Enhancing real-time cell culture monitoring: Automated Raman model optimization with Taguchi method.

Raman spectroscopy has found widespread usage in monitoring cell culture processes both in research and practical applications. However, commonly, preprocessing methods, spectral regions, and modeling parameters have been chosen based on experience or trial-and-error strategies. These choices can significantly impact the performance of the models. There is an urgent need for a simple, effective, and automated approach to determine a suitable procedure for constructing accurate models. This paper introduces the adoption of a design of experiment (DoE) method to optimize partial least squares models for measuring the concentration of different components in cell culture bioreactors. The experimental implementation utilized the orthogonal test table L25(56). Within this framework, five factors were identified as control variables for the DoE method: the window width of Savitzky-Golay smoothing, the baseline correction method, the order of preprocessing steps, spectral regions, and the number of latent variables. The evaluation method for the model was considered as a factor subject to noise. The optimal combination of levels was determined through the signal-to-noise ratio response table employing Taguchi analysis. The effectiveness of this approach was validated through two cases, involving different cultivation scales, different Raman spectrometers, and different analytical components. The results consistently demonstrated that the proposed approach closely approximated the global optimum, regardless of data set size, predictive components, or the brand of Raman spectrometer. The performance of models recommended by the DoE strategy consistently surpassed those built using raw data, underscoring the reliability of models generated through this approach. When compared to exhaustive all-combination experiments, the DoE approach significantly reduces calculation times, making it highly practical for the implementation of Raman spectroscopy in bioprocess monitoring.

Monitoring of sugars adsorption breakthrough curves with online Raman spectroscopy.

We introduce a new application for online Raman spectroscopy to monitor adsorption breakthrough curves of a glucose and xylose mixtures. Univariate and multivariate Partial Least Squares (PLS) calibration models are developed for each sugar when they are dissolved in water and in the case of the ethanol addition as a cosolvent. The models are validated by performing actual breakthrough experiments in a liquid phase using a column packed with a zeolite adsorbent. The first statistical moments of predicted curves are compared to the reference curves obtained with offline High-Performance Liquid Chromatography (HPLC). Glucose and xylose univariate predictions in the presence or absence of ethanol in the mixture are accurate and no improvements are found with the PLS models. Spectral subtraction coupled with the first derivative proved to be effective pretreatments to develop robust univariate models.

Chemical Characterization, Sensory Definition and Prediction Model of the Cider Dryness from New York State Apples.

Cider is a fermented drink obtained from apple juice. As a function of the used apple cultivar, cider can be classified in four different categories (dry, semi-dry, semi-sweet, sweet), distinguished by the attribute of "dryness," which reflects the sweetness and softness perceived. The dryness level is defined by scales (IRF, NYCA scales) based on the residual sugar, titratable acidity and tannin contents. Despite some adjustments, these scales show limitations in the prediction of actual perceived dryness, as they cannot consider the complicated interrelation between combined chemical compounds and sensory perception. After defining the perceived sensory dryness and its sensory description by using the quantitative descriptive analysis (QDA) method, a multivariate approach (PLS) was applied to define a predictive model for the dryness and to identify the chemical compounds with which it was correlated. Three models were developed, based on three different sets of chemical parameters, to provide a method that is easily applicable in the ordinary production process of cider. The comparison between the predicted rating and the relative scales scores showed that the models were able to predict the dryness rating in a more effective way. The multivariate approach was found to be the most suitable to study the relation between chemical and sensory data.

Prediction of tannin content and quality parameters in astringent persimmons from visible and near-infrared spectroscopy.

Introduction: Tannin content and postharvest quality characteristics of persimmon fruit are often determined by the destructive analysis that consumes time, does not allow the acquisition of data from the same fruit continuously, and requires expensive high-performance equipment. This research was done to investigate the potential for non-destructive estimation of astringency and quality parameters in persimmon fruit based on visible/near-infrared (VNIR) spectra. Methods: VNIR spectra readings, the reference tannin content, and quality parameters were measured from fruits of "Cheongdo-Bansi" and "Daebong" persimmon cultivars at harvest and throughout the ripening/deastringency period. The spectra readings from half of the total fruit were utilized for the calibration set, while the other half readings were used for the prediction set. To develop models correlating the spectra data to the measured reference parameters data, the partial least square regression (PLSR) method was utilized. Results and discussion: In the case of 'Daebong', the coefficients of determination (R2) between VNIR spectra and the actual measured values of TSS, firmness, simple sugars, and tannin content were (0.95, 0.94, 0.96, and 0.96) and (0.93, 0.89, 0.96, and 0.93), for the calibration and prediction sets, respectively. Similarly, the R2-values of (0.86, 0.93, 0.79, and 0.81) and (0.83, 0.91, 0.75, and 0.75) were recorded in 'Cheongdo-Bansi' for the calibration and prediction sets, respectively. Additionally, the acquired data were divided into two sets in a 3:1 ratio to develop predictive models and to validate the models in multiple regressions. PLSR models were developed in multiple regression to estimate the tannin content of both cultivars from firmness and simple sugars with R2-values of 0.83 and 0.79 in 'Cheongdo-Bansi' for the calibration and prediction sets, respectively, whereas, R2-values of 0.80 and 0.84 were recorded in 'Daebong' for the calibration and prediction sets, respectively. The overall findings of this study showed the possibility of using VNIR spectra for the prediction of postharvest quality and tannin contents from intact persimmon fruit with quick, chemical-free, and low-cost assessment methods. Also, the multiple regression using physicochemical parameters could fairly predict the tannin content in persimmon fruit though destructively but save time and low-cost.

Growth, physiological, and temperature characteristics in chinese cabbage pakchoi as affected by Cd- stressed conditions and identifying its main controlling factors using PLS model.

BACKGROUND: Although hormesis induced by heavy metals is a well-known phenomenon, the involved biological mechanisms are not fully understood. Cadmium (Cd) is a prevalent heavy metal in the environment. Exposure of Cd, via intake or consumption of Cd-contaminated air or food, poses a huge threat to human health. Chinese cabbage pakchoi (Brassica chinensis L.) is widely planted and consumed as a popular vegetable in China. Therefore, studying the response of Chinese cabbage pakchoi to Cd- stressed conditions is critical to assess whether cabbage can accumulate Cd and serve as an important Cd exposure pathway to human beings. In this study, we investigated the influence of Cd stress on growth, photosynthetic physiology, antioxidant enzyme activities, nutritional quality, anatomical structure, and canopy temperature in Chinese cabbage pakchoi. A partial least squares (PLS) model was used to quantify the relationship between physical and chemical indicators with Cd accumulation in cabbage, and identify the main controlling factors. RESULTS: Results showed that Cd stress significantly inhibited cabbage's growth and development. When Cd stress was increased, the phenotypic indicators were significantly reduced. Meanwhile, Cd stress significantly enhanced the oxidative stress response of cabbage, such as the activities of catalase (CAT), superoxide dismutase (SOD), peroxidase (POD), ascorbate peroxidase (APX), and the content of malondialdehyde (MDA) in leaves. Such a change tended to increase fenestrated tissues' thickness but decrease the thickness of leaf and spongy tissues. Moreover, Cd stress significantly increased soluble sugar, protein, and vitamin C contents in leaves as well as the temperature in the plant canopy. The PLS model analysis showed that the studied phenotypic and physicochemical indicators had good relationships with Cd accumulation in roots, shoots, and the whole plant of cabbage, with high coefficient of determination (R2) values of 0.891, 0.811, and 0.845, and low relative percent deviation (RPD) values of 3.052, 2.317, and 2.557, respectively. Furthermore, through analyzing each parameter's variable importance for projection (VIP) value, the SOD activity was identified as a key factor for indicating Cd accumulation in cabbage. Meanwhile, the effects of CAT on Cd accumulation in cabbage and the canopy mean temperature were also high. CONCLUSION: Cd stress has significant inhibitory effects and can cause damage cabbage's growth and development, and the SOD activity may serve as a key factor to indicate Cd uptake and accumulation in cabbage.

Assessment of the Effect of Selenium Supplementation on Production of Selected Cytokines in Women with Hashimoto's Thyroiditis.

The impact of selenium on the course of Hashimoto's thyroiditis (HT) was mainly assessed by monitoring the titer of antithyroid autoantibodies in most of the studies conducted hitherto. On the other hand, the imbalance in activity of T cells such as Th1, Th2, Th17, and Treg may be relevant in the pathogenesis of this disease. Hence, the assessment of changes in the secretion of cytokines by these cells during selenium supplementation in patients with HT seems to be an important issue and was the main goal of this study. A further aim was to search for correlations among these cytokines, as well as markers of thyroid function, selenium/iodine status in the body, and other biochemical parameters. The group of 29 women with newly diagnosed Hashimoto's thyroiditis was supplemented with selenium in a dose of 100 µg/day for 6 months. Immunological parameters: interferon γ, tumor necrosis factor α, chemokine CXCL10, interleukin 4, interleukin 1ß, interleukin 17, transforming growth factor ß, and C-reactive protein, as well as selenium status parameters were determined in serum twice, i.e., before and after supplementation. Selenium supplementation was associated with a change in the production of two cytokines: interferon γ and interleukin 1ß, for which a decrease and an increase in concentration were observed, respectively. The partial least squares (PLS) model revealed the presence of many relevant correlations among analyzed parameters. The stage of HT development, degree of thyroid dysfunction, and selenium supplementation of diet are interdependent factors which shape the profile of some cytokines secreted by cells participating in the autoimmunity process.

Rapid monitoring of flavonoid content in sweet tea (Lithocarpus litseifolius (Hance) Chun) leaves using NIR spectroscopy.

BACKGROUND: Sweet tea, which functions as tea, sugar and medicine, was listed as a new food resource in 2017. Flavonoids are the main medicinal components in sweet tea and have significant pharmacological activities. Therefore, the quality of sweet tea is related to the content of flavonoids. Flavonoid content in plants is normally determined by time-consuming and expensive chemical analyses. The aim of this study was to develop a methodology to measure three constituents of flavonoids, namely, total flavonoids, phloridin and trilobatin, in sweet tea leaves using near-infrared spectroscopy (NIR). RESULTS: In this study, we demonstrated that the combination of principal component analysis (PCA) and NIR spectroscopy can distinguish sweet tea from different locations. In addition, different spectral preprocessing methods are used to establish partial least squares (PLS) models between spectral information and the content of the three constituents. The best total flavonoid prediction model was obtained with NIR spectra preprocessed with Savitzky-Golay combined with second derivatives (SG + D2) (RP2 = 0.893, and RMSEP = 0.131). For trilobatin, the model with the best performance was developed with raw NIR spectra (RP2 = 0.902, and RMSEP = 2.993), and for phloridin, the best model was obtained with NIR spectra preprocessed with standard normal variate (SNV) (RP2 = 0.818, and RMSEP = 1.085). The coefficients of determination for all calibration sets, validation sets and prediction sets of the best PLS models were higher than 0.967, 0.858 and 0.818, respectively. CONCLUSIONS: The conclusion indicated that NIR spectroscopy has the ability to determine the flavonoid content of sweet tea quickly and conveniently.

Corrigendum: Risk Perception and PTSD Symptoms of Medical Staff Combating Against COVID-19: A PLS Structural Equation Model.

[This corrects the article DOI: 10.3389/fpsyt.2021.607612.].

Shiitake cultivation as biological preprocessing of lignocellulosic feedstocks - Substrate changes in crystallinity, syringyl/guaiacyl lignin and degradation-derived by-products.

Formulation of substrates based on three hardwood species combined with modulation of nitrogen content by whey addition (0-2%) was investigated in an experiment designed in D-optimal model for their effects on biological preproceesing of lignocellulosic feedstock by shiitake mushroom (Lentinula edodes) cultivation. Nitrogen loading was shown a more significant role than wood species for both mushroom production and lignocellulose degradation. The fastest mycelial colonisation occurred with no nitrogen supplementation, but the highest mushroom yields were achieved when 1% whey was added. Low nitrogen content resulted in increased delignification and minimal glucan consumption. Delignification was correlated with degradation of syringyl lignin unit, as indicated by a significant reduction (41.5%) of the syringyl-to-guaiacyl ratio after cultivation. No significant changes in substrate crystallinity were observed. The formation of furan aldehydes and aliphatic acids was negligible during the pasteurisation and fungal cultivation, while the content of soluble phenolics increased up to seven-fold.

Method operable design region obtained with a partial least squares model inversion in the determination of ten polycyclic aromatic hydrocarbons by liquid chromatography with fluorescence detection.

A chromatographic method with the Analytical Quality by Design (AQbD) methodology is developed for the simultaneous determination by HPLC-FLD of ten PAHs (naphthalene, phenanthrene, anthracene, fluoranthene, pyrene, chrysene, benzo[a]anthracene, perylene, benzo[b]fluoranthene, and benzo[a]pyrene), widely spread in the environment. The construction of the Method Operable Design Region (MODR) is conducted, for the first time, via the inversion of a multiresponse Partial Least Squares (PLS2) model, which is needed to maintain the correlations among the Critical Method Parameters (CMP), among the Critical Quality Attributes (CQA), and the covariance between one another. The five CMP considered were the composition of the mobile phase (water, methanol, acetonitrile), flow rate, and column temperature. The eight CQA were linked to resolution between peaks recorded in the same emission wavelength (greater than 1.4) and the total time (less than 15 minutes). By systematic use of experimental design and parallel coordinates plots to explore the Pareto optimal front obtained with the PLS2 model inversion, the computed MODR is formed by convex combinations of eight specific settings of Critical Method Parameters that have a mobile phase with percentages of water between 37 and 38 %, of methanol from 13 and 22 %, and of acetonitrile between 41 and 49 %, together with a flow rate between 1.47 and 1.50 mL min-1, and column temperature between 41.9 and 44.0 °C in their adequate combinations. All the chromatographic peaks are well resolved, with total time varying between 12.96 and 15.66 min inside the estimated MODR and the analytical method is accurate with CCß between 0.9 and 7.0 µg L-1 with probability of both false positive and false negative equal to 0.05.

Robustness of hyperspectral imaging and PLSR model predictions of intramuscular fat in lamb M. longissimus lumborum across several flocks and years.

The percentage of intramuscular fat content of lamb meat is a key index of consumer acceptability. Hyperspectral imaging is a potential technique for in-line measurements of intramuscular fat in fresh meat. However, little work has been conducted to investigate the robustness of hyperspectral imaging data and associated multivariate models over time. Fifteen trials consisting of eight independent flocks across five years were used to quantify robustness of partial least squares regression (PLSR) models developed using data collected with the same imaging system. Two models were developed; one using data from the first year of the trials, and a progressive model that cumulatively includes data in chronological order. The two models performed similarly, in terms of the coefficient of determination (R2), standard error of prediction (SEP) and bias, when experimental conditions were consistent. However, under varying imaging conditions, the progressive model was able to account for this variability resulting in higher R2 and lower SEP.

Modelling strawberry quality in a longitudinal study under the marketing concept of branding.

BACKGROUND: Marketing strategies, such as branding, redefine how consumers perceive quality and create new requirements related to season length and quality homogeneity, among others. For short-day (SD) strawberry cultivar brands, the commercial season is short due to a dependency on temperature and photoperiod. A plausible strategy to extend the commercialization period is to use different varieties within a single brand; however, this has led to inconsistent quality in other fruit crops. A form of quality assessment to evaluate the impact of a multi-varietal brand on sensory quality is a critical longitudinal study with several sources of variability, such as the inherent variation among assessors and fruit replicates that can affect the reliability of the results. Therefore, this study aimed to develop a methodology to assess the sensorial and physicochemical quality of strawberry brands in two contexts: a short-term season composed of two SD cultivars and a long-term season with one SD and one day-neutral (DN) cultivar. RESULTS: New statistical models are proposed in this study. An ANOVA mixed model with assessors and replicates as random terms and a multiple factor analysis highlighted a lack of homogeneity with regard to texture parameters and sourness, while partial least square models identified aroma and sweetness as the best quality indicators. CONCLUSIONS: This work has successfully illustrated a methodology that is capable of handling critical aspects of longitudinal studies by using univariate models that account for different sources of variability and constrained multivariate models to relate parameters with overall liking. A long-term brand is a viable solution to valorise strawberries, as parameter heterogeneity did not affect overall quality.

Risk Perception and PTSD Symptoms of Medical Staff Combating Against COVID-19: A PLS Structural Equation Model.

Medical staff were battling against coronavirus disease 2019 (COVID-19) at the expense of their physical and mental health, particularly at risk for posttraumatic stress disorder (PTSD). In this case, intervening PTSD of medical staff and preparing them for future outbreaks are important. Previous studies showed that perceived stress was related to the development of PTSD. Hence, in this study, the association between risk perception of medical staff and PTSD symptoms in COVID-19 and the potential links were explored. Three hundred four medical staff's exposure to COVID-19 patients, risk perception for working during COVID-19, PTSD symptoms, anxiety, and sleep quality were measured. Mediation analysis tested the indirect effects of anxiety and sleep quality on the relationship between risk perceptions and PTSD symptoms; 27.6% of participants were deemed as having probable PTSD diagnosis. Mediation analysis showed a significant chain-mediating effect of anxiety and sleep quality on the relationships between risk perceptions and PTSD symptoms; higher risk perceptions were related to increased anxiety, worsened sleep quality, and severe PTSD symptoms. Conclusively, medical staff have a high prevalence of PTSD symptoms after 3 months of COVID-19. Their PTSD symptoms were associated with the perceived risk level through the potential links with anxiety and sleep quality. Therefore, risk perception could be critical for our medical staff's responses to public health emergencies. It could be plausible to intervene in the perceived stress to alleviate aroused anxiety and improve sleep quality and thereby deter the development of PTSD.

A comparative study of ATR-FTIR, UV-visible and fluorescence spectroscopy combined with chemometrics for quantification of squalene in extra virgin olive oils.

Attenuated total reflectance-Fourier transform infrared (ATR-FTIR), ultraviolet-visible (UV-Vis), and fluorescence (FL) spectroscopic techniques, combined with partial least-square (PLS) regression with various spectral derivatization methods were tested for the quantitative determination of squalene content of extra virgin olive oils (EVOOs). A set of 90 calibration standards covering a wide range of squalene (3.25-12.54 mg/kg) was used to build up the calibration models. The root mean square error of calibration (RMSEC), the root mean square error of cross-validation (RMSECV), and the root mean square error of prediction (RMSEP) were calculated for evaluation of 18 different calibration models. 50 different brands of EVOOs, which are also analyzed in terms of quality indexes, fatty acid composition, and squalene concentration (3.25-12.54 g/kg) were used for checking the predictive capacities of the calibration models. The best predictions were achieved using normal spectra in FL spectroscopy with the lowest RMSEC of 0.1065, RMSEV of 0.1310, and RMSEP of 0.1500 in the spectral region 250-730 nm. Thus, FL spectroscopy combined with PLS regression is proposed as a fast, accurate, and environmentally friendly approach that can be efficiently used in determining squalene in EVOOs.

Hyperspectral-based Estimation of Leaf Nitrogen Content in Corn Using Optimal Selection of Multiple Spectral Variables.

Accurate and dynamic monitoring of crop nitrogen status is the basis of scientific decisions regarding fertilization. In this study, we compared and analyzed three types of spectral variables: Sensitive spectral bands, the position of spectral features, and typical hyperspectral vegetation indices. First, the Savitzky-Golay technique was used to smooth the original spectrum, following which three types of spectral parameters describing crop spectral characteristics were extracted. Next, the successive projections algorithm (SPA) was adopted to screen out the sensitive variable set from each type of parameters. Finally, partial least squares (PLS) regression and random forest (RF) algorithms were used to comprehensively compare and analyze the performance of different types of spectral variables for estimating corn leaf nitrogen content (LNC). The results show that the integrated variable set composed of the optimal ones screened by SPA from three types of variables had the best performance for LNC estimation by the validation data set, with the values of R2, root means square error (RMSE), and normalized root mean square error (NRMSE) of 0.77, 0.31, and 17.1%, and 0.55, 0.43, and 23.9% from PLS and RF, respectively. It indicates that the PLS model with optimally multitype spectral variables can provide better fits and be a more effective tool for evaluating corn LNC.

Media Disaster Reporting Effects on Public Risk Perception and Response to Escalating Tornado Warnings: A Natural Experiment.

Previous research has evaluated public risk perception and response to a natural hazards in various settings; however, most of these studies were conducted either with a single scenario or after a natural disaster struck. To better understand the dynamic relationships among affect, risk perception, and behavioral intentions related to natural disasters, the current study implements a simulation scenario with escalating weather intensity, and includes a natural experiment allowing comparison of public response before and after a severe tornado event with extensive coverage by the national media. The current study also manipulated the display of warning information, and investigated whether the warning system display format influences public response. Results indicate that (1) affect, risk perception, and behavioral intention escalated as weather conditions deteriorated, (2) responses at previous stages predicted responses at subsequent stages of storm progression, and (3) negative affect predicted risk perception. Moreover, risk perception and behavioral intention were heightened after exposure to the media coverage of an actual tornado disaster. However, the display format manipulation did not influence behavioral responses. The current study provides insight regarding public perception of predisaster warnings and the influence of exposure to media coverage of an actual disaster event.

Impact of blend properties on die filling during tableting.

Based on characterization of a wide range of fillers and APIs, thirty divergent blends were composed and subsequently compressed on a rotary tablet press, varying paddle speed and turret speed. The tablet weight variability was determined of 20 grab samples consisting of each 20 tablets. Additionally, the bulk residence time, ejection force, pre-compression displacement, main compression force, die fill fraction and feed frame fill fraction were determined during each run. Multivariate data analysis was applied to investigate the relation between the process parameters, blend characteristics, product and process responses. Blends with metoprolol tartrate as API showed high ejection forces. This behavior could be linked to the high wall friction value of metoprolol tartrate. The main responses related to the die filling could be predicted via a PLS model based on blend characteristics. Tablet weight variability was highly correlated with the variability on pre-compression displacement and main compression force. A good predictive model for tablet weight variability was obtained taking the porosity, wall friction angle, flowability, density, compressibility and permeability into account. Additionally, turret speed and paddle speed were included in the calibration of the model. The applied approach can save resources (material, time) during early drug product development.

Automatic untargeted metabolic profiling analysis coupled with Chemometrics for improving metabolite identification quality to enhance geographical origin discrimination capability.

Untargeted metabolic profiling analysis is employed to screen metabolites for specific purposes, such as geographical origin discrimination. However, the data analysis remains a challenging task. In this work, a new automatic untargeted metabolic profiling analysis coupled with a chemometric strategy was developed to improve the metabolite identification results and to enhance the geographical origin discrimination capability. Automatic untargeted metabolic profiling analysis with chemometrics (AuMPAC) was used to screen the total ion chromatographic (TIC) peaks that showed significant differences among the various geographical regions. Then, a chemometric peak resolution strategy is employed for the screened TIC peaks. The retrieved components were further analyzed using ANOVA, and those that showed significant differences were used to build a geographical origin discrimination model by using two-way encoding partial least squares. To demonstrate its performance, a geographical origin discrimination of flaxseed samples from six geographical regions in China was conducted, and 18 TIC peaks were screened. A total of 19 significant different metabolites were obtained after the peak resolution. The accuracy of the geographical origin discrimination was up to 98%. A comparison of the AuMPAC, AMDIS, and XCMS indicated that AuMPACobtained the best geographical origin discrimination results. In conclusion, AuMPAC provided another method for data analysis.

A prediction model of compressor with variable-geometry diffuser based on elliptic equation and partial least squares.

To achieve a much more extensive intake air flow range of the diesel engine, a variable-geometry compressor (VGC) is introduced into a turbocharged diesel engine. However, due to the variable diffuser vane angle (DVA), the prediction for the performance of the VGC becomes more difficult than for a normal compressor. In the present study, a prediction model comprising an elliptical equation and a PLS (partial least-squares) model was proposed to predict the performance of the VGC. The speed lines of the pressure ratio map and the efficiency map were fitted with the elliptical equation, and the coefficients of the elliptical equation were introduced into the PLS model to build the polynomial relationship between the coefficients and the relative speed, the DVA. Further, the maximal order of the polynomial was investigated in detail to reduce the number of sub-coefficients and achieve acceptable fit accuracy simultaneously. The prediction model was validated with sample data and in order to present the superiority of compressor performance prediction, the prediction results of this model were compared with those of the look-up table and back-propagation neural networks (BPNNs). The validation and comparison results show that the prediction accuracy of the new developed model is acceptable, and this model is much more suitable than the look-up table and the BPNN methods under the same condition in VGC performance prediction. Moreover, the new developed prediction model provides a novel and effective prediction solution for the VGC and can be used to improve the accuracy of the thermodynamic model for turbocharged diesel engines in the future.

Effect of Theobromine Consumption on Serum Lipoprotein Profiles in Apparently Healthy Humans with Low HDL-Cholesterol Concentrations.

Scope: Theobromine is a major active compound in cocoa with allegedly beneficial effect on high-density-lipoprotein-cholesterol (HDL-CH). We have investigated the effect of theobromine (TB) consumption on the concentrations of triglyceride (TG) and cholesterol (CH) in various lipoprotein (LP) subclasses. Methods: In a randomized, double-blind, placebo-controlled, cross-over study, 44 apparently healthy women and men (age: 60 ± 6 years, BMI: 29 ± 3 kg/m2) with low baseline HDL-CH concentrations consumed a drink supplemented with 500 mg/d theobromine for 4 weeks. TG and CH concentrations in 15 LP subclasses were predicted from diffusion-edited 1H NMR spectra of fasting serum. Results: The LP phenotype of the subjects was characterized by low CH concentrations in the large HDL particles and high TG concentrations in large VLDL and chylomicron (CM) particles, which clearly differed from a LP phenotype of subjects with normal HDL-CH. TB only reduced CH concentrations in the LDL particles by 3.64 and 6.79%, but had no effect on TG and CH in any of the HDL, VLDL and CM subclasses. Conclusion: TB was not effective on HDL-CH in subjects with a LP phenotype characterized by low HDL-CH and high TG in VLDL.