Trend Estimation Method Research Articles

Extraction of unadjusted estimates of prognostic association for meta-analysis: simulation methods as good alternatives to trend and direct method estimation

ObjectivesSystematic reviews and meta-analysis are the standard methods to assess the association between prognostic markers and major events/conditions. However, the summary measures reported are not always explicitly presented and therefore different indirect methods of extracting estimates have been proposed. The aim of this study is to present two new alternative methods for obtaining summary statistics to be included in a meta-analysis of prognostic studies based on simulating individual patient data and to compare them with the already known generalized least squares for trend (glst) estimation method and direct method. Study Design and SettingsWe have checked the performance of these methods using a between study comparison, including 122 studies, and a within study comparison, based on data from one of the studies. ResultsThe results obtained in this study show that glst estimation method appears to overestimate the effect size when reported information is incomplete. For the within-study comparison, the closest approximation to the direct estimates was obtained using the approach based on simulating individual patient data. ConclusionThe proposed simulation methods are a good alternative when other well-known indirect methods cannot be used.

Fibroblast growth factor 23 as a predictor of cardiovascular and all-cause mortality in prospective studies

Background and aimsThe prognostic value of fibroblast growth factor 23 (FGF23) for mortality remains controversial. We performed a meta-analysis of cohort studies to examine the controversial relationship between FGF23 and mortality. MethodsPubMed, EMBASE, the Cochrane Library databases and reference bibliographies were searched through September 2016 to identify prospective cohort studies with relative risks (RRs) and 95% confidence intervals (CIs) for FGF23 and mortality. A random effects model was used to pool the risk estimates. A dose-response analysis of the risk for all-cause mortality associated with FGF23 was conducted using the generalized least squares trend estimation method. ResultsNineteen prospective cohort studies were eligible for inclusion in this meta-analysis, of which 16 reported all-cause mortality and 9 reported cardiovascular mortality. During the follow-up periods ranging from 1 to 18.6 years, 5606 deaths occurred among 22,805 participants and 2458 cardiovascular deaths occurred among 28,845 participants. Elevated FGF23 was associated with an increased risk of all-cause mortality (RR 1.68; 95% CI 1.48–1.92) and cardiovascular mortality (RR 1.68; 95% CI 1.38–2.04) with moderate heterogeneity. These associations were not markedly modified by the geographic location, follow-up length, patient predisposition, FGF23 measurement or study quality. A sensitivity analysis yielded a similar effect on the pooled risk estimate. Evidence of a nonlinear relationship between FGF23 and all-cause mortality was observed in the dose-response analysis, with the risk gradually increasing as FGF23 increased. ConclusionsThis meta-analysis showed that individuals with increased plasma FGF23 levels might suffer a higher risk of all-cause mortality and cardiovascular mortality.

Multi-sliding time windows based changing trend of mean temperature and its association with the global-warming hiatus

Based on three global annual mean surface temperature time series and three Chinese annual mean surface air temperature time series, climate change trends on multiple timescales are analyzed by using the trend estimation method of multi-sliding time windows. The results are used to discuss the so-called global-warming hiatus during 1998–2012. It is demonstrated that different beginning and end times have an obvious effect on the results of the trend estimation, and the implications are particularly large when using a short window. The global-warming hiatus during 1998–2012 is the result of viewing temperature series on short timescales; and the events similar to it, or the events with even cold tendencies, have actually occurred many times in history. Therefore, the global-warming hiatus is likely to be a periodical feature of the long-term temperature change. It mainly reflects the decadal variability of temperature, and such a phenomenon in the short term does not alter the overall warming trend in the long term.

Damage evaluation by a guided wave-hidden Markov model based method

Guided wave based structural health monitoring has shown great potential in aerospace applications. However, one of the key challenges of practical engineering applications is the accurate interpretation of the guided wave signals under time-varying environmental and operational conditions. This paper presents a guided wave-hidden Markov model based method to improve the damage evaluation reliability of real aircraft structures under time-varying conditions. In the proposed approach, an HMM based unweighted moving average trend estimation method, which can capture the trend of damage propagation from the posterior probability obtained by HMM modeling is used to achieve a probabilistic evaluation of the structural damage. To validate the developed method, experiments are performed on a hole-edge crack specimen under fatigue loading condition and a real aircraft wing spar under changing structural boundary conditions. Experimental results show the advantage of the proposed method.

On trend estimation and significance testing for non-Gaussian and serially dependent data: quantifying the urbanization effect on trends in hot extremes in the megacity of Shanghai

Quantifying the urbanization effect on trends in climate extremes is important both for detection and attribution studies and for human adaptation; however, a fundamental problem is how to accurately estimate a trend and its statistical significance, especially for non-Gaussian and serially dependent data. In this paper, the choice of trend estimation and significance testing method is suggested as important for these kinds of studies, as illustrated by quantifying the urbanization effect on trends in seven hot-extreme indices for the megacity of Shanghai during 1961–2013. Both linear and nonlinear trend estimation methods were used. The trends and corresponding statistical significances were estimated by taking into account potential non-Gaussian and serial dependence in the extreme indices. A new method based on adaptive surrogate data is proposed to test the statistical significance of the ensemble empirical mode decomposition (EEMD) nonlinear trend. The urbanization contribution was found to be approximately 34 % (43 %) for the trend in the non-Gaussian distributed heat wave index based on nonparametric linear trend (EEMD nonlinear trend) estimation. For some of the other six hot-extreme indices analyzed, the urbanization contributions estimated based on linear and nonlinear trends varied greatly, with as much as a twofold difference between them. For the linear trend estimation itself, the ordinary least squares fit can give a substantially biased estimation of the urbanization contribution for some of the non-Gaussian extreme indices.

Effectiveness of Trend Estimation Methods

This paper analyses methods of trend estimation focusing on a process of calculation of their effectiveness. Shown a procedure that helps to choose : optimal parameters of trend estimation methods for trends with known characteristics (steepness of increasing or decreasing of trends, a frequency of alternation of regions with positive and negative derivatives etc.). All calculation are shown for data that represents near Earth's asteroids (NEA) semi%major axes distributio ns. The method is based on running of multiple iterations (in order to provide static stability of results) of a procedure that was designed to find a test signal after adding investigated trend and noise. It allows to calculate a probability of a signal detection after trend removing with using trend estimation methods with different parameters.

Comparison of forward versus backward walking using body weight supported treadmill training in an individual with a spinal cord injury: A single subject design

Purpose/Hypothesis: Body weight supported treadmill training (BWSTT) is a task-specific intervention that promotes functional locomotion. There is no research evaluating the effect of backward walking (BW) using BWSTT in individuals with spinal cord injury (SCI). The purpose of this single subject design was to examine the differences between forward walking (FW) and BW training using BWSTT in an individual with quadriparesis. Methods: The participant was a 57-year-old male with incomplete C3–C6 SCI. An ABABAB design (A = BW; B = FW; each phase = 3 weeks of biweekly sessions) was utilized. Outcome measures included: gait parameters; a timed 4-meter walk; the 5-repetition sit-to-stand test (STST); tandem stance time; and 6-minute walk test (6MWT). Data was analyzed with split level method of trend estimation. Results: Improvements in gait parameters, on the timed 4-meter walk, 6MWT, tandem balance and aerobic endurance were similar with FW and BW training. The only difference between FW and BW training was that BW training resulted in greater improvements in the STST. Conclusions: The results of this study suggest that in this individual backward walking training was advantageous, resulting in improved ability to perform the 5-repetition STST. It is suspected that these changes can be attributed to the differences in muscle activation and task difficulty between FW and BW.

Trend Change Detection in NDVI Time Series: Effects of Inter-Annual Variability and Methodology

Changing trends in ecosystem productivity can be quantified using satellite observations of Normalized Difference Vegetation Index (NDVI). However, the estimation of trends from NDVI time series differs substantially depending on analyzed satellite dataset, the corresponding spatiotemporal resolution, and the applied statistical method. Here we compare the performance of a wide range of trend estimation methods and demonstrate that performance decreases with increasing inter-annual variability in the NDVI time series. Trend slope estimates based on annual aggregated time series or based on a seasonal-trend model show better performances than methods that remove the seasonal cycle of the time series. A breakpoint detection analysis reveals that an overestimation of breakpoints in NDVI trends can result in wrong or even opposite trend estimates. Based on our results, we give practical recommendations for the application of trend methods on long-term NDVI time series. Particularly, we apply and compare different methods on NDVI time series in Alaska, where both greening and browning trends have been previously observed. Here, the multi-method uncertainty of NDVI trends is quantified through the application of the different trend estimation methods. Our results indicate that greening NDVI trends in Alaska are more spatially and temporally prevalent than browning trends. We also show that detected breakpoints in NDVI trends tend to coincide with large fires. Overall, our analyses demonstrate that seasonal trend methods need to be improved against inter-annual variability to quantify changing trends in ecosystem productivity with higher accuracy.

Dairy Consumption and Incidence of Hypertension

Observational and clinical studies suggest that dairy intake, particularly low-fat dairy, could have a beneficial effect on blood pressure. We performed a dose-response meta-analysis of prospective cohort studies on dairy intake and risk of hypertension in the general population. A systematic literature search for eligible studies was conducted until July 2011, using literature databases and hand search. Study-specific dose-response associations were computed according to the generalized least squares for trend estimation method, and linear and piecewise regression models were created. Random-effects models were performed with summarized dose-response data. We included 9 studies with a sample size of 57 256, a total of 15 367 incident hypertension cases, and a follow-up time between 2 and 15 years. Total dairy (9 studies; range of intake, ≈100–700 g/d), low-fat dairy (6 studies; ≈100–500 g/d), and milk (7 studies; ≈100–500 g/d) were inversely and linearly associated with a lower risk of hypertension. The pooled relative risks per 200 g/d were 0.97 (95% CI, 0.95–0.99) for total dairy, 0.96 (95% CI, 0.93–0.99) for low-fat dairy, and 0.96 (95% CI, 0.94–0.98) for milk. High-fat dairy (6 studies), total fermented dairy (4 studies), yogurt (5 studies), and cheese (8 studies) were not significantly associated with hypertension incidence (pooled relative risks of ≈1). This meta-analysis of prospective cohort studies suggests that low-fat dairy and milk could contribute to the prevention of hypertension, which needs confirmation in randomized controlled trials.

Construction, Management, and Performance of Sparse Markowitz Portfolios

We study different implementations of the sparse portfolio construction and rebalancing method introduced by Brodie et al. This technique is based on the use of a l1-norm (sum of the absolute values) type penalization on the portfolio weights vector that regularizes the Markowitz portfolio selection problem by automatically eliminating the dynamical redundancies present in the time evolution of asset prices. We make specific recommendations as to the different estimation techniques for the parameters needed in the use of the method and we prove its good performance in realistic situations involving different rebalancing frequencies and transaction costs. Our empirical findings show that the beneficial effects of the use of sparsity constraints are robust with respect to the choice of trend and covariance estimation methods used in its implementation.

Analysis of Temperature Trends and Spatial Variability in Huaibei Plain

Based on monthly data about temperature from 1957 to 2007 in Huaibei plain. Temperature variation trends and spatial differences were investigated by linear trend estimation method. Results show that annual mean temperature has increased about 1.2°C, with 0.25°C per ten years, which is especially obvious after the 1990s; Seasonal increasing trends are significant except summer, with increments of 1.6, 1.4 and 1.7°C in spring, autumn, and winter, about 0.33, 0.26, and 0.36°C per ten years, respectively; Warming rate from large to small was winter, spring and autumn. Suzhou and Bengbu are the most affected areas, which is warming obviously, increase temperature range is bigger; Dangshan, Bozhou, and Fuyang is the second one, warming of smaller amplitude.

Evolution of extreme temperatures in Belgium since the 1950s

The objective of the present study is to apply a wide range of efficient trend estimation methods for understanding how temperature extremes are locally changing. Temporal patterns of changes in extreme daily maximum or minimum temperature at homogeneous climate stations located in Belgium and their associations with changes in climate means are examined for the period 1952/1953 until present. A considerable amount of work is devoted to the formulation of extreme value models in the presence of non-stationarity. The covariate process is considered to be linear in time or/and in the North Atlantic Oscillation index as well. Additional insights on historical changes in frequency and amplitude of temperature extremes are obtained with the non-parametric quantile-perturbation approach.

Industrial business cycle and the cyclical fluctuations of sold industrial output according to PAC

The first aim of the paper is to identify business cycle in the Polish industry and in departments of the industry according to PAC. The second aim is the re-examination of the co-variability between the calculated fluctuations. When identifying the fluctuations, first one has to purify the time-series of incidental and seasonal fluctuations. According to that the time-series underwent adjustment procedure of seasonal correction Census X11. This way Henderson’s curves which reflect trends and seasonal fluctuations at the same time were obtained. The next step was to state the values of trends. To establish these values the method of stochastic trend estimation, which is called Hodrick-Prescott’s filter was used. Dividing empirical values of Henderson’s curves by the implementation arising from a use of HP we get (after multiplying by 100) a series showing percentage deviations form the trend line, which means seasonal fluctuations.The plan of the paper is as follows: introduction, section 1 definition of business cycle and data, section 2 with econometric methodology, section 3 with empirical analysis, and the conclusions.

Upward trend of streamflow and precipitation in a small, non-snow-fed, mountainous watershed in Nepal

Summary Hydroclimatic trend assessments of one of the rural, agricultural, non-snow-fed watersheds – the Jhikhu Khola Watershed (JKW) – in Nepal using a non-parametric trend estimation method (Mann–Kendall, Spearman’s rho, and modified Mann–Kendall for autocorrelated data) showed that annual average, maximum, and minimum flows are increasing. This trend of increasing flow is also seen in the annual average flow measured by one of the river gauging stations in the river basin that the watershed drains into, and is corroborated by Seasonal Kendall Test as well. This increase in streamflow coincides with the increasing rainfall trend of the yearly monsoon (June–September) and pre-monsoon (March–May) periods. The post-monsoon (October–February) period does not show any statistical trend. No consistent trend is observed in temperature changes for the whole watershed. Apparently there is no paradox in terms of scale (watershed and basin) and consistency between rainfall and runoff trends in this watershed. However, lack of a consistent trend in temperature changes in the study area may indicate a greater effect from local climatic control than global change on temperature. The lack of shift in the peak month of streamflow in this non-snow-fed watershed may also indicate that there are no major changes in the arrival of monsoon patterns in the study area.

A better way to estimate population trends

Estimation of a population trend from a time series of abundance data is an important task in ecology, yet such estimation remains logistically and conceptually challenging in practice. First, the extent to which unequal intervals in the time series, due to missing observations or irregular sampling, compromise trend estimation is not well‐known. Furthermore, the predominant trend estimation method (loglinear regression of abundance data against time) ignores the possibility of process noise, while an alternative method (the ‘diffusion approximation’) ignores observation error in the abundance data. State‐space models that account for both process noise and observation error exist but have been little used. We study an adaptation of the exponential growth state‐space (EGSS) model for use with missing data in the time series, and we compare its trend estimation to the status quo methods. The EGSS model provides superior estimates of trend across wide ranges of time series length and sources of variation. The performance of the EGSS model even with half of the counts in the time series missing implies that trend estimates may be improved by diverting effort away from annual monitoring and towards increasing time series length or improving precision of the abundance estimates for years that data are collected.

Efficacy of an Emotion-Focused Treatment for Prolonged Fatigue

Previous research findings have suggested a relationship between less adaptive emotional functioning and fatigue. The present study used a research design involving multiple baselines across participants to evaluate the efficacy of a new emotion-focused treatment for prolonged fatigue delivered in a cognitive behavioral therapy framework. The 13 adults participating in the study met the criteria for prolonged fatigue and provided fatigue baselines of 2, 5, or 8 weeks. The results indicated that the treatment was effective, with fatigue severity levels after the initiation of treatment significantly lower than that predicted by baseline patterns, as determined by the split median method of trend estimation. At 3-4 months after treatment, 8 of 11 clients who completed the treatment no longer met the criteria for prolonged fatigue.

Indicators of the Cyclically Adjusted Budget Balance: The Bank of Finland's Experience

Cyclically adjusted government budget balances have become increasingly popular as a means of analysing the fiscal situation and changes in policy that result from the intentional actions of the government. As the actual budget balances are affected both by cyclical factors ('automatic stabilisers') and structural ('discretionary') measures, they may not, in general, be very useful when seeking to assess the orientation of underlying fiscal policy and possible structural imbalances in the budget balance. The problem however is that there is no generally accepted method of calculating cyclically adjusted (structural) budget balances. The results tend to be fairly noisy and sensitive to the method of calculation. The purpose of this article is to highlight these issues by reviewing three estimation methods used by the Bank of Finland - GDP smoothing based on the Hodrick-Prescott trend estimation method, the production function approach and Blanchard's method - and the corresponding estimates of the cyclically adjusted budget balances for Finland.

A cross‐taxa phenological dataset from Mohonk Lake, NY and its relationship to climate

AbstractWe present a detailed analysis of a rare cross‐taxa native species phenology dataset (plant flowering, insect first sighting, and amphibian first sighting) from Mohonk Lake, NY. This dataset is highly unusual in North America for its longevity of record, consistency of methodology and location, diversity of species available, and availability of local daily meteorological data. For each phenology series, we examined the flowering and first sighting Julian calendar dates for the existence of temporal trends. Only one of the five animal species (katydid) showed any evidence for a significant trend in first sighting. In contrast, the plant species showed a rich mixture of temporal trends in flowering that could be divided into four classes: woody plant–no trend, woody plant–negative trend, herbaceous plant–negative trend, and herbaceous plant–positive trend. Many of the trends were found to be statistically significant and robust to the method of trend estimation. The data within each of the four plant classes were also pooled as anomalies to provide more complete temporal coverage for tests of trend robustness. The results were strongly consistent with the trends of the individual species within each class and highly significant statistically. We next correlated the flowering and first sighting dates against growing degree‐day (GDD) summations for each day of the year to measure the sensitivity of each species to this common form of climatic forcing on phenology. All species showed a significant sensitivity to GDD summations, with peak correlations falling on or near the varying median flowering or first sighting dates. These results were robust whether the GDD analyses were conducted over the complete set of observations for each species (beginning on or after 1928) or over a latter period with a more serially complete set of cross‐taxa observations (1970–2002). The GDD correlations indicate significant climate sensitivity in all species, but the different magnitudes of correlation and timing of maximum correlation imply that plant and animal responses to climate changes in the future will not be homogeneous for the tested species at Mohonk Lake. Copyright © 2007 Royal Meteorological Society

Statistical significance of trends and trend differences in layer‐average atmospheric temperature time series

This paper examines trend uncertainties in layer‐average free atmosphere temperatures arising from the use of different trend estimation methods. It also considers statistical issues that arise in assessing the significance of individual trends and of trend differences between data sets. Possible causes of these trends are not addressed. We use data from satellite and radiosonde measurements and from two reanalysis projects. To facilitate intercomparison, we compute from reanalyses and radiosonde data temperatures equivalent to those from the satellite‐based Microwave Sounding Unit (MSU). We compare linear trends based on minimization of absolute deviations (LA) and minimization of squared deviations (LS). Differences are generally less than 0.05°C/decade over 1959–1996. Over 1979–1993, they exceed 0.10°C/decade for lower tropospheric time series and 0.15°C/decade for the lower stratosphere. Trend fitting by the LA method can degrade the lower‐tropospheric trend agreement of 0.03°C/decade (over 1979–1996) previously reported for the MSU and radiosonde data. In assessing trend significance we employ two methods to account for temporal autocorrelation effects. With our preferred method, virtually none of the individual 1979–1993 trends in deep‐layer temperatures are significantly different from zero. To examine trend differences between data sets we compute 95% confidence intervals for individual trends and show that these overlap for almost all data sets considered. Confidence intervals for lower‐tropospheric trends encompass both zero and the model‐projected trends due to anthropogenic effects. We also test the significance of a trend in d(t), the time series of differences between a pair of data sets. Use of d(t) removes variability common to both time series and facilitates identification of small trend differences. This more discerning test reveals that roughly 30% of the data set comparisons have significant differences in lower‐tropospheric trends, primarily related to differences in measurement system. Our study gives empirical estimates of statistical uncertainties in recent atmospheric temperature trends. These estimates and the simple significance testing framework used here facilitate the interpretation of previous temperature trend comparisons involving satellite, radiosonde, and reanalysis data sets.

Statistical Analysis of Single-Subject Designs

To the Editor: The application of statistical analysis to single-subject designs is controversial and not straightforward. For this reason, we read the article by Nourbakhsh and Ottenbacher1 in the August 1994 issue of Physical Therapy with interest. The purposes of the study were to examine whether the use of different statistical analysis methods would lead to different statistical conclusions and to identify factors of graphed data that might have effected any observed difference in the statistical conclusions. The three methods studied were the C statistic,2 the two-standard deviation band method,3 and the split-middle method of trend estimation.4 To investigate the first purpose, the method of investigation was to analyze data from 42 AB graphs and to compare the statistical conclusions. The authors' methodological procedures for comparing the three techniques for agreement in statistical conclusions were limited and not entirely appropriate, for the following reasons:… [ ARTICLE][1] [1]: /lookup/volpage/74/768?iss=8