scholarly journals Development, validation and application of a machine learning model to estimate salt consumption in 54 countries

eLife ◽  
2022 ◽  
Vol 11 ◽  
Author(s):  
Wilmer Cristobal Guzman-Vilca ◽  
Manuel Castillo-Cara ◽  
Rodrigo M Carrillo-Larco
Keyword(s):  
Machine Learning ◽  
Salt Intake ◽  
Population Level ◽  
American Samoa ◽  
Population Based ◽  
Urine Samples ◽  
Spot Urine ◽  
Salt Consumption ◽  
Model Based ◽  
Machine Learning Model

Global targets to reduce salt intake have been proposed but their monitoring is challenged by the lack of population-based data on salt consumption. We developed a machine learning (ML) model to predict salt consumption at the population level based on simple predictors and applied this model to national surveys in 54 countries. We used 21 surveys with spot urine samples for the ML model derivation and validation; we developed a supervised ML regression model based on: sex, age, weight, height, systolic and diastolic blood pressure. We applied the ML model to 54 new surveys to quantify the mean salt consumption in the population. The pooled dataset in which we developed the ML model included 49,776 people. Overall, there were no substantial differences between the observed and ML-predicted mean salt intake (p<0.001). The pooled dataset where we applied the ML model included 166,677 people; the predicted mean salt consumption ranged from 6.8 g/day (95% CI: 6.8-6.8 g/day) in Eritrea to 10.0 g/day (95% CI: 9.9-10.0 g/day) in American Samoa. The countries with the highest predicted mean salt intake were in Western Pacific. The lowest predicted intake was found in Africa. The country-specific predicted mean salt intake was within reasonable difference from the best available evidence. A ML model based on readily available predictors estimated daily salt consumption with good accuracy. This model could be used to predict mean salt consumption in the general population where urine samples are not available.

scholarly journals Estimating salt consumption in 49 low-and middle-income countries: Development, validation and application of a machine learning model

2021 ◽  
Author(s):  
Wilmer Cristobal Guzman-Vilca ◽  
Manuel Castillo-Cara ◽  
Rodrigo M Carrillo-Larco
Keyword(s):  
Machine Learning ◽  
Salt Intake ◽  
Population Based ◽  
Urine Samples ◽  
Middle Income ◽  
Spot Urine ◽  
Salt Consumption ◽  
Middle Income Countries ◽  
Model Based ◽  
Low And Middle Income

Background: Global targets to reduce salt intake have been proposed but their monitoring is challenged by the lack of population-based data on salt consumption. We developed a machine learning (ML) model to predict salt consumption based on simple predictors, and applied this model to national surveys in low- and middle-income countries (LMICs).<br /><br />Methods: Pooled analysis of WHO STEPS surveys. We used 19 surveys with spot urine samples for the ML model derivation and validation; we developed a supervised ML regression model based on: sex, age, weight, height, systolic and diastolic blood pressure. We applied the ML model to 49 new STEPS surveys to quantify the mean salt consumption in the population.<br /><br />Results: The pooled dataset in which we developed the ML model included 45,152 people. Overall, there were no substantial differences between the observed (8.1 g/day (95% CI: 8.0-8.2 g/day)) and ML-predicted (8.1 g/day (95% CI: 8.1-8.2 g/day)) mean salt intake (p= 0.065). The pooled dataset where we applied the ML model included 157,699 people; the overall predicted mean salt consumption was 8.1 g/day (95% CI: 8.1-8.2 g/day). The countries with the highest predicted mean salt intake were in Western Pacific. The lowest predicted intake was found in Africa. The country-specific predicted mean salt intake was within reasonable difference from the best available evidence.<br /><br />Conclusions: A ML model based on readily available predictors estimated daily salt consumption with good accuracy. This model could be used to predict mean salt consumption in the general population where urine samples are not available.

scholarly journals Mean Dietary Salt Intake in Vanuatu: A Population Survey of 755 Participants on Efate Island

Nutrients ◽  
2019 ◽  
Vol 11 (4) ◽  
pp. 916
Author(s):  
Katherine Paterson ◽  
Nerida Hinge ◽  
Emalie Sparks ◽  
Kathy Trieu ◽  
Joseph Alvin Santos ◽  
...  
Keyword(s):  
Salt Intake ◽  
Urine Samples ◽  
World Health ◽  
Dietary Salt ◽  
Cross Sectional Survey ◽  
Cross Sectional ◽  
Middle Income ◽  
Dietary Salt Intake ◽  
Spot Urine ◽  
Salt Consumption

Non-communicable diseases are responsible for 63% of global deaths, with a higher burden in low- and middle-income countries. Hypertension is the leading cause of cardiovascular-disease-related deaths worldwide, and approximately 1.7 million deaths are directly attributable to excess salt intake annually. There has been little research conducted on the level of salt consumption amongst the population of Vanuatu. Based on data from other Pacific Island countries and knowledge of changing regional diets, it was predicted that salt intake would exceed the World Health Organization’s (WHO) recommended maximum of 5 g per day. The current study aimed to provide Vanuatu with a preliminary baseline assessment of population salt intake on Efate Island. A cross-sectional survey collected demographic, clinical, and urine data from participants aged 18 to 69 years in rural and urban communities on Efate Island in October 2016 and February 2017. Mean salt intake was determined to be 7.2 (SD 2.3) g/day from spot urine samples, and 5.9 (SD 3.6) g/day from 24-h urine samples, both of which exceed the WHO recommended maximum. Based on the spot urine samples, males had significantly higher salt intake than females (7.8 g compared to 6.5 g; p < 0.001) and almost 85% of the population consumed more than the WHO recommended maximum daily amount. A coordinated government strategy is recommended to reduce salt consumption, including fiscal policies, engagement with the food industry, and education and awareness-raising to promote behavior change.

scholarly journals Estimating sodium intake from spot urine samples at population level: a validation and application study in French adults

2019 ◽  
Vol 122 (2) ◽  
pp. 186-194 ◽  
Author(s):  
Elise Emeville ◽  
Camille Lassale ◽  
Katia Castetbon ◽  
Valérie Deschamps ◽  
Benoît Salanave ◽  
...  
Keyword(s):  
Salt Intake ◽  
Sodium Intake ◽  
Adult Population ◽  
Correlation Coefficients ◽  
Population Level ◽  
Urine Samples ◽  
Validation Sample ◽  
Application Study ◽  
Spot Urine ◽  
Calibration Coefficients

AbstractThe aim of this study was to assess the validity of the predictive INTERSALT equation using spot urine samples to estimate 24-h urinary Na (24-hUNa) excretion and daily Na intake among the French adult population. Among 193 French adults (‘validation sample’), we assessed the validity by comparing predicted 24-hUNa excretion from spot urine and measured 24-hUNa excretion from 24-h urine collections. Spearman correlation coefficients and Bland–Altman plots were used and we calculated calibration coefficients. In a nationally representative sample of 1720 French adults (‘application sample’), the calibrated predictive equation was then applied to the spot urine Na values to estimate 24-hUNa excretion and daily Na intake. In that sample, predicted Na intake was compared with that estimated from 24-h dietary recalls. Results were adjusted and corrected using calibration coefficients. In the validation sample, the measured 24-hUNa excretion was on average 14 % higher than the predicted 24-hUNa (+13 % for men and +16 % for women). Correlation between measured and predicted 24-hUNa excretion was moderate (Spearman r 0·42), and the Bland–Altman plots showed underestimation at lower excretion level and overestimation at higher level. In the application study, estimated daily salt intake was 8·0 g/d using dietary recalls, 8·1 g/d using predicted INTERSALT equation and 9·3 g/d after applying calibration coefficients calculated in the validation study. Despite overall underestimation of 24-hUNa excretion by spot urinary Na, the use of predictive INTERSALT equation remains an acceptable alternative in monitoring global Na intake/excreted in the French population but its use is not advised at the individual level.

scholarly journals Sodium and Salt Consumption in Latin America and the Caribbean: A Systematic-Review and Meta-Analysis of Population-Based Studies and Surveys

Nutrients ◽  
2020 ◽  
Vol 12 (2) ◽  
pp. 556 ◽  
Author(s):  
Rodrigo M Carrillo-Larco ◽  
Antonio Bernabe-Ortiz
Keyword(s):  
Systematic Review ◽  
Latin America ◽  
Salt Intake ◽  
Meta Analysis ◽  
Population Based ◽  
Urine Samples ◽  
World Health ◽  
Salt Consumption ◽  
The Caribbean ◽  
Sodium Consumption

Sodium/salt consumption is a risk factor for cardiovascular diseases. Although global targets to reduce salt intake have been established, current levels and trends of sodium consumption in Latin America and the Caribbean (LAC) are unknown. We conducted a systematic review and meta-analysis of population-based studies in which sodium consumption was analyzed based on urine samples (24 hour samples or otherwise). The search was conducted in Medline, Embase, Global Health, Scopus and LILACS. From 2350 results, 53 were studied in detail, of which 15 reports were included, providing evidence for 18 studies. Most studies were from Brazil (7/18) and six collected 24 hour urine samples. In the random effects meta-analysis, 12 studies (29,875 people) were analyzed since 2010. The pooled mean 24 hour estimated sodium consumption was 4.13 g/day (10.49 g/day of salt). When only national surveys were analyzed, the pooled mean was 3.43 g/day (8.71 g/day of salt); when only community studies were analyzed the pooled mean was 4.39 g/day (11.15 g/day of salt). Studies had low risk of bias. The estimated 24 hour sodium consumption is more than twice the World Health Organization recommendations since 2010. Regional organizations and governments should strengthen policies and interventions to measure and reduce sodium consumption in LAC.

scholarly journals Estimating mean population salt intake in Fiji and Samoa using spot urine samples

2019 ◽  
Vol 18 (1) ◽  
Author(s):  
Joseph Alvin Santos ◽  
Emalie Rosewarne ◽  
Martyna Hogendorf ◽  
Kathy Trieu ◽  
Arti Pillay ◽  
...  
Keyword(s):  
Developing Countries ◽  
Salt Intake ◽  
Intraclass Correlation ◽  
Population Level ◽  
Urine Samples ◽  
Urine Collection ◽  
Cross Sectional ◽  
Spot Urine ◽  
Paired Samples ◽  
Secondary Analyses

Abstract Background There is an increasing interest in finding less costly and burdensome alternatives to measuring population-level salt intake than 24-h urine collection, such as spot urine samples. However, little is known about their usefulness in developing countries like Fiji and Samoa. The purpose of this study was to evaluate the capacity of spot urine samples to estimate mean population salt intake in Fiji and Samoa. Methods The study involved secondary analyses of urine data from cross-sectional surveys conducted in Fiji and Samoa between 2012 and 2016. Mean salt intake was estimated from spot urine samples using six equations, and compared with the measured salt intake from 24-h urine samples. Differences and agreement between the two methods were examined through paired samples t-test, intraclass correlation coefficient analysis, and Bland-Altman plots and analyses. Results A total of 414 participants from Fiji and 725 participants from Samoa were included. Unweighted mean salt intake based on 24-h urine collection was 10.58 g/day (95% CI 9.95 to 11.22) in Fiji and 7.09 g/day (95% CI 6.83 to 7.36) in Samoa. In both samples, the INTERSALT equation with potassium produced the closest salt intake estimate to the 24-h urine (difference of − 0.92 g/day, 95% CI − 1.67 to − 0.18 in the Fiji sample and + 1.53 g/day, 95% CI 1.28 to 1.77 in the Samoa sample). The presence of proportional bias was evident for all equations except for the Kawasaki equation. Conclusion These data suggest that additional studies where both 24-h urine and spot urine samples are collected are needed to further assess whether methods based on spot urine samples can be confidently used to estimate mean population salt intake in Fiji and Samoa.

scholarly journals MODES: model-based optimization on distributed embedded systems

2021 ◽  
Author(s):  
Junjie Shi ◽  
Jiang Bian ◽  
Jakob Richter ◽  
Kuan-Hsun Chen ◽  
Jörg Rahnenführer ◽  
...  
Keyword(s):  
Machine Learning ◽  
Embedded Systems ◽  
Learning Model ◽  
Black Box ◽  
Distributed Embedded Systems ◽  
Data Set ◽  
Individual Model ◽  
Model Based ◽  
Machine Learning Model ◽  
Distributed Machine Learning

AbstractThe predictive performance of a machine learning model highly depends on the corresponding hyper-parameter setting. Hence, hyper-parameter tuning is often indispensable. Normally such tuning requires the dedicated machine learning model to be trained and evaluated on centralized data to obtain a performance estimate. However, in a distributed machine learning scenario, it is not always possible to collect all the data from all nodes due to privacy concerns or storage limitations. Moreover, if data has to be transferred through low bandwidth connections it reduces the time available for tuning. Model-Based Optimization (MBO) is one state-of-the-art method for tuning hyper-parameters but the application on distributed machine learning models or federated learning lacks research. This work proposes a framework $$\textit{MODES}$$ MODES that allows to deploy MBO on resource-constrained distributed embedded systems. Each node trains an individual model based on its local data. The goal is to optimize the combined prediction accuracy. The presented framework offers two optimization modes: (1) $$\textit{MODES}$$ MODES -B considers the whole ensemble as a single black box and optimizes the hyper-parameters of each individual model jointly, and (2) $$\textit{MODES}$$ MODES -I considers all models as clones of the same black box which allows it to efficiently parallelize the optimization in a distributed setting. We evaluate $$\textit{MODES}$$ MODES by conducting experiments on the optimization for the hyper-parameters of a random forest and a multi-layer perceptron. The experimental results demonstrate that, with an improvement in terms of mean accuracy ($$\textit{MODES}$$ MODES -B), run-time efficiency ($$\textit{MODES}$$ MODES -I), and statistical stability for both modes, $$\textit{MODES}$$ MODES outperforms the baseline, i.e., carry out tuning with MBO on each node individually with its local sub-data set.

scholarly journals Salt intake and blood pressure in Iranian children and adolescents: a population-based study

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Mohammad Hassan Emamian ◽  
Hossein Ebrahimi ◽  
Hassan Hashemi ◽  
Akbar Fotouhi
Keyword(s):  
Blood Pressure ◽  
Rural Areas ◽  
Urban Areas ◽  
Salt Intake ◽  
Daily Intake ◽  
Population Based ◽  
World Health ◽  
Second Phase ◽  
Salt Consumption ◽  
The Mean

Abstract Background Previous studies have reported a high prevalence of hypertension in Iranian students, especially in rural areas. The aim of this study was to investigate the daily intake of salt in students and its association with high blood pressure. Methods A random sub-sample was selected from the participants of the second phase of Shahroud schoolchildren eye cohort study and then a random urine sample was tested for sodium, potassium and creatinine. Urine electrolyte esexcretion and daily salt intake were calculated by Tanaka et al.’s formula. Results Among 1455 participants (including 230 participants from rural area and 472 girls), the mean age was 12.9 ± 1.7 year and the mean daily salt intake was 9.7 ± 2.6 g (95% CI 9.5–9.8). The mean salt consumption in rural areas [10.8 (95% CI 10.4–11.2)] was higher than urban areas [9.4 (95% CI 9.3–9.6)], in people with hypertension [10.8 (95% CI 10.3–11.3)] was more than people with normal blood pressure [9.4 (95% CI 9.3–9.6)], and in boys [9.8 (95% CI 9.7–10.0)] was more than girls [9.3 (95% CI 9.1–9.6)]. Higher age, BMI z-score, male sex and rural life, were associated with increased daily salt intake. Increased salt intake was associated with increased systolic and diastolic blood pressure. Conclusion Daily salt intake in Iranian adolescents was about 2 times the recommended amount of the World Health Organization, was higher in rural areas and was associated with blood pressure. Reducing salt intake should be considered as an important intervention, especially in rural areas.

Interpretable machine learning model to detect chemically adulterated urine samples analyzed by high resolution mass spectrometry

2021 ◽  
Vol 0 (0) ◽  
Author(s):  
Gabriel L. Streun ◽  
Andrea E. Steuer ◽  
Lars C. Ebert ◽  
Akos Dobay ◽  
Thomas Kraemer
Keyword(s):  
Machine Learning ◽  
Mass Spectrometry ◽  
Liquid Chromatography ◽  
High Resolution ◽  
Retention Time ◽  
Drug Testing ◽  
High Resolution Mass Spectrometry ◽  
Urine Samples ◽  
Machine Learning Model ◽  
Resolution Mass

Abstract Objectives Urine sample manipulation including substitution, dilution, and chemical adulteration is a continuing challenge for workplace drug testing, abstinence control, and doping control laboratories. The simultaneous detection of sample manipulation and prohibited drugs within one single analytical measurement would be highly advantageous. Machine learning algorithms are able to learn from existing datasets and predict outcomes of new data, which are unknown to the model. Methods Authentic human urine samples were treated with pyridinium chlorochromate, potassium nitrite, hydrogen peroxide, iodine, sodium hypochlorite, and water as control. In total, 702 samples, measured with liquid chromatography coupled to quadrupole time-of-flight mass spectrometry, were used. After retention time alignment within Progenesis QI, an artificial neural network was trained with 500 samples, each featuring 33,448 values. The feature importance was analyzed with the local interpretable model-agnostic explanations approach. Results Following 10-fold cross-validation, the mean sensitivity, specificity, positive predictive value, and negative predictive value was 88.9, 92.0, 91.9, and 89.2%, respectively. A diverse test set (n=202) containing treated and untreated urine samples could be correctly classified with an accuracy of 95.4%. In addition, 14 important features and four potential biomarkers were extracted. Conclusions With interpretable retention time aligned liquid chromatography high-resolution mass spectrometry data, a reliable machine learning model could be established that rapidly uncovers chemical urine manipulation. The incorporation of our model into routine clinical or forensic analysis allows simultaneous LC-MS analysis and sample integrity testing in one run, thus revolutionizing this field of drug testing.

Sign in / Sign up

Export Citation Format

Share Document