked for separately and not in an open question when interviewing children and adolescents on the telephone. Telephone interviews are a good method to achieve a high response rate and obtain valid data in studies with paediatric chronic pain patients.

Analytical results in longitudinal studies depended on target of inference and assumed mechanism of attrition

ObjectivesTo compare methods for analysis of longitudinal studies with missing data due to participant dropout and follow-up truncated by death. Study Design and SettingWe analyzed physical functioning in an Australian longitudinal study of elderly women where the missing data mechanism could either be missing at random (MAR) or missing not at random (MNAR). We assumed either an immortal cohort where deceased participants are implicitly included after death or a mortal cohort where the target of inference is surviving participants at each survey wave. To illustrate the methods a covariate was included. Simulation was used to assess the effect of the assumptions. ResultsIgnoring attrition or restricting analysis to participants with complete follow up led to biased estimates. Linear mixed model was appropriate for an immortal cohort under MAR but not MNAR. Linear increment model and joint modeling of longitudinal outcome and time to death were the most robust to MNAR. For a mortal cohort, inverse probability weighting and multiple imputation could be used, but care is needed in specifying dropout and imputation models, respectively. ConclusionAppropriate analysis methodology to deal with attrition in longitudinal studies depends on the target of inference and the missing data mechanism.

Reducing Longitudinal Attrition Through Facebook

Attrition of subjects is a concern when designing longitudinal field studies. One way to potentially reduce attrition in longitudinal studies may be to use a social network. In this study, two groups are contrasted. One group of participants completed a survey online and was redirected to a Facebook page for the purposes of the study, and was then invited to become Facebook friends. The second group completed the survey using pen and paper. In the second phase of the study, 6 months later, hospitality managers who had become friends of the study on Facebook were contacted via Facebook, while those who completed the paper survey were contacted via their personal e-mail addresses. The Facebook group had a response rate of 51.4%, while the e-mail group had a response rate of 15.2%. Our results suggest that Facebook might be an effective means of minimizing attrition in longitudinal studies.

Using multiple imputation to deal with missing data and attrition in longitudinal studies with repeated measures of patient-reported outcomes.

ObjectiveMissing data is a ubiquitous problem in studies using patient-reported measures, decreasing sample sizes and causing possible bias. In longitudinal studies, special problems relate to attrition and death during follow-up. We describe a methodological approach for the use of multiple imputation (MI) to meet these challenges.MethodsIn a cohort of patients treated with percutaneous coronary intervention followed with use of repetitive questionnaires and information from national registers over 3 years, only 417 out of 1,726 patients had complete data on all measure points and covariates. We suggest strategies for use of MI and different methods for dealing with death along with sensitivity analysis of deviations from the assumption of missing at random, all with the use of standard statistical software. The Mental Component Summary from Short Form 12-item survey was used as an example.ConclusionIgnoring missing data may cause bias of unknown size and direction in longitudinal studies. We have illustrated that MI is a feasible method to try to deal with bias due to missing data in longitudinal studies, including attrition and nonresponse, and should be considered in combination with analysis of sensitivity in longitudinal studies. How to handle dropout due to death is still open for debate.

Attrition in a 30-year follow-up of a perinatal birth risk cohort: factors change with age.

Background. Attrition is a major cause of potential bias in longitudinal studies and clinical trials. Attrition rate above 20% raises concern of the reliability of the results. Few studies have looked at the factors behind attrition in follow-ups spanning decades.Methods. We analyzed attrition and associated factors of a 30-year follow-up cohort of subjects who were born with perinatal risks for neurodevelopmental disorders. Attrition rates were calculated at different stages of follow-up and differences between responders and non-responders were tested. To find combinations of variables influencing attrition and investigate their relative importance at birth, 5, 9, 16 and 30 years of follow-up we used the random forest classification.Results. Initial loss of potential participants was 13%. Attrition was 16% at five, 24% at nine, 35% at 16 and 46% at 30 years. The only group difference that emerged between responders and non-responders was in socioeconomic status (SES). The variables identified by random forest classification analysis were classified into Birth related, Development related and SES related. Variables from all these categories contributed to attrition, but SES related variables were less important than birth and development associated variables. Classification accuracy ranged between 0.74 and 0.96 depending on age.Discussion. Lower SES is linked to attrition in many studies. Our results point to the importance of the growth and development related factors in a longitudinal study. Parents’ decisions to participate depend on the characteristics of the child. The same association was also seen when the child, now grown up, decided to participate at 30 years. In addition, birth related medical variables are associated with the attrition still at the age of 30. Our results using a data mining approach suggest that attrition in longitudinal studies is influenced by complex interactions of a multitude of variables, which are not necessarily evident using other multivariate techniques.

Reducing bias due to systematic attrition in longitudinal studies: The benefits of multiple imputation

Most longitudinal studies are plagued by drop-out related to variables at earlier assessments (systematic attrition). Although systematic attrition is often analysed in longitudinal studies, surprisingly few researchers attempt to reduce biases due to systematic attrition, even though this is possible and nowadays technically easy. This is particularly true for studies of stability and the long-term prediction of developmental outcomes. We provide guidelines how to reduce biases in such cases particularly with multiple imputation. Following these guidelines does not require advanced statistical knowledge or special software. We illustrate these guidelines and the importance of reducing biases due to selective attrition with a 25-year longitudinal study on the long-term prediction of aggressiveness and delinquency.

The telomere lengthening conundrum—artifact or biology?

Recent longitudinal studies of age-dependent leukocyte telomere length (LTL) attrition have reported that variable proportions of individuals experience LTL lengthening. Often, LTL lengthening has been taken at face value, and authors have speculated about the biological causation of this finding. Based on empirical data and theoretical considerations, we show that regardless of the method used to measure telomere length (Southern blot or quantitative polymerase chain reaction-based methods), measurement error of telomere length and duration of follow-up explain almost entirely the absence of age-dependent LTL attrition in longitudinal studies. We find that LTL lengthening is far less frequent in studies with long follow-up periods and those that used a high-precision Southern blot method (as compared with quantitative polymerase chain reaction determination, which is associated with larger laboratory error). We conclude that the LTL lengthening observed in longitudinal studies is predominantly, if not entirely, an artifact of measurement error, which is exacerbated by short follow-up periods. We offer specific suggestions for design of longitudinal studies of LTL attrition to diminish this artifact.

Attrition in longitudinal randomized controlled trials: home visits make a difference

BackgroundParticipant attrition in longitudinal studies can introduce systematic bias, favoring participants who return for follow-up, and increase the likelihood that those with complications will be underestimated. Our aim was to examine the effectiveness of home follow-up (Home F/U) to complete the final study evaluation on potentially “lost” participants by: 1) evaluating the impact of including and excluding potentially “lost” participants (e.g., those who required Home F/U to complete the final evaluation) on the rates of study complications; 2) examining the relationship between timing and number of complications on the requirement for subsequent Home F/U; and 3) determining predictors of those who required Home F/U.MethodsWe used data from a randomized controlled trial (RCT) conducted from 1991–1994 among coronary artery bypass graft surgery patients that investigated the effect of High mean arterial pressure (MAP) (intervention) vs. Low MAP (control) during cardiopulmonary bypass on 5 complications: cardiac morbidity/mortality, neurologic morbidity/mortality, all-cause mortality, neurocognitive dysfunction and functional decline. We enhanced completion of the final 6-month evaluation using Home F/U.ResultsAmong 248 participants, 61 (25%) required Home F/U and the remaining 187 (75%) received Routine F/U. By employing Home F/U, we detected 11 additional complications at 6 months: 1 major neurologic complication, 6 cases of neurocognitive dysfunction and 4 cases of functional decline. Follow-up of 61 additional Home F/U participants enabled us to reach statistical significance on our main trial outcome. Specifically, the High MAP group had a significantly lower rate of the Combined Trial Outcome compared to the Low MAP group, 16.1% vs. 27.4% (p=0.032). In multivariate analysis, participants who were ≥ 75 years (OR=3.23, 95% CI 1.52-6.88, p=0.002) or on baseline diuretic therapy (OR=2.44, 95% CI 1.14-5.21, p=0.02) were more likely to require Home F/U. In addition, those in the Home F/U group were more likely to have sustained 2 or more complications (p=0.05).ConclusionsHome visits are an effective approach to reduce attrition and improve accuracy of study outcome reporting. Trial results may be influenced by this method of reducing attrition. Older participants, those with greater medical burden and those who sustain multiple complications are at higher risk for attrition.

Missing value imputation in longitudinal measures of alcohol consumption

Attrition in longitudinal studies can lead to biased results. The study is motivated by the unexpected observation that alcohol consumption decreased despite increased availability, which may be due to sample attrition of heavy drinkers. Several imputation methods have been proposed, but rarely compared in longitudinal studies of alcohol consumption. The imputation of consumption level measurements is computationally particularly challenging due to alcohol consumption being a semi‐continuous variable (dichotomous drinking status and continuous volume among drinkers), and the non‐normality of data in the continuous part. Data come from a longitudinal study in Denmark with four waves (2003–2006) and 1771 individuals at baseline. Five techniques for missing data are compared: Last value carried forward (LVCF) was used as a single, and Hotdeck, Heckman modelling, multivariate imputation by chained equations (MICE), and a Bayesian approach as multiple imputation methods. Predictive mean matching was used to account for non‐normality, where instead of imputing regression estimates, "real" observed values from similar cases are imputed. Methods were also compared by means of a simulated dataset. The simulation showed that the Bayesian approach yielded the most unbiased estimates for imputation. The finding of no increase in consumption levels despite a higher availability remained unaltered.

Longitudinal Changes In Telomere Length In Patients with Dyskeratosis Congenita

AbstractAbstract 2230Dyskeratosis congenita (DC) is an inherited bone marrow failure syndrome with a complex clinical phenotype, including dysplastic nails, lacy reticular pigmentation, and oral leukoplakia (the diagnostic triad). Numerous other physical abnormalities may be present, in addition to cytopenias due to bone marrow failure, and a high risk of leukemia or solid tumors. However, many patients have no physical findings at diagnosis. Patients with DC have very short telomeres, and approximately one-half have a mutation in one of six genes important in telomere biology. Telomere length in leukocyte subsets, measured by automated flow fluorescence in situ hybridization (flow-FISH), is both sensitive and specific for identifying individuals with DC. Telomeres consist of nucleotide repeats and a protein complex at chromosome ends that are critical in chromosomal stability which shorten during normal cell division. Cross-sectional studies of normal individuals suggest that telomere length decreases with age in a sigmoid pattern from birth to old age. In a cross-sectional analysis of 26 patients with DC, we previously observed that telomere length appeared to be stable or even to slightly increase with age (BP Alter et al, Blood 110:149, 2007). Similar results were shown in 23 different DC patients by others (M Bessler et al, FEBS Lett 2010 in press). We speculated that these data were influenced by early presentation (or recognition) of clinically more severe patients, while patients with similar telomere length who were clinically milder were identified at older ages. In this pilot study, we examined, for the first time, the longitudinal age-association of telomere attrition in nine patients with DC who were followed for five to seven years (currently 8 – 50 years of age). These include three patients with mutations in TERC, and two each with TINF2, TERT, and DKC1 mutations. When first studied, four had normal hematopoiesis, three moderate cytopenias, one was receiving androgens, and one was on transfusions. Subsequently, one with normal hematopoiesis developed mild thrombocytopenia, one who was on transfusions responded to androgens, and one with moderate aplastic anemia became severe. In all cases, telomere length decreased with age. In a linear regression model, the average annual decrease in telomere length in lymphocytes was 167 base pairs/year (bp/yr) + 104, similar to the rate in granulocytes, 159 + 92 bp/yr. According to the literature, the rate of telomere attrition in longitudinal studies in normal blood is ∼45-50 bp/yr, albeit by methods other than flow-FISH; the rate of telomere shortening appears to decrease with increasing age. The average patient Z-scores at the beginning of the study were -3.9 standard deviations below the median for age in healthy normal controls, and were -4.3 at the end, consistent with the impression that DC patient telomeres shorten somewhat more than expected from normal aging. These data support the hypothesis that the earlier cross-sectional results for patients with DC indeed were influenced by the cross-sectional rather than longitudinal nature of the data. The current longitudinal data suggest that telomere shortening could possibly be accelerated in patients with DC, but larger studies are required. Our results indicate that patients with DC have telomeres that are much shorter than normal for their age, and that over time they continue to shorten, consistent with DC being classified as a disorder of premature aging. Disclosures:Lansdorp: Repeat Diagnostics: Equity Ownership.

Letter by Patel Regarding Article, “A Primer in Longitudinal Data Analysis”

HomeCirculationVol. 120, No. 4Letter by Patel Regarding Article, “A Primer in Longitudinal Data Analysis” Free AccessLetterPDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessLetterPDF/EPUBLetter by Patel Regarding Article, “A Primer in Longitudinal Data Analysis” Chirag B. Patel, MSE Chirag B. PatelChirag B. Patel Department of Diagnostic and Interventional Imaging, University of Texas Medical School at Houston, Houston, Tex Search for more papers by this author Originally published28 Jul 2009https://doi.org/10.1161/CIRCULATIONAHA.108.836767Circulation. 2009;120:e25To the Editor:I read with great interest the article titled “A Primer in Longitudinal Data Analysis” by Fitzmaurice and Ravichandran.1 Focusing on longitudinal data, the authors highlighted the important chasm between advancements in statistical methods and the analysis of current biomedical studies. Furthermore, the pros and cons of 2 particular approaches (analysis of response profiles and linear mixed-effects models) were well explained through case examples of previous studies. However, with the exception of a passing mention of cited sources for further reading (references 7 and 9 in the original article), the authors did not explain another important approach for analyzing longitudinal data: generalized estimating equations (GEEs).2GEEs can be used to model correlated data from repeated measures over the course of a longitudinal study. With respect to the defining features of longitudinal studies explained by Fitzmaurice and Ravichandran (eg, covariance structure and balanced versus unbalanced designs), GEEs have been shown to be more robust when missing data, imputation techniques, and other factors are considered.3 Of particular interest to longitudinal clinical studies is the use of GEEs to identify the best correlation structure and subset of covariates for a given model. Seemingly conflicting reports on the inefficiency of GEEs compared with independence estimating equations4,5 can be explained by differences in the type of data and covariate and correlation structures under study. The use of more recent models, such as the conditional second-order GEE estimator,6 has yielded improved efficiency. Furthermore, GEE models can be implemented in the software packages discussed by Fitzmaurice and Ravichandran.Clinician investigators and article reviewers would benefit greatly from knowing which model for the analysis of longitudinal data (eg, analysis of response profiles, linear mixed-effects models, GEE, independence estimating equation, or conditional second-order GEE) is most apt given a study’s design, data structure, and other relevant factors. Moreover, and in line with Fitzmaurice and Ravichandran’s intent, such an understanding will lead to an improved interpretation of longitudinal study results. This would get us “closer to reality” in terms of understanding the true impact of devices, pharmaceuticals, and other interventions for improved care of patients with cardiovascular risk factors and disease.DisclosuresNone. References 1 Fitzmaurice GM, Ravichandran C. A primer in longitudinal data analysis. Circulation. 2008; 118: 2005–2010.LinkGoogle Scholar2 Zeger SL, Liang KY, Albert PS. Models for longitudinal data: a generalized estimating equation approach. Biometrics. 1988; 44: 1049–1060.CrossrefMedlineGoogle Scholar3 Twisk J, de Vente W. Attrition in longitudinal studies: how to deal with missing data. J Clin Epidemiol. 2002; 55: 329–337.CrossrefMedlineGoogle Scholar4 Sutradhar BC, Das K. On the efficiency of regression estimators in generalised linear models for longitudinal data. Biometrika. 1999; 86: 459–465.CrossrefGoogle Scholar5 Fitzmaurice GM. A caveat concerning independence estimating equations with multivariate binary data. Biometrics. 1995; 51: 309–317.CrossrefMedlineGoogle Scholar6 Vonesh EF, Wang H, Nie LDM. Conditional second-order generalized estimating equations for generalized linear and nonlinear mixed-effects models. J Amer Statistical Assoc. 2002; 97: 271–283.CrossrefGoogle Scholar Previous Back to top Next FiguresReferencesRelatedDetailsCited By Rahman S, Sullivan J, Barger L, St. Hilaire M, O’Brien C, Stone K, Phillips A, Klerman E, Qadri S, Wright K, Halbower A, Segar J, McGuire J, Vitiello M, de la Iglesia H, Poynter S, Yu P, Sanderson A, Zee P, Landrigan C, Czeisler C and Lockley S (2021) Extended Work Shifts and Neurobehavioral Performance in Resident-Physicians, Pediatrics, 10.1542/peds.2020-009936, 147:3, Online publication date: 1-Mar-2021. July 28, 2009Vol 120, Issue 4 Advertisement Article InformationMetrics https://doi.org/10.1161/CIRCULATIONAHA.108.836767PMID: 19635976 Originally publishedJuly 28, 2009 PDF download Advertisement SubjectsEpidemiology