Network meta-analysis in health psychology and behavioural medicine: a primer

Progress in the science and practice of health psychology depends on the systematic synthesis of quantitative psychological evidence. Meta-analyses of experimental studies have led to important advances in understanding health-related behaviour change interventions. Fundamental questions regarding such interventions have been systematically investigated through synthesising relevant experimental evidence using standard pairwise meta-analytic procedures that provide reliable estimates of the magnitude, homogeneity and potential biases in effects observed. However, these syntheses only provide information about whether particular types of interventions work better than a control condition or specific alternative approaches. To increase the impact of health psychology on health-related policy-making, evidence regarding the comparative efficacy of all relevant intervention approaches – which may include biomedical approaches – is necessary. With the development of network meta-analysis (NMA), such evidence can be synthesised, even when direct head-to-head trials do not exist. However, care must be taken in its application to ensure reliable estimates of the effect sizes between interventions are revealed. This review paper describes the potential importance of NMA to health psychology, how the technique works and important considerations for its appropriate application within health psychology.     

How effective are interventions in improving dietary behaviour in low- and middle-income countries? A systematic review and meta-analysis

Several interventions encouraging people to change their diet have been tested in low- and middle-income countries (LMICs) but these have not been meta-synthesised and it is not known which elements of these interventions contribute to their effectiveness. The current review addressed these issues. Randomised controlled trials of dietary interventions in LMICs were eligible and identified via eight publication databases. Elements of both the intervention and comparison groups (e.g., behaviour change techniques (BCTs), delivery mode), participant characteristics and risk of bias were coded. Random effects meta-analysis of 76 randomised controlled trials found, on average, small- to medium-sized but highly heterogeneous improvement in dietary behaviour following an intervention. Small and homogeneous improvements were found for BMI/weight, waist- and hip-circumference, with medium-sized, but heterogeneous, improvements in blood pressure and cholesterol. Although many BCTs have yet to be tested in this context, meta-regressions suggested some BCTs (action planning, self-monitoring of outcome(s) of behaviour; demonstration of behaviour) as well as individually randomised trials, adult- or hypertensive-samples and lack of blinding were associated with larger dietary behaviour effect sizes. Interventions to encourage people from LMICs to change their diet produce, on average, small-to-medium-sized effects. These effects may possibly be increased through the inclusion of specific BCTs and other study elements.     

Heat of supersaturation-limited amyloid burst directly monitored by isothermal titration calorimetry

Amyloid fibrils form in supersaturated solutions via a nucleation and growth mechanism. Although the structural features of amyloid fibrils have become increasingly clearer, knowledge on the thermodynamics of fibrillation is limited. Furthermore, protein aggregation is not a target of calorimetry, one of the most powerful approaches used to study proteins. Here, with β₂-microglobulin, a protein responsible for dialysis-related amyloidosis, we show direct heat measurements of the formation of amyloid fibrils using isothermal titration calorimetry (ITC). The spontaneous fibrillation after a lag phase was accompanied by exothermic heat. The thermodynamic parameters of fibrillation obtained under various protein concentrations and temperatures were consistent with the main-chain dominated structural model of fibrils, in which overall packing was less than that of the native structures. We also characterized the thermodynamics of amorphous aggregation, enabling the comparison of protein folding, amyloid fibrillation, and amorphous aggregation. These results indicate that ITC will become a promising approach for clarifying comprehensively the thermodynamics of protein folding and misfolding.Aggregation has often been an obstacle to studying the structure, function, and physical properties of proteins. However, a large number of aggregates associated with serious diseases, including Alzheimer’s, Parkinson, and prion diseases (1, 2) promoted the challenge of studying protein misfolding and aggregation. Researchers succeeded in distinguishing amyloid fibrils and oligomers from other amorphous aggregates and characterized the ordered structures present in amyloid fibrils or oligomers, which led to the development of the field of amyloid structural biology (3–8). These advances have been attributed to various methodologies that are also useful for studying the structural properties of globular proteins. Even X-ray crystallography has become a powerful approach for studying amyloid microcrystals (5) or oligomers (9). The atomic details of amyloid fibrils are becoming increasingly clearer, and a cross-β structure was shown to be the main structural component of fibrils (5, 6, 8). Although tightly packed core regions of amyloid fibrils have been reported, the overall structures were shown to be dominated by common cross-β structures, which supported the argument for the main-chain dominated architecture in contrast to the side-chain dominated architecture of globular native states (10–12).These structural studies have been complemented by a series of efforts to clarify the mechanism for the formation of amyloid fibrils (i.e., amyloid fibrillation). The presence of a long lag time in spontaneous fibrillation and rapid fibrillation by the addition of preformed fibrils represent a similarity with the supersaturation-limited crystallization of substances (13–18). We have revisited “supersaturation” and argued its critical role for amyloid fibrillation (17–19). The role of supersaturation in neurodegenerative diseases at the proteome level has been reported recently (20).However, calorimetry, one of the most powerful methods used to study the thermodynamic properties of globular proteins (21–24), has not played a significant role in understanding protein aggregation. The aggregation of proteins following heat denaturation as monitored by differential scanning calorimetry is an infamous example demonstrating how aggregation can prevent exact analyses (25, 26). To date, few studies have investigated protein aggregation including amyloid fibrils with calorimetry (27–32). Our previous study on the exothermic heat effects accompanying fibril growth was achieved by monitoring the seed-dependent elongation of fibrils formed by β₂-microglobulin (β2m), a protein responsible for dialysis-related amyloidosis, using isothermal titration calorimetry (ITC) (28).In the present study using β2m, we succeeded in characterizing the total heat of spontaneous fibrillation and amorphous aggregation. An analysis of the heat burst associated with fibrillation or amorphous aggregation under various temperatures clarified their thermodynamic properties. The results obtained enabled the calorimetric characterization of amyloid fibrils and amorphous aggregates relative to that of the native globular structures, which opens a new field for the calorimetric study of protein aggregates.     

Museum specimens reveal loss of pollen host plants as key factor driving wild bee decline in The Netherlands

Evidence for declining populations of both wild and managed bees has raised concern about a potential global pollination crisis. Strategies to mitigate bee loss generally aim to enhance floral resources. However, we do not really know whether loss of preferred floral resources is the key driver of bee decline because accurate assessment of host plant preferences is difficult, particularly for species that have become rare. Here we examine whether population trends of wild bees in The Netherlands can be explained by trends in host plants, and how this relates to other factors such as climate change. We determined host plant preference of bee species using pollen loads on specimens in entomological collections that were collected before the onset of their decline, and used atlas data to quantify population trends of bee species and their host plants. We show that decline of preferred host plant species was one of two main factors associated with bee decline. Bee body size, the other main factor, was negatively related to population trend, which, because larger bee species have larger pollen requirements than smaller species, may also point toward food limitation as a key factor driving wild bee loss. Diet breadth and other potential factors such as length of flight period or climate change sensitivity were not important in explaining twentieth century bee population trends. These results highlight the species-specific nature of wild bee decline and indicate that mitigation strategies will only be effective if they target the specific host plants of declining species.Pollinating insects such as bees play an essential role in the pollination of wild plants (1) and crops (2). However, reported population declines in both wild and managed bees (3–5) have raised concerns about loss of pollination services and triggered interest in identifying the underlying causes for bee decline (6). Land use change and agricultural intensification are major drivers of biodiversity loss in general (7, 8) and are considered the most important environmental drivers of loss of wild bee diversity in particular (6, 9). It is generally believed that these drivers affect bees, which depend on floral resources in both their larval and adult life stages, through repercussions on the availability of floral resources in contemporary anthropogenic landscapes (9–11), but, so far, scientific evidence that loss of floral resources is driving bee decline is lacking. Nevertheless, current strategies to mitigate bee decline focus primarily on enhancing floral resources (12). To prioritize and develop effective mitigation strategies, it is essential to identify the mechanisms underlying bee population trends and assess whether these are mediated by floral resources.Although bees as a group are declining, individual species show more variable responses, with some species declining sharply while others remain stable or even increase under current land use change and agricultural intensification (3, 4, 13). These differential responses can be used to disentangle the effects of floral resource availability from those of other potential factors affecting bee population trends. The proportion of the floral resources in contemporary anthropogenic landscapes that can be used for forage by a bee species depends on its diet breadth and host plant preference, and it may be expected that species that have declined have a narrower diet breadth and prefer host plants that have declined (14, 15). However, diet breadth and host plant preference of bee species is difficult to assess. Presently observed host plant use does not necessarily reflect actual preference, as preferred host plants may have gone locally extinct and bees that have declined may have become restricted in their food choice in their remaining habitats (15). In addition, if host plant use is measured for more individuals of abundant, widespread species than for rare ones, an apparent link between diet breadth and population trend may simply arise as a sampling artifact (16). Furthermore, the relationship between host plant use and population trend may be confounded by species’ rarity prior to the onset of major environmental changes (17), as rarity in itself increases susceptibility to stochastic events (18) and has been shown to be one of the most important factors predicting population decline in various taxa (19–21). Surprisingly, to our knowledge, none of the studies that have so far examined the relationship between diet breadth and/or host plant preference and bee population trends have taken species’ initial rarity into account (e.g., refs. 3, 4, 15, and 22). Other factors, such as body size (4, 23), phenology (4, 22), and sensitivity to climate change (4, 24, 25) may be associated with bee decline as well, and, to date, the relative importance of diet breadth and pollen host plant preference in explaining bee population trends remains unclear.Here we solve this problem by analyzing historical pollen preferences of wild bees (15). Bees are generally more selective in their choice of food plants when foraging for pollen (source of protein and minerals for both larvae and adults) than nectar (source of energy) (26, 27). Distributional changes in plant species from which pollen is collected therefore probably exerts a larger influence on bee populations than changes in nectar plants. We investigate whether and to what extent loss of preferred floral resources drives bee population trends in The Netherlands, one of the most human-modified and intensively farmed countries in the world. Over the course of the twentieth century, agriculture has intensified in The Netherlands (Fig. S1) and the area of seminatural habitat preferred by bees has diminished to only one-fifth of the area at the beginning of the twentieth century (Fig. S2). More than half of the bee species are currently on the national Red List (28). As such, this country is a particularly suitable study area to identify critical factors associated with bee population decline.We assessed pollen host plant use of bee species independently from their population trends by analyzing pollen loads on the bodies of bee specimens that were collected before 1950 (15), before the onset of agricultural intensification in The Netherlands. Altogether, our analysis included trend and trait data of 57 bee species in 10 genera and 4 subfamilies (Table S1). We calculated population trend indices for bee species and their host plants (period 1902–1949 vs. 1975–1999) using extensive national species distribution datasets (13, 29). Linear mixed models, with bee subfamily as a random factor to account for phylogeny, and a multimodel inference approach were used to examine the relationship between bee population trends and pollen host plant use, simultaneously taking into account differences in species’ rarity before the onset of agricultural intensification and other factors that have been proposed to explain bee population trends.     

Deoxygenation of the Baltic Sea during the last century

Deoxygenation is a global problem in coastal and open regions of the ocean, and has led to expanding areas of oxygen minimum zones and coastal hypoxia. The recent expansion of hypoxia in coastal ecosystems has been primarily attributed to global warming and enhanced nutrient input from land and atmosphere. The largest anthropogenically induced hypoxic area in the world is the Baltic Sea, where the relative importance of physical forcing versus eutrophication is still debated. We have analyzed water column oxygen and salinity profiles to reconstruct oxygen and stratification conditions over the last 115 y and compare the influence of both climate and anthropogenic forcing on hypoxia. We report a 10-fold increase of hypoxia in the Baltic Sea and show that this is primarily linked to increased inputs of nutrients from land, although increased respiration from higher temperatures during the last two decades has contributed to worsening oxygen conditions. Although shifts in climate and physical circulation are important factors modulating the extent of hypoxia, further nutrient reductions in the Baltic Sea will be necessary to reduce the ecosystems impacts of deoxygenation.Dead zones are hypoxic (low-oxygen) areas unable to support most marine life, and over the past 50 y they have spread rapidly in the open ocean (1) as well as in coastal ecosystems (2). Global warming is thought to be a major driver for these changes (3), although biogeochemical factors have also been recognized, especially in coastal marine ecosystems (4, 5). In the Baltic Sea, the present spread of hypoxia is the combined result of climate changes influencing deepwater oxygenation (6) and increased eutrophication (7, 8), resulting in a hypoxic area ranging between 12,000 and 70,000 km² with an average of 49,000 km² over the time period 1961–2000 (7). Here, we separate the effects of the two factors on oxygen conditions.Physical factors are an important consideration in whether an ecosystem will experience hypoxia. The Baltic Sea is naturally prone to hypoxia due to a restricted water exchange with the ocean and a long residence time above 30 y (9, 10). Saltier, denser water from the North Atlantic flows over a series of shallow sills in the Danish Straits to ventilate waters below the permanent halocline and are governed by meteorological-induced variations in sea levels (11), displaying variations at decadal scales (12, 13). The dense saltwater inflows bring new supplies of oxygen to bottom waters, but at the same time enhance stratification, creating larger bottom areas that experience hypoxia (14). In particular, the ventilation of the deeper waters is attributed to events of larger inflows of high-saline water (>17), termed Major Baltic Inflows (MBIs), that have been less frequent in the last three decades (6).Climate warming decreases oxygen solubility due to higher water temperature, increases stratification, and enhances respiration processes (15). Climate warming is likely to be accompanied by increased precipitation and inflows of freshwater and nutrients to coastal waters in many areas of the globe. Increasing nutrient inputs from land stimulates primary production and export of organic material to the deep waters, thereby disrupting the subtle natural balance between oxygen supply from physical processes and oxygen demand from consumption of organic material. However, the importance of decreasing oxygenation versus increasing nutrient inputs for explaining the recent spread of hypoxia is not known (6, 7).Water column measurements of dissolved oxygen concentrations began around 1900 with more regularly spaced measurements commencing in the 1960s (Fig. S1), allowing a more consistent assessment of the spatial extent of hypoxia (7, 14). The sparse temporal and spatial resolution of oxygen data before 1960 allowed only assessing hypoxia at specific locations (16) or specific years (17). To our knowledge, our study is the first to report basin-wide trends of stratification and oxygen conditions from 1898 to present, and here we will focus on the two basins that have perennial hypoxia—the Bornholm Basin and the Gotland Basin (Fig. S2). These two basins are connected via a channel with a sill depth of 60 m.     

Nonlinear permanent migration response to climatic variations but minimal response to disasters

We present a microlevel study to simultaneously investigate the effects of variations in temperature and precipitation along with sudden natural disasters to infer their relative influence on migration that is likely permanent. The study is made possible by the availability of household panel data from Indonesia with an exceptional tracking rate combined with frequent occurrence of natural disasters and significant climatic variations, thus providing a quasi-experiment to examine the influence of environment on migration. Using data on 7,185 households followed over 15 y, we analyze whole-household, province-to-province migration, which allows us to understand the effects of environmental factors on permanent moves that may differ from temporary migration. The results suggest that permanent migration is influenced by climatic variations, whereas episodic disasters tend to have much smaller or no impact on such migration. In particular, temperature has a nonlinear effect on migration such that above 25 °C, a rise in temperature is related to an increase in outmigration, potentially through its impact on economic conditions. We use these results to estimate the impact of projected temperature increases on future permanent migration. Though precipitation also has a similar nonlinear effect on migration, the effect is smaller than that of temperature, underscoring the importance of using an expanded set of climatic factors as predictors of migration. These findings on the minimal influence of natural disasters and precipitation on permanent moves supplement previous findings on the significant role of these variables in promoting temporary migration.Human migration has been identified as a potentially important response to climate change. Where climate change makes habitation in certain places less desirable or even impossible, people may respond by moving elsewhere. However, the idea that environmental change induces people to migrate remains a widely contested topic, especially given recent findings suggesting that environmental changes may also constrain movement (1–3). Historically, there has been a paucity of empirical demonstrations of environmental effects on population mobility, partly due to sparse data and partly because migration studies have tended to focus on further exploration of social and economic predictors of migration that have already been established as primary drivers. More recently however, new empirical approaches to exploring the relationship between migration and climate change have emerged.Based on a review of the existing literature (SI Text, Literature Review), there is conflicting evidence on the effects of climatic variations and natural disasters on migration, partly arising from the inability to distinguish permanent moves from temporary ones, especially in the case of macrolevel studies that analyze aggregate flows of people. Furthermore, the effects may vary significantly by distance of migration destination, which may also confound the overall effect of environmental factors on migration. Above all, most studies at the microlevel do not simultaneously examine the effects of both disasters and climatic changes on migration, and often use only one aspect of climate, generally variation in rainfall. However, precipitation and temperature are historically correlated and to infer an unbiased effect of either one on migration probability, both need to be included in the model (4).The current study therefore attempts to improve on the existing studies. This study is, to our knowledge, the first at a microlevel to simultaneously explore the effects of sudden natural disasters and climatic variations on permanent migration of the whole household. (Household migration can take several forms: migration of a single member or individual migration; migration of one or more members of the household or split household migration; and migration of whole household, which includes migration of the entire household along with the head of household.) In doing so, we test the effect of temperature along with precipitation on migration decision. We include a summary that allows a quick comparison of the methodology used in our study compared with previous studies in terms of the choice of environmental variables to predict migration (Table S1). Prior studies have primarily examined individual migration behavior, which may capture both temporary and permanent migration. In contrast, province-level migration of entire households, as we show (with the use of data that follows households over a period of 15 y), tends to be more permanent. (We hereafter use the term “permanent.”) To our knowledge, the only other studies that focus on household migration explore the mechanisms by which natural disasters can deter migration of the household or its relatives (5, 6). By studying migration behavior of the whole household, we are able to focus on permanent migration, and therefore directly test how sudden disasters along with variations in rainfall and temperature affect permanent and relatively longer distance (province-to-province) migration as opposed to temporary movement; this allows us to complement the findings in some existing studies that generally conclude that natural disasters and rainfall result in temporary and short-distance moves while providing new evidence on the temperature–migration link in the context of microlevel studies.To achieve these goals, we chose Indonesia as our study site because as the world’s largest archipelago situated in a tectonically active location, the country is highly exposed to both geologic and climatic hazards (SI Text, Background on Indonesia). In addition, as the world’s fourth most populous country with ∼40% of the labor force engaged in agriculture and more than 60% of the total population living in the coastal areas, the country is extremely vulnerable to the impacts of climate variations and extreme events (SI Text, Background on Indonesia). At the same time, the availability of household panel data with an exceptional tracking rate allows us to use a difference-in-differences approach to study the migration of households before and after disasters as well as establish a plausibly causal link between climatic variations and migration. We use the Indonesian Family Life Survey (IFLS), a household panel survey representative of ∼83% of the Indonesian population from 13 of the 27 provinces in 1993 (Fig. S1). The IFLS provides data on 7,185 original households, followed over a period of 15 y (Materials and Methods). The migration outcome we predict captures migration of households, which is likely permanent. We track whole-household migration for a 15-y period during which the original households are followed with a very high retention rate even after they migrate. Some 95% of the migrant households end up migrating only once and do not return to their original province during the entire 15-y period. Therefore, at least in the Indonesian case, the provincial level migration of households seems to be permanent (see SI Text, Internal Migration in Indonesia, for more on internal migration in Indonesia).The household panel data are supplemented with data on natural disasters taken from DesInventar database, which provides disaster-related data using information on disasters of different intensities based on preexisting official data, academic records, newspaper sources, and institutional reports. Different types of disasters may not all affect migration in the same way, thus cancelling out each other’s effects on migration. For our analysis, we therefore separately estimate the effects of each type of disaster. Furthermore, we use multiple measures to capture the different ways in which a disaster may affect the population, for three primary reasons. First, we believe that the intensity of disaster rather than simply the occurrence of disaster should more accurately predict its effect. Measures such as number of deaths, number injured, number of houses destroyed, and amount of financial loss (captured by monetary loss measured in Indonesian Rupiah) from each type of disaster aggregated at the province level are used to capture both the frequency as well as the intensity of disasters. Second, relying on a single measure of a disaster may not capture its overall effect because a certain event may result in large financial losses but inflict little physical harm. Third, using alternate measures of disaster allows one to compare and confirm results, and can be used as a robustness check (Materials and Methods).Finally, for measures of climatic variations, we construct estimates of average temperature and precipitation for each province during each observation interval (Materials and Methods), because temperature and rainfall variations together provide a more complete measure of the extent of climate variations that may affect migration (7).     

Porcelain laminate veneers: Clinical survey for evaluation of failure

ObjectiveTo investigate the association of the failure of porcelain laminate veneers with factors related to the patient, material, and operator.MethodsThis clinical survey involved 29 patients (19 women and 10 men) and their dentists, including undergraduate and postgraduate dental students and dental interns. Two questionnaires were distributed to collect information from participants. All patients were clinically examined. Criteria for failure of the porcelain laminate veneers included color change, cracking, fracture, and/or debonding.ResultsA total of 205 porcelain laminate veneers were evaluated. All of the restorations were fabricated from IPS e.max Press and cemented with Variolink Veneer (Ivoclar Vivadent, Schaan, Principality of Liechtenstein) or RelyX veneer cement (3M ESPE, St. Paul, MN, USA). The preparations were generally located in enamel (58.6%), and most veneers had an overlapped design (89.7%). Ten patients (34.48%) showed veneer failure, most often in terms of color change (60%). Overall, 82.8% of patients were satisfied with their restorations.ConclusionInsufficient clinical skills or operator experience resulted in restoration failure in one-third of patients.