In recent years, the global economic crisis has had a profound impact on individuals across various dimensions, including financial stability, employment security, access to essential services, and overall well-being [1]. Furthermore, an economic recession often exacerbates health inequalities across different social groups, disproportionately affecting vulnerable populations such as low-income families, marginalized communities, and public sector workers [2].

Since 2022, Sri Lanka has been facing severe economic pressure, the most challenging situation to have occurred in the past seven decades [3]. The economic downturn had a profound impact at the household level, where its effects became most visible. Many Sri Lankan families struggled to navigate a range of daily challenges, including managing reduced incomes, meeting financial obligations, and safeguarding the well-being of their family members [4]. The strain on household resources not only disrupted everyday routines but also heightened emotional and psychological stress across all age groups.

In addition to the general population, employees in both the state and private sectors have been significantly affected by the ongoing economic crisis, experiencing a range of health issues as a result [5]. The financial strain has contributed to increased levels of stress, anxiety, and burnout among workers, particularly as they struggle to cope with rising living costs, job insecurity, and reduced access to essential services. These conditions have created a high-pressure environment that adversely impacts both physical and mental well-being [6]. According to the International Labor Organization (ILO), the employment sector in Sri Lanka was severely affected by the economic crisis, resulting higher unemployment rate in 2023, compared to the year 2019 [7].

Teachers play an essential role in shaping the academic and personal development of their students [8]. Consequently, teachers continuously invest in both their personal and professional development to maintain and enhance the quality of education they deliver. This commitment includes pursuing further education, participating in training programs, engaging in reflective practice, and adapting to evolving pedagogical methods [9]. Furthermore, teachers’ mental well-being is closely linked to students'psychological health and is a significant contributor to positive educational outcomes among school children [10]. However, despite their critical role, recent studies suggest that teachers are considerably vulnerable to mental health challenges [11, 8]. The demographic variables such as age, gender, living environment, socioeconomic status, and professional factors such as service years, type of schools, and work environment are significantly associated with negative mental health outcomes among teachers [12, 9].

School teachers are particularly vulnerable to severe mental health issues during times of social crisis, as they often serve as the primary point of contact between students, families, and the broader education system. For instance, schoolteachers in Sri Lanka experienced a higher prevalence of psychological distress that has resulted due to social insecurity, isolation, and this abrupt transition from in-person to online teaching modes were major contributors to elevated psychological distress among school teachers during the COVID-19 pandemic [13]. Such disruptions not only impact teachers'mental health but can also reduce job satisfaction and potentially compromise the overall quality of education [14]. While the effects of pandemics on teachers'mental well-being are well-documented, very handful studies have examined the impact of other social challenges, such as financial crises, on psychological instability among teachers across the globe [15]. Although several investigations have been conducted to observe the association between the economic crisis on mental wellbeing among selected professional categories in Sri Lanka [16], the existing literature provides limited insight into how this financial hardship is associated with the mental health status of school teachers in Sri Lanka.

The twelve-item General Health Questionnaire (GHQ-12) is widely recognized as one of the most effective tools for identifying psychiatric morbidity in the general population [17]. This self-assessment instrument has been extensively used worldwide and translated into multiple languages [18]. Research commonly employs a validated Sinhala version of the GHQ-12 in Sri Lanka. One of its key strengths is its simplicity and efficiency, making it a practical tool for diverse settings [19]. Although the GHQ-12 has been validated in Sinhala, its applicability across different demographic groups may require further evaluation, as variations in comprehension could impact response accuracy. Therefore, another aim of this study is to validate the GHQ-12 version specifically for this professional group. Additionally, understanding the relationship between economic challenges and coping strategies is crucial for improving teachers'job satisfaction and well-being. Insights from this study could inform the development of targeted interventions to mitigate the adverse effects of economic hardships on educators. Further uncovering this association may enable teachers to address this issue among teachers and help to improve the academic performance and mental health status of school children as there is a strong relationship between teachers'psychological health and the educational outcomes of school children.

A web-based cross-sectional survey was conducted from 20 March to 30 April 2024 to assess the association between ongoing economic crises and teachers'mental health outcomes.

There are 10,146 government schools in Sri Lanka, and more than 240,000 teachers are employed to teach around four million students [20]. Teachers attached to government schools in Sri Lanka were invited to participate in the study. Several reasons justified the selection of government school teachers as the target population for the present study. Firstly, government teachers represent a relatively standardized and homogeneous group in terms of salary structure, job security, and working conditions. Secondly, as public sector employees, they are subject to economic pressures and policy shifts that differ markedly from those experienced by teachers in private or international schools. Furthermore, government teachers are employed across both urban and rural regions, enabling the study to capture geographic variations while maintaining consistency in employment structure and benefits.

The sample size was calculated using the formula provided by Lwanga and Lemeshow [21]. To estimate the sample, the prevalence of mental health distress among teachers in Sri Lanka, reported at 27.2% [22], was used as the expected prevalence (p). The Z-value corresponding to a 95% confidence interval (Z = 1.96) and a precision level (d) 0.05 were applied. The formula used was n = Z² * P(1-P)/d², resulting in a sample size of 305. To account for potential dropouts and non-responses during data collection, the final sample size was increased by 10%, bringing it to 335 participants. To meet the required minimum sample size, participants were selected to represent various types of schools (Type 1, Type 2, 1 C, and AB schools) across Sri Lanka. Only teachers from government schools were included in the study, while those from private schools or those with self-reported chronic mental health conditions were excluded.

A self-administered questionnaire consisting of three sections was used to collect data. The first section gathered general demographic information, including age, gender, monthly income, and years of service. The second section focused on changes in response to economic pressures, with questions addressing how household income was managed, the additional costs required to purchase essentials, reductions in spending on specific needs, and alternative income-generating activities adopted to cope with financial strain. In this section, participants were asked to self-report whether they had experienced a reduction in their monthly income during a specified period (e.g., “Have you experienced a reduction in your monthly income compared to the pre-crisis period?”). The economic burden was assessed using a set of composite indicators reflecting increased financial pressure on households, such as the need to apply for loans, pawn valuables, or seek external financial assistance. Financial hardship was evaluated by asking participants whether they encountered difficulties in managing their household expenses during the same period. The final section employed a validated version of the General Health Questionnaire (GHQ-12) [17] to assess participants'mental health outcomes. This 12-item questionnaire measured the severity of mental health issues over the past few weeks using a four-point Likert scale (ranging from 0 to 3).

All sections of the questionnaire were translated into Sinhala and Tamil. However, the already validated Sinhala version [19] of the GHQ-12 was used in this survey, while the English version of the GHQ-12 was professionally translated into Tamil for this study. To ensure reliability, the Tamil translation was back-translated into English. The finalized questionnaire, including an information sheet and a consent form, was formatted into a Google Form. A pre-test was conducted with a small sample (n = 20) to ensure the clarity and cultural sensitivity of the questions. Since this was a self-administered questionnaire, each section was thoroughly explained to facilitate accurate responses. The web-based questionnaire was then distributed to eligible study participants via WhatsApp and email.

The data were collected from the online questionnaire and organized in a Microsoft Excel spreadsheet. A constant value imputation method was used to identify the missing values and eliminate duplicate entries. However, neither missing values nor duplicate entries were identified. The data were analyzed using the Statistical Package for the Social Sciences (SPSS), version 23.0.

In this study, the General Health Questionnaire (GHQ-12) was scored using a simple Likert scale ranging from 0 to 3, resulting in a total possible score between 0 and 36. These total scores were then categorized into five levels of psychological distress for interpretation: scores of 1–10 indicated low distress, 11–12 were considered typical, 13–15 represented above-average distress, 16–20 suggested the presence of psychological distress, and scores above 20 indicated severe psychological distress.

Demographic and socio-economic characteristics, along with self-reported general and mental health assessed through the GHQ-12, were analyzed. The factor structure of the GHQ-12 was evaluated using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA). Bartlett's test of sphericity (p < 0.001) and the Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy (0.918) confirmed the suitability of the data for factor analysis. EFA was conducted using principal component analysis with Varimax rotation to identify underlying factor structures. Subsequently, CFA was performed to test the fit of four proposed models:

Descriptive analysis was conducted to extract mean values, standard deviations (SDs), frequencies, and percentages for both categorical and continuous variables with 95% confidence intervals (95% CIs), using a significance level of p < 0.05. The associations between demographic variables and mental health outcomes were evaluated using the chi-square test. Furthermore, a multivariate regression analysis was performed to identify significant predictors of distress level of the study participants.

With a response rate of 84.5%, a total of 283 teachers completed the questionnaire, all of whom used the Sinhala version. Among the participants, 65% were female (n = 184), and the majority (24.0%) were aged 25 to 30 years. Additionally, 82.3% of the respondents were married, and most (61.5%) reported an average monthly family income of 50,000 to 100,000 LKR. Just over one-third of the teachers in this study represented national schools in Sri Lanka, while nearly 29.0% had 10 to 15 years of service experience. Furthermore, nearly one in five participants reported living with a chronic disease (Table 1).

Over two-thirds (81.6%) of the research participants reported that their monthly income was insufficient to cover all their expenses. As a result, all study participants reduced household spending in reaction to financial strain. Specific areas where costs were curtailed included medicine (77.7%), transportation (62.5%), entertainment (91.5%), and food (47.3%). Additionally, 77.7% of participants limited their monthly savings. In response to these financial challenges, the majority sought ways to earn extra income. Approximately half of the teachers in this survey indicated they required an additional 15,000 to 20,000 LKR to meet their monthly needs. Consequently, 77.7% of participants engaged in alternative income-generating activities. Among those who sought extra income, 18.0% took on part-time jobs, 10.2% obtained instant loans, and 26.5% used pawn services.

Moreover, a staggering 98.2% reported reducing food expenses during the economic crisis. Specifically, 92.2% decreased their consumption of fruits and vegetables, while 92.6% reduced their intake of protein sources such as meat, fish, and eggs. Additionally, 76.0% reported an increased consumption of bakery products (Table 2).

Table 3 presents the mean score of GHQ items, while Table 4 indicates the distribution of study participants based on their level of psychological distress. The mean General Health Questionnaire (GHQ) score was 15.15 (SD ± 8.14, 95% CI). The skewness value suggests a right-skewed distribution, indicating that the data has a longer tail on the right side. This implies that a few higher values are pulling the distribution toward the right. Meanwhile, the kurtosis value indicates a sharper peak, suggesting that the distribution is more peaked compared to a normal distribution. Among the participants, 33.6% experienced low distress, while 13.4% showed evidence of psychological distress. Additionally, 30.4% of participants reported experiencing severe distress (Table 4).

Table 5 summarizes the Pearson correlation coefficients for all items. The strongest correlation was observed between Item 3 and Item 5 (r = 0.822), while the weakest association was between Item 3 and Item 12 (r = 0.158). The reliability analysis of the inter-item co-relation indicated 0.454 (0.3–0.8) moderate mean-correlation, affirming the acceptability of conducting a factor analysis.The factor analysis suitability was explained using a KMO value of 0.918, which surpassed the recommended value of 0.6 Bartlett Test of sphericity (approximate chi-square = 3000.390, df = 66, p < 0.001). Then, the Principal Axis Factoring (PAF) with oblique rotation was conducted to determine the factor structure. Aa. A two-factor structure was identified following an oblique rotation principal component analysis, collectively explaining 62% of the total variance (Table 6). The two factors were extracted based on an eigenvalue greater than one and a visual inspection of the scree plot. The further range of factor loading for factor one was 0.69 to 0.82, while factor loading for factor two was 0.52 to 0.87. In addition, the cut-off for retaining items < 0.49 was suppressed. Factor one was labeled as social dysfunction, whereas factor two was labeled as depression and anxiety.

In the Confirmatory Factor Analysis (CFA), the Root Mean Square Error of Approximate (RMSEA) was greater than 0.08 for all the models. In addition to that, the Goodness of Fit Index (GFI), and Comparative Fit Index (CFI) were greater than 0.95 for model 1 and model 2, while Tusker- Lewis Index (TLI) showed a good fit for all the models. Nevertheless, the Standardized Root Mean Square Residual (SRMR) values for all models remained within the acceptable range, as summarized in Table 7.

The results of the Confirmatory Factor Analysis (CFA) are also illustrated in Fig. 1. In the model, Factor 1, which comprised Items 1 through 8, was designed to assess social performance disturbances. Factor 2, consisting of Items 9 through 12, measured mental health disturbances. A strong positive correlation was observed between the two factors, suggesting a close relationship between social functioning and psychological distress in the study population.

Confirmatory Factor Structure for twelve-item, Factor 1 assesses the social dysfunction while factor 2 assesses the psychological distress

Full size image

The multivariate regression analysis revealed that the selected factors collectively accounted for approximately 29% of the variance in participants'distress levels (R² = 0.29). After controlling for factors such as age, income, marital status, type of school, and years of service, the odds of experiencing depression were 2.9 times (95% CI: 1.270–1.634, β = −0.29) for individuals who experienced a loss in monthly income. Additionally, those without monthly savings had a 1.7 times likelihood (95% CI: 0.436–6.312, β = −0.45) of experiencing a reduced quality of life. Moreover, reducing monthly spending on entertainment (OR = 2.78, 95% CI: 1.324–5.123, β = −0.42), cutting back on household essentials (OR = 0.28, 95% CI: 1.424–5.616, β = −0.51), and seeking additional income to meet routine expenses (OR = 2.5, 95% CI: 1.214–4.113, β = −0.52) were significantly associated with increased odds of depression among the participants (Table 8).

This pioneering study examines the psychometric properties of the General Health Questionnaire (GHQ-12) specifically for school teachers while investigating psychological distress among government educators during Sri Lanka's severe economic crisis. Using a methodological approach combining Exploratory Factor Analysis (EFA) to identify underlying structures and Confirmatory Factor Analysis (CFA) for validation, researchers collected comprehensive data on demographics, economic adaptations, financial coping mechanisms, and psychological well-being.

Findings revealed that a significant majority of participating teachers reported insufficient monthly income to meet basic needs, forcing widespread lifestyle modifications. Specifically, teachers reduced expenditures on transportation (primary reduction), followed by cuts to entertainment, medicine, food, and communication services. The financial shortfall was substantial, with over half of the participants indicating they required an additional 15,000–20,000 LKR monthly to cover essential expenses. To address this income gap, teachers employed multiple financial coping strategies, including pursuing self-employment opportunities, borrowing money from various sources, pawning valuable possessions, and taking high-interest instant loans. The psychological impact of these economic pressures was severe, with approximately one-third of participants experiencing clinical levels of psychological distress. Multivariate logistic regression analysis identified five significant predictors of severe distress among teachers: reduced income compared to pre-crisis levels, depleted savings, necessity to cut back on essential needs, inability to afford entertainment or social activities, and pressure to seek supplementary income sources. These findings highlight the profound intersection between economic hardship and psychological wellbeing in the teaching profession during national financial crises.

The current study identified a higher prevalence of psychological distress among teachers, aligning with recent research [9, 25,26,27]. However, our findings diverge from several studies that documented lower psychological distress rates in teacher populations [28, 29]. This discrepancy may stem from demographic variations, as participants in these contradictory studies predominantly represented high-income communities and diverse ethnic backgrounds. Furthermore, differences in working environments likely contributed to the elevated distress levels observed in our study.

The factor analysis in this study was conducted exclusively on responses from the Sinhala version of the questionnaire, as the Tamil-translated version received no responses. Psychometric analysis of the GHQ-12 revealed two distinct factors through exploratory factor analysis (EFA): Factor 1 (social dysfunction) and Factor 2 (depression and anxiety), which provided the best fit for the data. Together, these factors explained approximately 62% of the total variance, consistent with previous findings from the same population [30]. In this survey, both factors demonstrated strong validity with factor loadings exceeding 0.5, aligning with earlier evidence [30]. Items categorized under Factor 1 (social dysfunction) included'capable of making decisions,''playing a useful part in things,''able to concentrate,''enjoying day-to-day activities,''losing confidence,'and'thinking of oneself as worthless.'Factor 2 (depression and anxiety) comprised items such as'facing up to problems,''feeling reasonably happy,''feeling unhappy and depressed,''constantly under strain,''losing sleep over worry,'and'unable to overcome difficulties.'While our two-factor structure aligns with previous research by Lutke et al. [30], it differs from many other GHQ-12 studies that have proposed three-factor models [31]. For instance, Lee and Kim [32] identified sleep disturbance, social performance, and loss of confidence as key factors, while Anjara et al. [33] outlined distress, anxiety, and social function. These variations in factor structures likely stem from differences in sample characteristics and sizes, with many previous studies utilizing substantially larger samples than the current investigation.

In Sri Lanka, teachers earn approximately 50,000 LKR monthly, which is substantially below other government employees'salaries and roughly half of private-sector executive compensation [34]. This financial constraint became particularly acute during the data collection period, when inflation fluctuated between 6.4% and 6.5%, causing prices of some goods to triple [35]. Consequently, study participants have restricted essential purchases to remain within their limited budgets.

This study identified a positive correlation between reduced essential expenditures and increased psychological distress among participants. While establishing a direct causal relationship between these economic adaptations and participants'psychological traits remains challenging, evidence suggests that adapting to these restricted living conditions likely increases vulnerability to psychological problems. This aligns with previous findings among both the same target population [15] and broader populations [36]. Confirming this relationship definitively will require further research addressing multiple societal factors, including the specific working conditions of Sri Lankan teachers. Although our study did not explicitly examine how lifestyle modifications affect teachers'academic engagement, existing literature indicates that financial pressures significantly impact teachers'professional performance.

The present study revealed that over half of the participants engaged in additional income-generating activities. Seeking secondary employment represents a recognized resilience strategy among teachers confronting economic hardship [15]. Teachers typically pursue socially acceptable supplementary income sources for survival—in Zimbabwe, for instance, educators reported involvement in cross-border trading and gold mining, mirroring coping mechanisms employed by other citizens. In this investigation, most teachers selected education-related side jobs, particularly private tutoring [37]. However, the teaching profession encompasses multiple responsibilities including lesson delivery, assessment administration, and co-curricular activity facilitation. Despite spending considerable hours at school, teachers require additional time for curriculum preparation and personal obligations, as their time quality significantly impacts student educational outcomes [38, 39]. The widespread pursuit of supplementary income among teachers potentially compromises instructional quality, thereby affecting the nation's overall educational standards. Consequently, future research should examine the implications of these supplementary activities, as such insights are essential for developing policies that simultaneously support teacher welfare and educational excellence.

The multivariate logistic regression analysis in this study identified that a loss of monthly income and lack of savings are strong predictors of increased psychological impairment among participants. This association is consistent with findings from studies on teachers in Thailand [40] and China [41]. Financial difficulties among teachers often result in reduced opportunities for entertainment and recreation [29]. The present study confirms that teachers have significantly curtailed their recreational activities due to economic pressures, contributing to heightened distress, as supported by recent evidence [42]. These findings highlight the importance of providing recreational opportunities for teachers to maintain a healthy work-life balance and improve mental health outcomes. In addition to cutting back on entertainment, a notable portion of the participants in this study reported reducing their daily essentials to manage their limited salary. Similar findings have been observed in other contexts, where teachers in Western countries also reported cutting costs on food, transportation, communication, and other necessities [43, 15]. The results of this study guarantee the application of policies and guidelines to encourage teachers'recreational activities to raise the standard of instruction.

Finally, the majority of study participants reported reducing their expenditures on protein sources, fruits, and vegetables. Economic downturns typically drive inflation, leading to disproportionate increases in food prices. Individuals consequently seek more affordable alternatives, often sacrificing nutritional quality [44]. Poultry products face particularly steep price increases due to rising production and maintenance costs. Consistent with our findings, previous research confirms that during economic recessions, people gravitate toward less nutritious options while reducing consumption of fruits, vegetables, and protein sources like poultry [45]. This dietary shift raises potential public health concerns, including nutrient deficiencies, increased risk of non-communicable diseases, and compromised immune function [46]. However, our participants reported dietary habits based on memory recall—a notable limitation. Future research should employ more robust dietary assessment tools to quantify specific food group consumption and evaluate whether participants meet recommended intake levels for nutritious foods. Such findings could inform both immediate interventions and long-term strategies to improve dietary behaviors among teachers in similar socioeconomic contexts.

A key strength of this study is its pioneering identification of the relationship between Sri Lankan teachers'mental health outcomes and the country's ongoing economic recession. This research also represents the first investigation of the factorial structure and psychometric properties of the GHQ-12 within this specific Sri Lankan population. Additionally, the study's sample encompasses a diverse range of demographic characteristics found in Sri Lankan schools, enhancing the representativeness of the findings. The results yield valuable insights that could inform policy interventions aimed at improving the quality of Sri Lanka's educational system.

Despite its strengths, this study has several limitations that warrant consideration in future research. First, it overlooked workplace environment factors, which significantly influence teacher mental health outcomes. Second, the findings are limited to government school teachers in Sri Lanka and cannot be generalized to the broader teaching population. The self-administered questionnaire methodology introduces potential concerns regarding data accuracy, as financial and psychological information may be affected by social desirability bias and stigma-related underreporting. The study's statistical power was potentially constrained by its relatively small sample size; future investigations should incorporate larger participant pools to enhance the robustness of findings. Finally, the lower-than-expected response rate, likely attributable to the online survey format, suggests this data collection approach may have created accessibility barriers for some teachers..

In conclusion, this study investigated the psychological well-being of teachers amid Sri Lanka's ongoing economic crisis, validating the factorial structure of the General Health Questionnaire-12 (GHQ-12) for this population. The findings revealed widespread financial hardship, with most participants reporting decreased family income and insufficient funds to maintain their standard of living. To cope with these financial pressures, many teachers were forced to reduce spending on necessities including food, medicine, transportation, communication, and entertainment. Significantly, half of the participants needed to generate supplemental income—approximately equal to half their regular salary—to meet basic needs, compelling many to pursue additional employment opportunities. The study demonstrated that one-third of teachers experienced severe psychological distress, with results establishing a strong correlation between financial-driven lifestyle modifications and mental health deterioration. These findings highlight the profound impact of economic instability on both the financial security and psychological wellness of educational professionals in Sri Lanka.

According to the current study, the survey population had a higher prevalence of mental discomfort. At the time the participants completed the questionnaire, there were no comprehensive interventions or programs led by the Ministry of Education in Sri Lanka specifically aimed at promoting the mental well-being of school teachers. Thus, it is advised that long-lasting interventions be created to enhance teachers'mental health results. The goals of these interventions must be to improve teachers'resilience and coping mechanisms while offering them psychological assistance to enhance their mental health and general wellness. Government policies should also prioritize raising the standard of living for educators by offering sufficient financial assistance to reduce economic strain, which hurts mental health stability and the capacity to combine work and family obligations. This might discourage people from looking for other sources of income, which would greatly improve educational achievements.

The datasets used for the analysis and data collection tools can be accessed online repository (https://drive.google.com/drive/folders/1vt3BstsYn5-oFY7wQGsksESW8jyywCIS?usp=drive_link).

GHQ:

General Health Questionnaire

SPSS:

Statistical Package for the Social Sciences

EFA:

Exploratory Factor Analysis

CFA:

Confirmatory Factor Analysis

KMO:

Kaiser-Meyer-Olkin

ILO:

International Labor Organization

Download references

C.P. designed the research, conducted data analysis, and prepared the original draft. S.P. contributed to the methodology validation and interpretation of results. L.P. and M.S. reviewed, edited, and refined the manuscript. All authors have read the final version of the manuscript and approved it for publication.

Correspondence to C. P. Senevirathne.

This study was conducted in accordance with the principles outlined in the Declaration of Helsinki. Participants were provided with an information sheet detailing the study procedures and were required to give informed consent before participation. The ethical approval for this study was obtained from the Ethics Review Committee of Faculty of Applied Sciences, Rajarata University of Sri Lanka, under the reference number ERC/04/028.

This manuscript does not contain any personally identifiable information, such as names, contact details, or identity card numbers of the study participants. Additionally, it does not include any sources requiring additional consent, such as images or videos.

The authors declare no competing interests.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions