Kamis, 06 Oktober 2016

logistics distribution controlling

Fusing Data Mining, Machine Learning and

Traditional Statistics to Detect Biomarkers

Associated with Depression

Joanna F. Dipnall1,2*, Julie A. Pasco1,3,4,5, Michael Berk1,5,6,7,8, Lana J. Williams1,

Seetal Dodd1,5,6, Felice N. Jacka1,6,9,10, Denny Meyer2

1 IMPACT Strategic Research Centre, School of Medicine, Deakin University, Geelong, VIC, Australia,

2 Department of Statistics, Data Science and Epidemiology, Swinburne University of Technology,

Melbourne, VIC, Australia, 3 Department of Medicine, The University of Melbourne, St Albans, VIC,

Australia, 4 Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, VIC,

Australia, 5 University Hospital Geelong, Barwon Health, Geelong, VIC, Australia, 6 Department of

Psychiatry, The University of Melbourne, Parkville, VIC, Australia, 7 Florey Institute of Neuroscience and

Mental Health, Parkville, VIC, Australia, 8 Orygen, the National Centre of Excellence in Youth Mental Health,

Parkville, VIC, Australia, 9 Centre for Adolescent Health, Murdoch Children’s Research Institute, Melbourne,

Australia, 10 Black Dog Institute, Sydney, Australia

 These authors contributed equally to this work.

* jdipnall@deakin.edu.au

Abstract

Background

Atheoretical large-scale data mining techniques using machine learning algorithms have

promise in the analysis of large epidemiological datasets. This study illustrates the use of a

hybrid methodology for variable selection that took account of missing data and complex

survey design to identify key biomarkers associated with depression from a large epidemio-
logical study.

Methods

The study used a three-step methodology amalgamating multiple imputation, a machine

learning boosted regression algorithm and logistic regression, to identify key biomarkers

associated with depression in the National Health and Nutrition Examination Study (2009–

2010). Depression was measured using the Patient Health Questionnaire-9 and 67 bio-
markers were analysed. Covariates in this study included gender, age, race, smoking, food

security, Poverty Income Ratio, Body Mass Index, physical activity, alcohol use, medical

conditions and medications. The final imputed weighted multiple logistic regression model

included possible confounders and moderators.

Results

After the creation of 20 imputation data sets from multiple chained regression sequences,

machine learning boosted regression initially identified 21 biomarkers associated with

depression. Using traditional logistic regression methods, including controlling for possible

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 1 / 23

OPEN ACCESS

Citation: Dipnall JF, Pasco JA, Berk M, Williams LJ,

Dodd S, Jacka FN, et al. (2016) Fusing Data Mining,

Machine Learning and Traditional Statistics to Detect

Biomarkers Associated with Depression. PLoS ONE

11(2): e0148195. doi:10.1371/journal.pone.0148195

Editor: Mansour Ebrahimi, Qom University, ISLAMIC

REPUBLIC OF IRAN

Received: September 30, 2015

Accepted: January 14, 2016

Published: February 5, 2016

Copyright: © 2016 Dipnall et al. This is an open

access article distributed under the terms of the

Creative Commons Attribution License, which permits

unrestricted use, distribution, and reproduction in any

medium, provided the original author and source are

credited.

Data Availability Statement: Original and cleaned

data for the NHANES data used in this study is open

access and located at the URL http://wwwn.cdc.gov/

Nchs/Nhanes/Search/nhanes09_10.aspx A Stata

syntax file containing a template for enabling those to

implement this methodology on other data files has

been provided as part of the Supplementary

information. This template also explains the NHANES

biomarker variables used for this study and

references key NHANES analytical information.

Funding: Michael Berk is supported by a NHMRC

Senior Principal Research Fellowship 1059660 and

Lana J Williams is supported by a NHMRC Career

confounders and moderators, a final set of three biomarkers were selected. The final three

biomarkers from the novel hybrid variable selection methodology were red cell distribution

width (OR 1.15; 95% CI 1.01, 1.30), serum glucose (OR 1.01; 95% CI 1.00, 1.01) and total

bilirubin (OR 0.12; 95% CI 0.05, 0.28). Significant interactions were found between total bili-
rubin with Mexican American/Hispanic group (p = 0.016), and current smokers (p<0.001).

Conclusion

The systematic use of a hybrid methodology for variable selection, fusing data mining tech-
niques using a machine learning algorithm with traditional statistical modelling, accounted

for missing data and complex survey sampling methodology and was demonstrated to be a

useful tool for detecting three biomarkers associated with depression for future hypothesis

generation: red cell distribution width, serum glucose and total bilirubin.

Background

Over the last two decades there has been a steady rise in the use of data mining techniques

across a number of disciplines. Data mining incorporates a path to knowledge discovery and is

a meaningful process for discovering patterns in data by exploring and modelling large quanti-
ties of data [1, 2]. The distinction between statistics and data mining has been attributed to the

nature of the analysis; statistics deals with primary analysis, whereas data mining deals with

secondary analysis [3] that learns from data [4]. Data mining incorporates machine-learning

algorithms to learn, extract and identify useful information and subsequent knowledge from

large databases [2].

For a number of years, data mining techniques incorporating machine learning algorithms

have been used for ‘big data’ analytics in both marketing and finance (i.e. cost saving, sales

opportunity) [5]. Supervised machine learning techniques have been successfully used in these

industries for feature or variable reduction to produce highly predictive models [6]. However,

it has only been over the last 10 years that data mining techniques have been used in medical

research, primarily in neuroscience and biomedicine [7, 8]. More recently, psychiatry has

begun to utilize the benefits of these techniques to gain further insight into the genetic makeup

of mental illness [9]. However, the implementation of data mining techniques in large epidemi-
ological studies has not been fully explored.

Epidemiological observational studies are based on a particular study population, often fol-
lowed over a period of time, and usually involving no intervention other than the administra-
tion of questionnaires and the carrying out of medical and laboratory or biomarker

examinations. These studies have been used to quantify prevalence and risk factors for diseases

within the population [10]. Sample sizes for these types of studies often comprise of some thou-
sands of individuals. Incorporating a methodology involving data mining techniques using a

machine learning algorithm with traditional statistics for variable selection in these studies

would augment the effective knowledge discovery processes of data mining and rigors of

machine learning algorithms with the well-established metrics of traditional statistics. A hybrid

methodology as such could take account of common analytical issues associated with these

types of studies, such as large numbers of variables, a complex survey design and missing data.

For these reasons, the aim of this study was to develop a systematic and sound hybrid

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 2 / 23

Development Fellowship 1064272. The funders had

no role in study design, data collection and analysis,

decision to publish, or preparation of the manuscript.

Competing Interests: JFD has no conflicts of

interest. JAP has recently received grant/research

support from the National Health and Medical

Research Council (NHMRC), BUPA Foundation,

Amgen, GlaxoSmithKline, Osteoporosis Australia,

Barwon Health and Deakin University. MB has

received Grant/Research Support from the NIH,

Cooperative Research Centre, Simons Autism

Foundation, Cancer Council of Victoria, Stanley

Medical Research Foundation, MBF, NHMRC,

Beyond Blue, Rotary Health, Geelong Medical

Research Foundation, Bristol Myers Squibb, Eli Lilly,

Glaxo SmithKline, Meat and Livestock Board,

Organon, Novartis, Mayne Pharma, Servier and

Woolworths, has been a speaker for Astra Zeneca,

Bristol Myers Squibb, Eli Lilly, Glaxo SmithKline,

Janssen Cilag, Lundbeck, Merck, Pfizer, Sanofi

Synthelabo, Servier, Solvay and Wyeth, and served

as a consultant to Astra Zeneca, Bioadvantex, Bristol

Myers Squibb, Eli Lilly, Glaxo SmithKline, Janssen

Cilag, Lundbeck Merck and Servier. Drs Copolov, MB

and Bush are co-inventors of provisional patent

02799377.3-2107-AU02 “Modulation of physiological

process and agents useful for same”. MB and Laupu

are co-authors of provisional patent 2014900627

“Modulation of diseases of the central nervous

system and related disorders”. MB is supported by a

NHMRC Senior Principal Research Fellowship

1059660. LJW is supported by a NHMRC Career

Development Fellowship 1064272. SD has received

grants/research support from the Stanley Medical

Research Institute, NHMRC, Beyond Blue, ARHRF,

Simons Foundation, Geelong Medical Research

Foundation, Fondation FondaMental, Eli Lilly, Glaxo

SmithKline, Organon, Mayne Pharma and Servier,

speaker’s fees from Eli Lilly, advisory board fees from

Eli Lilly and Novartis, and conference travel support

from Servier. FNJ has received Grant/Research

support from the Brain and Behaviour Research

Institute, the National Health and Medical Research

Council (NHMRC), Australian Rotary Health, the

Geelong Medical Research Foundation, the Ian

Potter Foundation, Eli Lilly, the Meat and Livestock

Board and The University of Melbourne and has

received speakers honoraria from Sanofi-Synthelabo,

Janssen Cilag, Servier, Pfizer, Health Ed, Network

Nutrition, Angelini Farmaceutica, and Eli Lilly. This

does not alter the authors’ adherence to all the PLOS

ONE policies on sharing data and materials, as

detailed online in the guide for authors. DM has

recently received grant/research support from the

Australian research Council (ARC), Mental Illness

Research Fund (MIRF), Victorian Department of

methodology involving these elements to perform variable selection that would be appropriate

for use in large epidemiological studies.

To test the proposed hybrid methodology, data from a large cross section population based

U.S. epidemiology study was utilised to identify key biomarkers for depression. Depression is a

serious medical illness, with the World Health Organization estimating that 350 million people

worldwide are affected by depression, with depressive disorders ranking second in terms of

global disability burden and depression expected to be the number one health concern in both

developed and developing nations by 2020 [11–13].

Biomarkers are used in medicine as indicators of risk, diagnosis or trait, disease state or acu-
ity, stage of illness, treatment response and prognosis [14]. Identifying diagnostic biomarkers

of depression may help in its detection and furthermore, circumvent the onset. Despite inten-
sive approaches into the investigation of biomarkers in psychiatry, limitations of sensitivity

and specificity of single biomarkers have made it impractical to assess an individual’s clinical

situation and make determinations regarding diagnosis and prognosis on the basis of biomark-
ers. Thus, establishing key informative biomarkers would be of benefit to psychiatry [15]. The

nature of this type of data is such that missing data often needs to be accommodated using

multiple imputation and the complex survey sampling methodology taken into account. To

date, conventional variable selection methods do not deal with both these issues associated

with this type of data.

The proposed novel hybrid methodology involved a three-step approach, combining data

mining techniques using the machine learning algorithm of boosted regression and bagging,

with traditional statistical techniques. To take into account missing data and complex survey

sampling methodologies, this new three-step approach to variable selection was developed.

This methodology was effectively used to identify key biomarkers for depression in a large

cross-sectional population-based U.S. epidemiological study.

Methods

Study design and participants

Data from the National Health and Nutrition Examination Survey (NHANES) (2009–2010)

[16] were utilised. Relevant NHANES data files were downloaded from the website and inte-
grated using the Data Integration Protocol In Ten-Steps (DIPIT) [17]. NHANES is a cross-sec-
tional, population-based study of approximately 10,000 non-institutionalised U.S. civilians

aged 18 to 80 years, conducted in two-year blocks. A four-stage sampling methodology was

applied: counties; segments within counties; households within segments; and finally, individu-
als within households. Data were collected from 15 different locations across 50 states of Amer-
ica and the District of Columbia. In addition, oversampling of subgroups of the population of

particular public health interest was performed to increase the reliability and precision of pop-
ulation estimates [16]. Finally, subsamples for the mental health and laboratory components of

the survey were chosen, at random, with a sampling frame especially designed to reduce

respondent fatigue and help scheduling in the biomarker collection [16, 18].

Over 250 biomarkers were available for 5,546 participants. Duplicate biomarkers (n = 75),

biomarkers with a high missing count predominantly due to the sampling protocol (n = 94)

and low incidence (n = 26) were excluded from the analysis. The final set consisted of 67 bio-
markers for inclusion. Of the 5,546 participants, 5.6% were excluded from the analysis due to

having six or more missing data values across the 67 biomarkers, and a further six outlier cases

were removed to enable the multiple imputation to converge. The final sample size included in

this research study was 5,227.

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 3 / 23

Justice, Beyond Blue, Swinburne University of

Technology.

Abbreviations: BMI, Body Mass Index; DIPIT, Data

Integration Protocol In Ten-Steps; DSM-IV, Diagnostic

and Statistical Manual of Mental Disorder Fourth

Edition; FSSM, Food Security Survey Module; MDD,

Major Depressive Disorder; MEC, Mobile

Examination Center; MART, Multiple Additive

Regression Trees; NCHS, National Center for Health

Statistics; NHANES, National Health and Nutrition

Examination Survey; PHQ-9, Patient Health

Questonnaire-9; WHO, World Health Organization

(WHO) World Health Organization.

NHANES received approval from the National Center for Health Statistics (NCHS)

research ethics review board and informed consent was obtained for all participants. Use of

data from the NHANES 2009–2010 is approved by the National Center for Health Statistics

(NCHS) Research Ethics Review Board (ERB) Approval for NHANES 2009–2010 (Continua-
tion of Protocol #2005–06).

Study Measurements

The Patient Health Questonnaire-9 (PHQ-9) [19] was used to assess depressive symptoms

(depression). The PHQ-9 is a well-validated, self-report tool for detecting and monitoring

depression, with good concordance with a clinical diagnosis of major depressive disorder

(MDD) [20]. Items assess the presence of nine Diagnostic and Statistical Manual of Mental

Disorder Fourth Edition (DSM-IV) depression symptoms over the past two weeks, and are

scored on a four-point scale indicating the degree of severity from 0 (not at all) to 3 (nearly

every day). Items were then summed to form a total severity score ranging from 0 to 27 where

those with a total score of 10 or more were considered depressed (i.e. moderately to severely

depressed) [21].

Health report status was measured as an ordinal five-point rating scale (excellent, very

good, good, fair, poor) and incorporated as predictor in the multiple imputation models as an

indicator of general health.

Blood and urine samples were collected in the Mobile Examination Center (MEC) and

shipped weekly for laboratory analyses. Specific laboratory techniques for each test are avail-
able from the NHANES Laboratory Procedures Manual [22].

Demographic variables were self-reported and included: age, gender, and race (collapsed

into four groups: Mexican American and other Hispanic, Non-Hispanic White, Non-Hispanic

Black, and other).

Current and past smoking status was determined from self-reported questions from the

smoking cigarette use component from NHANES. Smoking status was categorized into those

who were never, former or current smokers. Physical activity was grouped into active (low to

high activity) and not active from the physical activity component from NHANES. Alcohol

regularity per month was calculated from the self-report question regarding how often alcohol

was drunk over the past 12 months, calculated from the weekly, monthly and yearly figures.

Food insecurity, a measure relating to the limited or uncertain access to food due to inade-
quate financial resources, was determined from the 10-item Food Security Survey Module

(FSSM) [23] and categorized into two groups: full food security (no affirmative response in any

of the 10 items); and food insecurity (at least one affirmative responses).

Body measurements were collected in the MEC by trained health technicians. Weight and

height were measured and body mass index (BMI) calculated as weight/height2 (kg/m2

). BMI

was categorised as underweight (BMI < 18.5 kg/m2

), normal (BMI between 18.5 to under 25

kg/m2

), overweight (between 25 to under 30 kg/m2

) and obese (BMI 30kg/m2

) [24].

Participants were identified as having Type 2 diabetes if they (i) reported that either a doctor

or health professional had told them they had diabetes or sugar diabetes (other than during

pregnancy); (ii) reported the use of hyperglycaemic agents, with the medication bottle seen by

the interviewer [25]; and/or (iii) diabetes based on their fasting blood glucose >126 mg/dL (7.0

mmol/L) [26–29] and glycated haemoglobin level of >6.5% [30–32].

Participants were identified as having cardiovascular disease using the self-report questions

regarding past history of heart conditions (i.e. congestive heart failure, coronary heart disease,

angina/angina pectoris, heart attack, stroke) which have been shown to have high concordance

with laboratory measures for cardiovascular disease [33, 34].

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 4 / 23

Participants were also identified as having the inflammatory conditions of arthritis and can-
cer or malignancy using the self-report questions.

Participants were asked to bring currently used medications to the assessment. The Lexicon

Plus1 proprietary comprehensive database of Cerner Multum, Inc. was used to assist with data

collection, data editing and release of the NHANES medication data. All prescription and some

non-prescription drug products available in the U.S. were coded using the Lexicon Plus medi-
cation database. For the current analyses, Lexicon Plus medication categories with higher than

2% incidence were included: anti-infective, cardiovascular agents, central nervous system

agents, coagulation modifiers, gastrointestinal agents, hormones/hormone modifiers, meta-
bolic agents, nutritional products, psychotherapeutic agents, respiratory agents, and topical

agents.

Methodology

The proposed hybrid methodology was applied to this study to select key biomarkers associ-
ated with depression from the NHANES data set (Fig 1).

STEP 1: Multiple Imputation (MI) for Missing Data

In many ‘big data’ situations, missing data are not an issue due to the large volume of observa-
tions and variables or features available. In contrast, missing data in studies with small sample

sizes can influence the results greatly [35, 36]. There have been a number of missing data pro-
cedures suggested in the literature over the last decades: listwise deletion; pairwise deletion;

mean substitution; regression imputation; Maximum Likelihood (ML) estimation [37, 38].

However, most of these methods can only be used when there is no pattern for the missing

data. The choice of method used for dealing with missing data is often less important when the

proportion of missing data is less than 5% [39]. However, it is not unusual for the proportion

of missing data in large epidemiological studies to exceed this percentage, thereby potentially

reducing statistical power, producing biased parameters and increasing the risk of a Type I

error [35]. All these issues could be detrimental to traditional statistical multivariate regression

models.

Multiple imputation is a useful flexible strategy for addressing the missing value problem.

Multiple imputation is considered when the missingness is not totally random, depending on

observed or unobserved values. However, this method is applicable even when the pattern of

missing data is not random. Multiple imputation using chained sequences of equations is a

flexible imputation method with the ability to handle different variable types, as an imputation

model is generated for each variable with missing data [40]. This method is also referred to as

fully conditional specification [41] and sequential regression multivariate imputation [42]. A

selected imputation model generates ‘mi’ complete imputed sets of data. Analysis is then per-
formed and results are pooled. The number of ‘mi’ sets depends on the amount of missing

data, nature of the data and analysis model used. More than 50 imputed sets may be required

to obtain stable results [43], but between 5 to 20 has been considered appropriate to reduce the

imputation sampling error when fractions of missing data are low [44].

Prior to imputing the missing data, missing values are often tested for Missing At Random

(MAR) or Missing Completely At Random (MCAR) [38, 45]. MAR occurs when the probabil-
ity of being missing is dependent on the variables measured in the study, but independent of

those not measured in the study. To test MAR, logistic regression is often performed with a

missing data indicator created for each potential predictor (i.e. 1 for missing, 0 for non-miss-
ing). No significant relationship between the missingness indicators and the outcome of inter-
est suggests MAR. MCAR occurs when the probability of being missing is independent of all

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 5 / 23

the variables in the study (those measured and not measured) The assumption that the data are

missing completely at random (MCAR) can be assessed using Little's MCAR chi-squared test,

which can be applied to a large set of variables [46, 47], including all the predictors simulta-
neously. A significant result rejects the null hypothesis that the data is MCAR. Rejection of

MAR or MCAR confirms the need for multiple imputation for the missing values in the data.

As the highest missing data percentage for this study data set was above 5% and the assump-
tions of MAR and MCAR were rejected, multiple imputation was required. The multiple impu-
tation framework of inference was developed by Rubin [38] and implemented in Stata.

Multiple imputation was performed, with chained sequences of equations [40, 48] using all bio-
markers, age, race, and a measure of self-report health status, but run separately for males and

females due to possible differences in biomarker importance between the sexes [49]. A mix of

binary logistic and linear regression, using chained equations, was used, contingent on the

nature of the biomarker imputed. This study involved 20 chained sequences with the primary

data set of 5,227 observations. Thus, the combined original and imputed data sets contained

Fig 1. Hybrid Methodology Steps.

doi:10.1371/journal.pone.0148195.g001

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 6 / 23

109,767 observations. Convergence for each of the chained equations was achieved within an

acceptable 20 imputed data sets [44].

STEP 2: Technique for Initial Selection of Predictors (Boosted

Regression)

There are many potential statistical and machine learning algorithms available for variable

selection [2], but boosted regression is recognised as being particularly effective [50]. This tech-
nique has often been thought of as a ‘bucket’ or ensemble method (a method that averages over

multiple classifiers), and has gained popularity amongst mainstream statisticians due to its pre-
dictive accuracy and ease of interpretation [51, 52]. Boosted regression has been used in studies

involving animal ecology [53], for example to build predictive models for fish and coral metrics

[54], and to identify factors predicting the reproductive success of dominant pairs of clown

anemonefish [55]. It has also been used in the development of the Chicago Adolescent Depres-
sion Risk Assessment (CADRA) index from an array of baseline social and cognitive vulnera-
bility and mood risk factors [56].

Boosting was invented by Freund and Schapire for regression trees [57] and translated to

the logistic regression model by Hastie, Tibshirani, and Friedman [54]. One of the advantages

with boosted regression is that the algorithm can accommodate any type of variable (continu-
ous, categorical, censored), any type of likelihood loss function (Gaussian, Binomial, Poisson,

and robust), and can deal with highly correlated predictors [58]. In addition it can automati-
cally accommodate non-linearity and interaction effects between predictor variables. The

boosting regression algorithm fits a sequence of simple trees based on a set of data splitting

rules to provide a more accurate estimate of the outcome with over-fitting avoided using a vali-
dation data set and shrinkage. Each tree consists of a series of yes/no questions which is applied

to each observation in the training data. Predictions are obtained using the average of the out-
comes for the observations found in each terminal node. Residuals are computed as the differ-
ence between the true and outcome values. Each successive tree is built for the prediction of the

residuals of the preceding tree. Essentially the technique successively gives larger weights to

observations that are repeatedly misclassified.

The final classifier consists of a weighted average of previous classifiers, with lower weights

(shrinkage) for later trees in order to avoid over-fitting. Shrinkage is accomplished by adding a

parameter λ to the last regression tree of residuals, where λ = 1 corresponds to no shrinkage.

Typically λ is 0.1 or smaller, with λ = 0.01 or λ = 0.001 common. Smaller shrinkage values

require larger number of iterations which can be computationally demanding.

The steps in the boosted regression technique incorporating the shrinkage parameter:

• An initial guess for the predicted outcomes is created. Compute the residuals based on the

current model.

• For each regression tree:

• Fit a regression tree to the residuals

• Compute the predicted outcome of the residuals for each terminal node

• Use the regression tree to predict the new set of residuals

• Update the boosting regression model to reflect the current regression tree, applying the

shrinkage parameter to the last regression tree of residuals

• This model is applied to the validation data and if there is an improvement in predictive

accuracy over the results for the previous iteration the next iteration commences. If not,

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016

then training stops and no more iterations are executed and no more residual trees are

developed.

In order to improve the robustness of the final model without introducing bias ‘bagging’ was

used [59]. Bagging is sometimes referred to as bootstrap aggregating and was originally devel-
oped by Breiman as a method for creating multiple versions of a predictor to form an aggregated

predictor [60]. Combining this method with boosting, bagging randomly selects a proportion of

the residuals at each iteration to build a new tree. While not all observations are used in each

iteration, all observations are eventually used across all iterations. Thus, the tree analysis is run

on multiple similar datasets, and the aggregate results are used to construct the final model and

relevant statistics. Friedman recommends bagging with 50% of the database [59].

The final boosted regression tree model tends to be more robust than a single regression

tree model, enabling complex functions and interactions to be modelled without assumptions

[53]. Since no probability values (p-value) are produced from the boosted regression algorithm,

the relative importance of variables has been used to pick likely predictors [59]. Each variable is

assigned a relative importance percentage or contribution, where the sum of the standardized

values of all variable importances add to 100%. The importance is calculated from the number

of times a variable is selected for splitting, weighted by the squared improvement to the model

fit achieved from each such split, averaged over all trees. Higher values of importance indicate

stronger contributions to outcomes values. With a binary outcome measure the importance

represents the percentage of the log likelihood explained by each variable. For the proposed

methodology, the importance percentage for each predictor will be of particular interest for the

variable selection process.

The analysis approach for selecting the preliminary subset of predictors for the proposed

methodology is one of inclusion rather than exclusion. A potential predictor was selected based

on the relative importance percentage in the original data set and the average importance per-
centage statistics across the 20 imputed data sets. The cut-off relative importance percentage is

therefore relatively inclusive, ensuring that a reasonable percentage of the total log likelihood is

retained.

For the NHANES study, depression was a binary indicator. The boosted regression method

used the Multiple Additive Regression Trees (MART) boosting algorithm [54, 59] imple-
mented by Schonlau [61]. This method was performed on the original data set of 5,227 obser-
vations with missing data, then for each of the 20 imputed data sets. For each run, validation

was performed by randomly splitting each data set into 60% training and 40% validation ensur-
ing that additional boosting iterations were employed only when improved prediction accuracy

was achieved with the validation data set. After some initial testing, the final shrinkage parame-
ter used was 0.001 with the recommended 50% of the residuals used to fit each individual tree

(50% bagging) [59]. The maximum number of boosting interactions (i.e. number of terminal

nodes plus 1) allowed was 6, being marginally higher than the default (i.e. 5) and within the

recommended range by Hastie, Tibshirani and Friedman [54]. Finally, a random-number seed

was used to generate the same sequence of random numbers for bagging, ensuring that the

results could be reproduced. Biomarkers with the greatest relative importance for the predic-
tion of depression were identified.

To ensure the inclusion criterion was implemented and a reasonable proportion of the total

relative importance was retained, cut-off percentages of 4%, 3% and 2% were tested. It was

decided to use 2% to ensure more than 50% of the total relative importance was included in this

stage of variable selection. Thus, a biomarker was included in the initial biomarker selection if

its relative importance percentage in the original data set was higher than 2%, or its average

importance percentage statistics across the 20 imputed data sets yielded average of at least 2%.

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 8 / 23

STEP 3: Traditional statistical regression

At the heart of data mining is the concept of data splitting to address overfitting the data from

the machine learning techniques used. Data are randomly split into two groups: a training set

and a validation set. Thus, the imputed data set was split into training and validation data sets.

For the NHANES study, the imputed data set was split into approximately 50% training and

50% validation data for the traditional statistical analysis. Each set maintained the balance in

regard to depression levels and imputation sets.

The standard approach of univariate regressions for each of the potential predictors selected

from step two were then performed, with the variable of interest as the outcome variable. Pre-
dictors with the strongest relationship with the outcome for both these data sets were chosen

based on traditional statistical 95% confidence (i.e. p<0.05). As required for the NHANES

data, the study’s complex survey sampling methodology was taken into account using univari-
ate weighted logistic regressions. These regressions, were conducted for each selected bio-
marker from the machine learning boosted regression algorithm with depression as the

outcome. Biomarkers having a significant relationship with depression were chosen (i.e.

p<0.05) at the final variable selection stage.

Using the selected biomarkers from the univariate analysis, a traditional statistical multivariate

regression model was then estimated for both the training and validation data sets. Multicollinear-
ity and mediation relationships were tested and predictors removed accordingly. For the NHANES

data, it was important that this step took account of the complex four-staged survey design.

Finally, demographic and medical covariates and significant interactions were included in

this multivariate model to control for any confounders and account for any moderation effect

with models fitted using both the training and validation imputed data sets. The final model

was also fitted using the original set of observations (with missing values) to obtain an indica-
tive measure for goodness of fit and to ensure consistent direction and significance for the

important biomarkers identified using the primary data set and the combined data set with

imputation for missing values.

All statistical procedures were performed using Stata V14 software (StataCorp., 2014). A

Stata plugin was used for the boosted regression component of the analysis [61].

Results

Estimated statistics for each of the covariates and the significance of the relationship with

depression for each covariate is presented in Table 1. The complex survey design of NHANES

is taken into account in these estimates.

STEP 1: Multiple Imputation results

The impact of missing data across the 67 laboratory data was quantified [17]. The results from

the imputed weighted logistic regression with the missing data indicator for each biomarker as

the outcome and depression as the predictor are reported in the last column in Table 1. Signifi-
cant results for several biomarkers indicated that missing biomarker data affected the depres-
sion status (i.e. p<0.05), confirming that the data was not MAR. Little's test also provided

evidence against the assumption that all laboratory data were MCAR (p<0.001). The missing

data were therefore imputed using multiple imputation.

STEP 2: Machine Learning boosted regression results

Multiple imputation produced 20 separate data sets with distinct imputed missing values for

the boosted regression step. The relative importance statistics are reported in Table 2 for the

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 9 / 23

Table 1. Estimated covariate statistics.

Covariate Proportion Std Error 95% CI Low 95% CI High p-value

Depressed

Not Depressed 0.923 0.005 0.913 0.934

Depressed 0.077 0.005 0.066 0.087

Gender

Male 0.496 0.006 0.483 0.509 Reference

Female 0.504 0.006 0.491 0.517 <0.001

Age Group

18–34 years 0.329 0.009 0.309 0.348 Reference

35–44 years 0.210 0.008 0.193 0.226 0.457

45–54 years 0.223 0.007 0.207 0.238 0.117

55–64 years 0.176 0.006 0.162 0.189 0.272

65+ years 0.063 0.004 0.056 0.071 0.006

Race

Mexican American/Hispanic 0.147 0.031 0.082 0.213 0.010

Non-Hispanic White 0.678 0.035 0.602 0.753 Reference

Non-Hispanic Black 0.110 0.009 0.089 0.130 <0.001

Other 0.065 0.009 0.046 0.084 0.542

Smoking

Current Smoker 0.219 0.008 0.201 0.237 <0.001

Former Smoker 0.226 0.014 0.195 0.257 0.800

Never Smoked 0.555 0.019 0.515 0.595 Reference

Food Security

Full food security 0.782 0.013 0.755 0.809 <0.001

Food insecurity 0.218 0.013 0.191 0.245 Reference

BMI Category

Underweight 0.018 0.003 0.012 0.024 0.396

Normal 0.295 0.014 0.266 0.324 Reference

Overweight 0.331 0.011 0.307 0.355 0.358

Obese 0.356 0.011 0.333 0.379 0.035

Physical Activity

Low to high activity 0.451 0.018 0.411 0.490 Reference

No low to high activity 0.549 0.018 0.510 0.589 0.882

Diabetes Status

No Diabetes 0.910 0.005 0.900 0.920 Reference

Diabetes 0.090 0.005 0.080 0.100 0.001

Cardiovascular Disease Status

No Cardiovascular Disease 0.924 0.007 0.910 0.938 Reference

Cardiovascular Disease 0.076 0.007 0.062 0.090 0.009

Arthritis Status

No Arthritis 0.928 0.005 0.918 0.938 Reference

Arthritis 0.072 0.005 0.062 0.082 <0.001

Cancer or malignancy

No Cancer or malignancy 0.901 0.008 0.884 0.919 Reference

Cancer or malignancy 0.099 0.008 0.081 0.116 0.925

Use Central Nervous System

No 0.854 0.008 0.836 0.871 Reference

Medication(s)

(Continued)

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 10 / 23

Table 1. (Continued)

Yes 0.146 0.008 0.129 0.164 <0.001

Use Psychotherapeutic Agents

No 0.915 0.006 0.902 0.928 Reference

Yes 0.085 0.006 0.072 0.098 <0.001

Mean Std Error 95% CI Low 95% CI High p-value

Family Poverty Income Ratio (PIR) 3.013 0.044 2.918 3.108 <0.001

Number of times drink alcohol pm 6.235 0.290 5.615 6.856 0.005

Note: Multiple-imputation, survey estimation. Based on 20 imputations, primary N = 5,227. P-value indicates the significance of biomarker with depression.

doi:10.1371/journal.pone.0148195.t001

Table 2. Boosted regression statistics.

Biomarker Original data set Imputation sets 1 to 20

Mean Std Dev Min Max

T.gondii antibodies (IU/ml) 0.220 0.562 0.444 0.326 2.388

Blood lead (ug/dL) 1.482 1.658 0.066 1.537 1.753

Mercury, total (ug/L) 1.958 1.847 0.110 1.628 2.049

Mercury, inorganic (ug/L) 0.290 1.668 0.116 1.358 1.788

White blood cell count (1000 cells/uL) 1.243 1.126 0.065 1.000 1.277

Lymphocyte percent (%) 1.331 0.978 0.078 0.780 1.176

Monocyte percent (%) 1.996 1.595 0.172 1.371 1.904

Segmented neutrophils percent (%) 1.240 1.004 0.082 0.856 1.121

Eosinophils percent (%) 1.770 0.971 0.126 0.819 1.406

Basophils percent (%) 0.565 0.585 0.051 0.503 0.690

Lymphocyte number (1000 cells/uL) 0.754 0.912 0.171 0.691 1.559

Monocyte number (1000 cells/uL) 0.835 0.492 0.043 0.392 0.552

Segmented neutrophils num (1000 cell/uL) 1.138 1.103 0.075 0.950 1.282

Eosinophils number (1000 cells/uL) 0.229 0.129 0.026 0.052 0.157

Basophils number (1000 cells/uL) 0.069 0.084 0.012 0.056 0.100

Red blood cell count (million cells/uL) 0.829 1.083 0.148 0.898 1.595

Hemoglobin (g/dL) 1.717 2.942 0.152 2.667 3.210

Hematocrit (%) 0.847 2.798 0.148 2.533 3.103

Mean cell volume (fL) 0.494 0.943 0.082 0.853 1.154

Mean cell hemoglobin (pg) 0.848 1.307 0.045 1.251 1.432

Mean Cell hemoglobin concentration (MCHC) (g/dL) 5.139 3.380 0.105 3.195 3.610

Red cell distribution width (%) 3.423 1.918 0.105 1.685 2.058

Platelet count (1000 cells/uL) 2.395 2.632 0.112 2.443 2.842

Mean platelet volume (fL) 1.257 1.348 0.095 1.214 1.519

Blood cadmium (nmol/L) 5.306 4.159 0.136 3.908 4.447

Glycohemoglobin (%) 1.526 1.830 0.107 1.604 1.986

C-reactive protein(mg/dL) 1.877 2.264 0.107 2.024 2.405

Direct HDL-Cholesterol (mg/dL) 1.097 1.149 0.104 0.946 1.328

RBC folate (ng/mL) 0.536 0.779 0.097 0.588 1.021

Serum folate (ng/mL) 1.955 1.840 0.163 1.486 2.087

Cotinine (ng/mL) 2.011 1.982 0.277 1.542 2.739

Urinary Total NNAL (ng/mL) 3.226 5.579 0.496 4.544 6.459

(Continued)

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 11 / 23

original data set and across the 20 imputed data sets. Based on the selection criterion explained

above, 21 biomarkers were selected for the next stage of the analysis. The selected biomarkers

consistently explained more that 50% of the total relative importance for both the original data

set (53.85%) and for the mean values computed across the 20 imputed data sets (53.33%).

Table 2. (Continued)

Biomarker Original data set Imputation sets 1 to 20

Mean Std Dev Min Max

Albumin (g/dL) 0.914 0.543 0.050 0.459 0.662

Alanine aminotransferase ALT (U/L) 0.843 0.836 0.033 0.791 0.889

Aspartate aminotransferase AST (U/L) 0.727 0.577 0.053 0.462 0.694

Alkaline phosphotase (U/L) 2.992 1.966 0.086 1.825 2.184

Blood urea nitrogen (mg/dL) 0.840 0.875 0.088 0.718 1.061

Total calcium (mg/dL) 0.671 0.610 0.053 0.528 0.753

Cholesterol (mg/dL) 0.373 0.793 0.086 0.680 1.076

Bicarbonate (mmol/L) 3.769 1.840 0.089 1.661 2.035

Creatinine (mg/dL) 3.286 3.041 0.189 2.491 3.370

Gamma glutamyl transferase (U/L) 1.023 0.689 0.045 0.628 0.823

Glucose, serum (mg/dL) 4.911 2.487 0.168 2.140 2.742

Iron, refigerated (ug/dL) 1.752 1.443 0.088 1.288 1.559

Lactate dehydrogenase (U/L) 1.050 1.102 0.066 0.973 1.202

Phosphorus (mg/dL) 0.556 0.587 0.040 0.514 0.668

Total bilirubin (mg/dL) 3.366 2.555 0.291 1.841 2.853

Total protein (g/dL) 0.639 0.930 0.064 0.820 1.055

Triglycerides (mg/dL) 0.987 1.285 0.170 1.126 1.867

Uric acid (mg/dL) 2.598 2.401 0.116 2.232 2.675

Sodium (mmol/L) 0.957 0.299 0.034 0.242 0.366

Potassium (mmol/L) 0.654 0.664 0.059 0.559 0.778

Chloride (mmol/L) 2.810 1.629 0.065 1.460 1.726

Osmolality (mmol/Kg) 0.749 1.165 0.061 1.019 1.263

Globulin (g/dL) 0.501 0.679 0.062 0.529 0.778

Total Cholesterol (mg/dL) 0.457 0.609 0.058 0.488 0.714

Albumin, urine (ug/mL) 1.247 1.459 0.085 1.326 1.680

Creatinine, urine (umol/L) 0.963 0.992 0.108 0.769 1.193

First albumin creatinine ratio (mg/g) 0.723 0.787 0.118 0.636 1.110

Second albumin (ug/mL) 0.975 2.314 1.184 1.071 5.309

Second creatinine (mg/dL) 1.768 2.056 0.525 1.459 3.329

Second albumin creatinine ratio (mg/g) 1.176 3.321 0.825 1.783 5.326

The volume of urine collection #1 1.485 2.082 0.092 1.948 2.336

Urine #1 Flow Rate 3.262 2.921 0.720 1.743 4.456

Urine osmolality (mOsm/kg) 1.308 1.675 0.179 1.352 2.242

Hepatitis A Antibody 0.031 0.089 0.028 0.046 0.151

Hepatitis B surface antibody 0.034 0.051 0.018 0.023 0.082

Note: Highlighted indicate biomarker selected for univariate logistic regression. Validation on original plus each imputed data set. Random splitting of

60:40 training:validation, λ = 0.001, 50% bagging, 6 maximum number of boosting interactions. Original pseudo-R2 = 0.032, imputed data set pseudo-R2

ranged from 0.044 to 0.052. Variables selected at this step accounted for more than 50% of the total relative importance: original data was 53.85%; mean

of 20 imputation sets was 53.33%.

doi:10.1371/journal.pone.0148195.t002

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 12 / 23

STEP 3: Traditional statistical regression results

The training data yielded five biomarkers with significant univariate relationship with depres-
sion: hemoglobin (g/dL), Red cell distribution width (%), blood cadmium (nmol/L), cotinine

(ng/mL), and total bilirubin (mg/dL) (Table 3). The validation data confirmed the five bio-
markers from the training data with the addition of glucose, serum (mg/dL) (Table 3). Since

the p-value from the training data was 0.05 for this sixth biomarker it was decided to include

this variable in the final model.

A multiple logistic regression with these predictors suggested that the effect of hemoglobin

was fully mediated by total bilirubin and that cotinine was fully mediated by blood cadmium.

Both haemoglobin and cotinine were therefore excluded from the final model, leaving four key

biomarkers: red cell distribution width, blood cadmium, serum glucose and total bilirubin.

The direction and significance of the four biomarkers were tested with the combined

imputed data sets, allowing for a 50:50 random split between training and validation data. Con-
sistent results were obtained for the multiple logistic regression model in the training and vali-
dation data (Table 4).

The final model controlling for potential confounders and including significant interactions

are reported in Table 5.

Only Central nervous system medications and / or psychotherapeutic agents and diabetes

were included in the final multivariate model. The final validated model, using the original

Table 3. Univariate Logistic Regression statistics.

TRAINING VALIDATION

Biomarker Odds Ratio Std. Err. p-value CI Low CI High Odds Ratio Std. Err. p-value CI Low CI High

Hemoglobin (g/dL) 0.85 0.057 0.042 0.73 0.99 0.85 0.059 0.048 0.72 1.00

Hematocrit (%) 0.95 0.024 0.084 0.90 1.01 0.95 0.025 0.077 0.89 1.01

MCHC (g/dL) 0.85 0.115 0.261 0.63 1.15 0.89 0.131 0.454 0.65 1.23

Red cell distribution width (%) 1.20 0.080 0.024 1.03 1.40 1.20 0.079 0.023 1.03 1.40

Platelet count (1000 cells/uL) 1.00 0.002 0.058 1.00 1.01 1.00 0.002 0.083 1.00 1.01

Blood cadmium (nmol/L) 1.07 0.018 0.014 1.02 1.11 1.07 0.018 0.012 1.02 1.11

C-reactive protein(mg/dL) 1.18 0.106 0.091 0.97 1.44 1.15 0.098 0.141 0.95 1.38

Cotinine (ng/mL) 1.00 0.001 0.011 1.00 1.00 1.00 0.001 0.009 1.00 1.00

Urinary Total NNAL (ng/mL) 1.07 0.092 0.444 0.87 1.32 1.10 0.117 0.390 0.85 1.44

Alkaline phosphotase (U/L) 1.00 0.004 0.197 1.00 1.01 1.00 0.004 0.262 1.00 1.01

Bicarbonate (mmol/L) 0.93 0.045 0.156 0.83 1.03 0.92 0.044 0.133 0.83 1.03

Creatinine (mg/dL) 0.87 0.409 0.770 0.30 2.48 0.87 0.430 0.778 0.28 2.67

Glucose, serum (mg/dL) 1.00 0.002 0.050 1.00 1.01 1.01 0.002 0.039 1.00 1.01

Total bilirubin (mg/dL) 0.19 0.093 0.009 0.06 0.58 0.24 0.112 0.016 0.08 0.71

Uric acid (mg/dL) 0.97 0.069 0.655 0.83 1.13 0.94 0.062 0.345 0.81 1.08

Chloride (mmol/L) 0.98 0.039 0.700 0.90 1.08 0.98 0.038 0.624 0.90 1.07

Second albumin (ug/mL) 1.00 0.001 0.477 1.00 1.00 1.00 0.001 0.290 1.00 1.00

Second creatinine (mg/dL) 1.00 0.002 0.438 1.00 1.00 1.00 0.001 0.479 1.00 1.00

Second albumin creatinine ratio (mg/g) 1.00 0.000 0.516 1.00 1.00 1.00 0.000 0.315 1.00 1.00

The volume of urine collection #1 1.00 0.001 0.813 1.00 1.00 1.00 0.001 0.651 1.00 1.00

Urine #1 Flow Rate 0.89 0.125 0.438 0.65 1.23 0.86 0.136 0.382 0.60 1.24

Note: Bold Biomarker indicates selection. Multiple imputation logistic regression used taking account of the survey design of NHANES with 15 strata, 31

Primary Sampling Units (PSU).

doi:10.1371/journal.pone.0148195.t003

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 13 / 23

data set of observations including missing values, took account of the complex survey data of

NHANES and the test for goodness of fit was not significant (F(9,8) = 2.05, p = 0.163) indicat-
ing that the model provided a good fit to the data (76). In addition, the odds ratios and p-values

were consistent with the final model in Table 5.

In Table 5, three of the four biomarkers remained significant predictors of depression at the

95% level (p<0.05) after controlling for the potential confounders: red cell distribution width,

serum glucose and total bilirubin. Significant interactions were found between bilirubin and

the Mexican American / Hispanic group compared to the Non-Hispanic White group, and

those who were current smokers compared to those who have never smoked.

Blood cadmium was not found to be significant in the final model (p = 0.180). However, sig-
nificant interactions (p<0.05) were found between blood cadmium and the 45–54 age group

when compared to the younger 18–34 age group, and the diabetic group when compared to the

non-diabetic group.

Discussion

The novel proposed hybrid methodology of fusing data mining techniques using a boosted

regression machine learning algorithm with traditional statistical techniques offers an inte-
grated approach to variable selection from a large number of features for large epidemiological

studies. The methodology uses a systematic stepped approach that can be applied to large epi-
demiological population-based studies using complex survey designs.

The methodology ensures that missing data are quantified and addressed using multiple

imputation techniques which are appropriate even when the data is not missing at random;

however, should multiple imputation not be required, then the methodology is still appropriate

using only the original data set. The boosted regression ensures that variable importance is reli-
ably measured taking into account observations that are difficult to predict while at the same

time shrinking the effects of these difficult observations in order to avoid over-fitting. This

method ensures that multicollinearity does not obscure the influence of important predictors,

giving all variables a chance to shine, while taking into account any non-linearity or interaction

effects.

An inclusive choice of cut-off values ensures that the regression boosting only reduces the

number of predictors from 67 to 21. However, this is enough to remove the multicollinearity

problem and to allow a conventional logistic regression analysis that takes account of the com-
plex survey sampling design to be employed in order to further reduce the number of

Table 4. Final Four biomarkers from boosted regression.

Biomarker Training Validation

Odds

Ratio

Std.

Err.

p-

value

95% CI

Low

95% CI

High

Odds

Ratio

Std.

Err.

p-

value

95% CI

Low

95% CI

High

Red cell distribution

width

1.159 0.079 0.057 0.995 1.350 1.161 0.080 0.063 0.990 1.362

Blood cadmium (nmol/

L)

1.060 0.017 0.020 1.015 1.107 1.060 0.017 0.017 1.016 1.106

Glucose, serum (mg/dL) 1.005 0.002 0.066 1.000 1.009 1.005 0.002 0.051 1.000 1.010

Total bilirubin (mg/dL) 0.241 0.112 0.016 0.082 0.703 0.315 0.143 0.034 0.111 0.895

Constant 0.017 0.014 0.001 0.002 0.116 0.012 0.011 0.001 0.002 0.094

Note: Multiple imputation logistic regression using subpopulation based on a random split of approximately 50:50 train:validation (n = 2,590 train:

n = 2,637 validation).

doi:10.1371/journal.pone.0148195.t004

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 14 / 23

Table 5. Final Multivariate Logistic Regression.

Odds Ratio Std. Err. p-value 95% CI Low 95% CI High

Biomarkers

Red cell distribution width 1.145 0.067 0.037 1.009 1.298

Blood cadmium (nmol/L) 1.024 0.018 0.182 0.987 1.063

Glucose, serum (mg/dL) 1.005 0.002 0.009 1.001 1.008

Total bilirubin (mg/dL) 0.116 0.049 <0.001 0.047 0.284

Gender

Male (Reference) 1.000

Female 1.610 0.312 0.027 1.064 2.439

Age group (years)

18–34 (Reference) 1.000

35–44 1.287 0.346 0.364 0.724 2.288

45–54 1.993 0.419 0.005 1.271 3.124

55–64 1.475 0.398 0.171 0.828 2.627

65+ 0.660 0.421 0.525 0.169 2.585

Race

Non-Hispanic White (Reference) 1.000

Mexican Amer/Hispanic 0.409 0.150 0.029 0.186 0.898

Non-Hispanic Black 0.842 0.391 0.718 0.312 2.278

Other 1.552 1.527 0.662 0.189 12.743

Smoking

Never smoked (Reference) 1.000

Current smoker 0.382 0.162 0.039 0.154 0.946

Former smoker 0.506 0.326 0.308 0.127 2.013

Food Security

Food insecurity (Reference) 1.000

Full food security 0.492 0.093 0.002 0.328 0.737

Poverty Income Ratio (PIR) 0.787 0.058 0.006 0.671 0.923

BMI Category

Normal (Reference) 1.000

Underweight 2.497 1.630 0.182 0.617 10.101

Overweight 0.864 0.177 0.486 0.557 1.339

Obese 1.007 0.194 0.973 0.666 1.521

Inactivity

Active (Reference) 1.000

Inactive 1.104 0.135 0.431 0.850 1.433

Alcohol consumption per month 1.008 0.014 0.601 0.978 1.038

Diabetes

No (Reference) 1.000

Yes 0.798 0.210 0.407 0.454 1.403

Central Nervous System Meds

No (Reference) 1.000

Yes 2.279 0.292 <0.001 1.732 2.999

Psychotherapeutic Agents

No (Reference) 1.000

Yes 2.912 0.433 <0.001 2.118 4.002

Interactions:

Age Group by Blood cadmium

(Continued)

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 15 / 23

predictors from 21 to 4. Interestingly no biomarker interaction or polynomial terms were

found to be significant in the final multivariate logistic regression model making this a particu-
larly simple model which was also found to be robust in relation to important possible con-
founder variables. Three of the four important biomarkers remained significant to the 95%

level despite the inclusion of several confounder covariates and several important biomarker

interaction effects with these covariates.

The flexibility of the methodology is exemplified by its ability to take account of any sam-
pling methodology used for an epidemiological study (i.e. simple random samples through to

multi-staged complex survey sampling). This was demonstrated using the well-known

NHANES data set which used a complex four-staged sampling design encompassing clustered,

stratified and weighted data. There are currently no available stepwise or regularized regression

procedures which can be applied with complex sample designs making variable selection

impossible when there are initially 67 often highly correlated predictors. The boosted regres-
sion overcomes this problem by reducing the number of predictors to 21 and ensuring that

these predictors are not highly correlated, allowing the complex survey based logistic regression

model to be used to manually select the final three predictor variables.

Validation

There are a number of approaches for assessing models in data mining. This methodology

employs the criteria of validity, reliability, parsimony and usefulness to evaluate the effective-
ness of the proposed method.

A simulation was initially used to evaluate the effectiveness of the boosted regression and

importance percentage metric over a traditional backward stepwise variable selection method

to test the impact of a highly correlated predictor in the model such as the case with the

Table 5. (Continued)

Odds Ratio Std. Err. p-value 95% CI Low 95% CI High

18–34 (Reference) 1.000

35–44 0.974 0.024 0.308 0.924 1.027

45–54 0.950 0.019 0.025 0.910 0.993

55–64 0.946 0.029 0.088 0.887 1.009

65+ 0.966 0.062 0.606 0.842 1.110

Diabetes by Blood cadmium

No (Reference) 1.000

Yes 1.115 0.049 0.025 1.016 1.225

Race by Bilirubin

Non-Hispanic White (Reference) 1.000

Mexican Amer/Hispanic 3.946 1.988 0.016 1.342 11.603

Non-Hispanic Black 1.715 1.333 0.499 0.325 9.050

Other 0.484 0.628 0.585 0.030 7.777

Smoking by Bilirubin

Never smoked (Reference) 1.000

Current smoker 9.131 4.193 <0.001 3.418 24.398

Former smoker 2.676 2.396 0.290 0.394 18.187

Constant 0.040 0.030 0.001 0.008 0.200

Note: Multiple imputation logistic regression taking account of the complex survey design of NHANES with 15 strata, 31 PSUs. (n = 3,326).

doi:10.1371/journal.pone.0148195.t005

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 16 / 23

biomarker data in this study. A simple random sample of the same size as the original

NHANES data was generated (n = 5,227) for six potential predictors. The correlation for two

predictors was set to 0.8 to indicate multicollinearity, with the remaining correlations emulated

correlations found in the NHANES biomarker data. Two simulations were generated: one with

the same betas for all predictors; the other with varying betas. In both simulations the stepwise

regression failed to drop one of the two highly correlated predictors, indicating possible estima-
tion instability and structural misspecification [62]. In contrast, the boosted regression down-
graded the importance of one of the two highly collinear predictors and produced importance

rankings that reflected the beta values used for the other variables. This validates the use of

boosted regression for variable selection over stepwise regression.

To further assess the validity of using boosted regression, a lasso regression was used as a

comparative machine learning technique for variable selection [63, 64]. This method is similar

to boosted regression in terms of its ability to handle a large number of predictors. Using a

L1-norm penalty for shrinkage in order to prevent the problem of overfitting, the lasso regres-
sion algorithm was employed in the R statistical software using the glmnet package [65]. Con-
sistent with the boosted regression step, the algorithm was performed on the original data and

20 imputed data sets and included cross-validation on 60:40 train:test data. The lasso regres-
sion validated the results of the boosted regression analysis but provided a much less parsimo-
nious solution and it was more difficult to combine the variable selection results across the

imputed data sets. Results from the lasso analysis contained non-zero coefficients only for the

selected variables, making the number of selections a suitable variable importance measure for

each variable. Two of the final three biomarkers from the proposed methodology were chosen

by the lasso regression in at least 20 of the 21 data sets, and the other was in 16 of the 21 data

sets (Table 6). However, for the lasso regression the number of selected variables varied

between 18 and 34 for these 21 data sets, with an average of 25 variables selected, making the

lasso regression much less parsimonious than the boosted regression method.

Running a multiple regression directly on the 21 variables selected from the boosted regres-
sion on the training and validation data sets, without first using univariate regressions to reduce

Table 6. Top 15 biomarkers selected from lasso regression.

Biomarker Frequency

Blood cadmium (nmol/L) 21

Blood urea nitrogen (mg/dL) 21

Glucose, serum (mg/dL) 21

Blood lead (ug/dL) 20

Cotinine (ng/mL) 20

Total bilirubin (mg/dL) 20

Mercury, total (ug/L) 20

Platelet count (1000 cells/uL) 19

Mercury, inorganic (ug/L) 18

Globulin (g/dL) 18

Red blood cell count (million cells/uL) 17

Red cell distribution width (%) 16

Albumin (g/dL) 16

Phosphorus (mg/dL) 16

Direct HDL-Cholesterol (mg/dL) 15

Note: Bold represents the final 3 biomarkers selected from proposed methodology.

doi:10.1371/journal.pone.0148195.t006

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 17 / 23

the number of variables further, produced some unstable results across the training and valida-
tion data sets. The odds ratio for Haemoglobin differed drastically between the two models

which was consistent with haemoglobin having singularity with haematocrit (r = 0.968) [66]. In

addition, predictor significance was inconsistent: total bilirubin was the only significant predictor

in the training model and no predictors were significant in the validation model. Finally, the mul-
tiple regression models did not highlight mediation effects as significant univariate relationships

with the outcome were hidden when all predictors were included in the multivariate models.

Results are considered reliable if the same variables are selected regardless of the data supplied.

At step 3 this methodology employs the robust data mining technique of randomly splitting the

data into testing and validation data sets at multiple stages of the traditional statistical regression

to ensure the variables selected and significance are reliable. In addition, at step 2, validation was

performed for each run during the boosted regression by randomly splitting each data set into

training and validation sets to make sure that boosting iterations were employed only when

improved prediction accuracy was achieved with the validation data set. Usefulness relates to

evaluating if the model provides useful information. This hybrid methodology identified three

biomarkers associated with depression from an initial set of 67 in the NHANES data set: red cell

distribution width, serum glucose, and total bilirubin. The direction of the relationship between

depression and the final three biomarkers are broadly concordant with current literature.

Red cell distribution width is a marker of the variability in size of red blood cells or erythro-
cytes and used for differential diagnosis of anaemia, especially iron deficiency anaemia. This

biomarker has been found as a predictor of mortality in the general population [67] and other

conditions such as cardiovascular complications [68], Alzheimer’s disease [69] and diabetes

[70]. In addition, this biomarker has been investigated as a substitute marker for inflammatory

conditions such as inflammatory cancers [71, 72]. Many previous studies have reported the

association between inflammatory markers and depression [73, 74], with depression found to

be comorbid with both diabetes [25, 75] and cardiovascular diseases [76, 77].

Fasting plasma or serum glucose tests are used to screen for and diagnose diabetes and pre-
diabetes and to monitor for hyperglycaemia or hypoglycaemia. Diabetes is often comorbid

with major depression [78, 79], but has a complex bidirectional relationship [75]. An associa-
tion between depression and glucose utilisation dysfunction has been documented [80].

Bilirubin has been reported to be an antioxidant [81, 82]. Depression has been found to be

associated with oxidative stress [83] and mild to moderately elevated serum bilirubin levels are

associated with better outcome in diseases involving oxidative stress [82]. The significant inter-
action found between bilirubin and smoking status is consistent with previous research indicat-
ing that smoking is associated with decreased serum bilirubin concentrations [84, 85].

Hemoglobin was found to be fully mediated by total bilirubin. Red blood cells are continu-
ously being broken down, with hemoglobin splitting into globin (protein), iron and heme. The

heme initially breaks apart into biliverdin which is reduced to bilirubin. Cotinine was found to

be fully mediated by blood cadmium.

Strengths and Limitations

The strength of this hybrid methodology over other variable selection methods is the potential

to adequately handle missing data and complex survey samples using a sound and systematic

multi-stepped approach to variable selection. The application of the data mining knowledge

discovery process to large epidemiological studies allows researchers to include a large array of

data (i.e. variables and observations) to be investigated to generate hypotheses that may have

been otherwise overlooked. This is probably true for this NHANES study, particularly in regard

to the total bilirubin finding.

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 18 / 23

The data mining method of splitting data files into training and validation minimises issues

of overfitting that is often problematic in traditional statistical techniques with a large number

of predictors. The boosted machine learning technique can accommodate different types of

variables and has been found to have high predictive accuracy, with shrinkage also used to

avoid over-fitting. The iterative learning nature of the algorithm provides researchers with rea-
sonable confidence in the results with its boosted handling of residuals at each iteration. The

relative importance measure produced by this technique has been demonstrated to be more

effective than the traditional coefficient measures produced by lasso regularized regression and

stepwise regression.

A limitation of the methodology is the potential computing power required to perform the

machine learning techniques when implementing a small shrinkage parameter. In addition,

this study implemented the recommended bagging and number of iterations for the machine

learning boosted regression algorithm, but it may be appropriate in the future to run the algo-
rithm on a number of different bagging percentages and number of iterations.

A limitation of this methodology is the complexity of the implementing a multi-stepped

variable selection approach compared to a simpler single-stepped variable selection approach.

However, unlike simpler single-stepped variable selection procedures such as stepwise regres-
sion and regularized regression this multi-stepped method can accommodate missing data

using multiple imputation combined with complex survey designs.

The NHANES study used in the example for this hybrid methodology is a large cross-sec-
tional study that contains both missing data and utilises a complex four-stage sampling meth-
odology. The limitations of this type of data restrict the ability to infer the direction of the

relationship between the key biomarkers and depression. The self-report instrument of depres-
sion, the PHQ-9, may have missed less severe cases of depression [19, 21]. It is also recognised

that the imbalance of depressive symptoms in the data set may have resulted in a prediction

bias towards the major classes [86].

Conclusion

An amalgamation of data mining techniques using a machine learning algorithm with tradi-
tional statistical techniques provided an effective systematic approach to variable selection in a

large epidemiological study, detecting three biomarkers associated with depression for future

hypothesis generation: red cell distribution width, serum glucose and total bilirubin. The use of

this novel methodology yielded results concordant with previous research, taking account of

the missing data and complex survey design of NHANES. This methodology highlights the

effectiveness of implementing a hybrid of big data and small sample techniques for variable

and potential hypothesis generation.

Supporting Information

S1 Stata Syntax File. Hybrid Methodology Template.

(TXT)

Acknowledgments

MB is supported by a NHMRC Senior Principal Research Fellowship 1059660 and LJW is sup-
ported by a NHMRC Career Development Fellowship 1064272.

The authors would like to thank the referees and editors of this issue for their valuable and

insightful comments and suggestions that have improved this paper.

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 19 / 23

Author Contributions

Analyzed the data: JFD. Wrote the paper: JFD JAP MB LJW SD FNJ DM. Designed and per-
formed the statistical methodology and analysis: JFD.

References

1. First International Conference on Knowledge Discovery and Data Mining (KDD-95) Montreal1995.

Available from: http://www.sigkdd.org/kdd/1995/.

2. Witten IH, Frank E, Hall MA. Data Mining: Practical Machine Learning Tools and Techniques: Practical

Machine Learning Tools and Techniques: Morgan Kaufmann; 2011.

3. Yoo I, Alafaireet P, Marinov M, Pena-Hernandez K, Gopidi R, Chang J- F, et al. Data Mining in Health-
care and Biomedicine: A Survey of the Literature. Journal of Medical Systems. 2012; 36(4):2431–48.

4. Friedman JH. Data Mining and Statistics: What's the connection? Computing Science and Statistics.

1998:3–9.

5. Linoff GS, Berry MJ. Data Mining Techniques: For Marketing, Sales, And Customer Relationship Man-
agement Author: Gordon S. Linoff, Michael J. Be. 2011.

6. Zhang D, Zhou L. Discovering golden nuggets: data mining in financial application. Systems, Man, and

Cybernetics, Part C: Applications and Reviews, IEEE Transactions on. 2004; 34(4):513–22.

7. Chaovalitwongse W, Pardalos PM, Xanthopoulos P. Computational Neuroscience: Springer; 2010.

8. Seref O, Kundakcioglu OE, Pardalos PM. Data mining, systems analysis, and optimization in biomedi-
cine: American Institute of Physics Inc.; 2007.

9. Tovar D, Cornejo E, Xanthopoulos P, Guarracino MR, Pardalos PM. Data mining in psychiatric

research. Methods in Molecular Biology (Clifton, NJ). 2012; 829:593–603.

10. Pearce N. Classification of epidemiological study designs. International journal of epidemiology. 2012;

41(2):393–7. doi: 10.1093/ije/dys049 PMID: 22493323

11. Lim SS, Vos T, Flaxman AD, Danaei G, Shibuya K, Adair-Rohani H, et al. A comparative risk assess-
ment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions,

1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. The lancet. 2013; 380

(9859):2224–60.

12. Murray CJ, Lopez AD. Measuring the global burden of disease. N Engl J Med. 2013; 369(5):448–57.

doi: 10.1056/NEJMra1201534 PMID: 23902484

13. Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLoS

Med. 2006; 3(11):e442. PMID: 17132052

14. Davis J, Maes M, Andreazza A, McGrath J, Tye S, Berk M. Towards a classification of biomarkers of

neuropsychiatric disease: from encompass to compass. Mol Psychiatry. 2015; 20(2):152–3. doi: 10.

1038/mp.2014.139 PMID: 25349167

15. Boksa P. A way forward for research on biomarkers for psychiatric disorders. Journal of psychiatry &

neuroscience: JPN. 2013; 38(2):75.

16. Centers for Disease Control and Prevention National Center for Health Statistics. National Health and

Nutrition Examination Survey: Analytic Guidelines, 1999–2010 2013.

17. Dipnall JF, Berk M, Jacka FN, Williams LJ, Dodd S, Pasco JA. Data Integration Protocol In Ten-steps

(DIPIT): A new standard for medical researchers. Methods. 2014.

18. Centers for Disease Control and Prevention National Center for Health Statistics. National Health and

Nutrition Examination Survey 1999–2014 Survey Content Brochure. 2014.

19. Kroenke K, Spitzer RL. The PHQ-9: a new depression diagnostic and severity measure. Psychiatric

Annals. 2002; 32(9):509–15.

20. Kroenke K, Spitzer RL, Williams JB, Löwe B. The patient health questionnaire somatic, anxiety, and

depressive symptom scales: a systematic review. Gen Hosp Psychiatry. 2010; 32(4):345–59. doi: 10.

1016/j.genhosppsych.2010.03.006 PMID: 20633738

21. Kroenke K, Spitzer RL, Williams JB. The PHQ-9. J Gen Intern Med. 2001; 16(9):606–13. PMID:

11556941

22. (CDC). CfDCaP. National Center for Health Statistics (NCHS): NHANES Laboratory Procedure Man-
ual. 2009–2010.

23. Council NR, Council NR. Food insecurity and hunger in the United States: An assessment of the mea-
sure. National Academies Press, Washington, DC.; 2006.

24. Organization WH, Organization WH. Global database on body mass index. 2006.

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 20 / 23

25. Mezuk B, Johnson-Lawrence V, Lee H, Rafferty JA, Abdou CM, Uzogara EE, et al. Is ignorance bliss?

Depression, antidepressants, and the diagnosis of prediabetes and type 2 diabetes. Health Psychol.

2013; 32(3):254. doi: 10.1037/a0029014 PMID: 23437855

26. World Health Organisation (WHO). Definition, diagnosis and classification of diabetes mellitus and its

complications: Part; 1999.

27. American Diabetes Association. The American Diabetes Association (ADA) has been actively involved

in the development and dissemination of diabetes care standards, guidelines, and related documents

for many years. Introduction. Diabetes Care. 2009; 32:S1. doi: 10.2337/dc09-S001 PMID: 19118285

28. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care.

2010; 33(Supplement 1):S62–S9.

29. Alberti KGMM, Zimmet PZ,. Definition, diagnosis and classification of diabetes mellitus and its compli-
cations. Part 1: diagnosis and classification of diabetes mellitus. Provisional report of a WHO consulta-
tion. Diabet Med. 1998; 15(7):539–53. PMID: 9686693

30. Nathan DM, Kuenen J, Borg R, Zheng H, Schoenfeld D, Heine RJ. Translating the A1C assay into esti-
mated average glucose values. Diabetes Care. 2008; 31(8):1473–8.

31. Patel P, Macerollo A. Diabetes mellitus: diagnosis and screening. Diabetes. 2010; 100:13.

32. Committee IE. International Expert Committee report on the role of the A1C assay in the diagnosis of

diabetes. Diabetes Care. 2009; 32(7):1327–34. doi: 10.2337/dc09-9033 PMID: 19502545

33. Gaziano TA, Young CR, Fitzmaurice G, Atwood S, Gaziano JM. Laboratory-based versus non-labora-
tory-based method for assessment of cardiovascular disease risk: the NHANES I Follow-up Study

cohort. The Lancet. 2008; 371(9616):923–31.

34. Pandya A, Weinstein MC, Gaziano TA. A comparative assessment of non-laboratory-based versus

commonly used laboratory-based cardiovascular disease risk scores in the NHANES III population.

PLoS One. 2011; 6(5):e20416. doi: 10.1371/journal.pone.0020416 PMID: 21655241

35. Sterne JA, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, et al. Multiple imputation for missing

data in epidemiological and clinical research: potential and pitfalls. BMJ: British Medical Journal.

2009;338.

36. Wood AM, White IR, Thompson SG. Are missing outcome data adequately handled? A review of pub-
lished randomized controlled trials in major medical journals. Clinical trials. 2004; 1(4):368–76. PMID:

16279275

37. Anderson TW. Maximum likelihood estimates for a multivariate normal distribution when some observa-
tions are missing. Journal of the american Statistical Association. 1957; 52(278):200–3.

38. Rubin DB. Inference and missing data. Biometrika. 1976; 63(3):581–92.

39. Tabachnick BG, Fidell L. Using Multivariate Statistics: International Edition: Pearson; 2012.

40. White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues and guidance for

practice. Stat Med. 2011; 30(4):377–99. doi: 10.1002/sim.4067 PMID: 21225900

41. Van Buuren S. Multiple imputation of discrete and continuous data by fully conditional specification.

Stat Methods Med Res. 2007; 16(3):219–42. PMID: 17621469

42. Raghunathan TE, Lepkowski JM, Van Hoewyk J, Solenberger P. A multivariate technique for multiply

imputing missing values using a sequence of regression models. Survey methodology. 2001; 27

(1):85–96.

43. Kenward MG, Carpenter J. Multiple imputation: current perspectives. Stat Methods Med Res. 2007; 16

(3):199–218. PMID: 17621468

44. StataCorp L. Stata multiple-imputation reference manual. 2015.

45. Van Ness PH, Murphy TE, Araujo KL, Pisani MA, Allore HG. The use of missingness screens in clinical

epidemiologic research has implications for regression modeling. J Clin Epidemiol. 2007; 60(12):1239–

45. PMID: 17998078

46. Little RJ. A test of missing completely at random for multivariate data with missing values. Journal of

the American Statistical Association. 1988; 83(404):1198–202.

47. Li C. Little's test of missing completely at random. Stata Journal. 2013; 13(4):795–809.

48. Royston P, White IR. Multiple imputation by chained equations (MICE): implementation in Stata. Jour-
nal of Statistical Software. 2011; 45(4):1–20.

49. Essau CA, Lewinsohn PM, Seeley JR, Sasagawa S. Gender differences in the developmental course

of depression. J Affect Disord. 2010; 127(1):185–90.

50. Friedman J, Hastie T, Tibshirani R. Additive logistic regression: a statistical view of boosting (with dis-
cussion and a rejoinder by the authors). The annals of statistics. 2000; 28(2):337–407.

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 21 / 23

51. Opitz D, Maclin R. Popular ensemble methods: An empirical study. Journal of Artificial Intelligence

Research. 1999:169–98.

52. Bauer E, Kohavi R. An empirical comparison of voting classification algorithms: Bagging, boosting, and

variants. Machine learning. 1999; 36(1–2):105–39.

53. Elith J, Leathwick JR, Hastie T. A working guide to boosted regression trees. Journal of Animal Ecol-
ogy. 2008; 77(4):802–13. doi: 10.1111/j.1365-2656.2008.01390.x PMID: 18397250

54. Hastie T, Tibshirani R, Friedman J, Franklin J. The elements of statistical learning: data mining, infer-
ence and prediction. The Mathematical Intelligencer. 2005; 27(2):83–5.

55. Buston PM, Elith J. Determinants of reproductive success in dominant pairs of clownfish: a boosted

regression tree analysis. Journal of Animal Ecology. 2011; 80(3):528–38. doi: 10.1111/j.1365-2656.

2011.01803.x PMID: 21284624

56. Van Voorhees BW, Paunesku D, Gollan J, Kuwabara S, Reinecke M, Basu A. Predicting future risk of

depressive episode in adolescents: the Chicago Adolescent Depression Risk Assessment (CADRA).

The Annals of Family Medicine. 2008; 6(6):503–11. doi: 10.1370/afm.887 PMID: 19001302

57. Freund Y, Schapire RE. A decision-theoretic generalization of on-line learning and an application to

boosting. Journal of computer and system sciences. 1997; 55(1):119–39.

58. De'Ath G. Boosted trees for ecological modeling and prediction. Ecology. 2007; 88(1):243–51. PMID:

17489472

59. Friedman J, Hastie T, Tibshirani R. The elements of statistical learning: Springer series in statistics

Springer, Berlin; 2001.

60. Breiman L. Bagging predictors. Machine learning. 1996; 24(2):123–40.

61. Schonlau M. Boosted regression (boosting): An introductory tutorial and a Stata plugin. Stata Journal.

2005; 5(3):330.

62. Farrar DE, Glauber RR. Multicollinearity in regression analysis: the problem revisited. The Review of

Economic and Statistics. 1967:92–107.

63. Tibshirani R. Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society

Series B (Methodological). 1996:267–88.

64. Tibshirani R. Regression shrinkage and selection via the lasso: a retrospective. Journal of the Royal

Statistical Society: Series B (Statistical Methodology). 2011; 73(3):273–82.

65. Friedman J, Hastie T, Tibshirani R. glmnet: Lasso and elastic-net regularized generalized linear mod-
els. R package version. 2009;1.

66. Schaefer RL. Alternative estimators in logistic regression when the data are collinear. Journal of Statis-
tical Computation and Simulation. 1986; 25(1–2):75–91.

67. Perlstein TS, Weuve J, Pfeffer MA, Beckman JA. Red blood cell distribution width and mortality risk in a

community-based prospective cohort. Arch Intern Med. 2009; 169(6):588–94. doi: 10.1001/

archinternmed.2009.55 PMID: 19307522

68. Kimmenade RR, Mohammed AA, Uthamalingam S, Meer P, Felker GM, Januzzi JL. Red blood cell dis-
tribution width and 1-year mortality in acute heart failure. Eur J Heart Fail. 2010; 12(2):129–36. doi: 10.

1093/eurjhf/hfp179 PMID: 20026456

69. Öztürk ZA, Ünal A, Yiğiter R, Yesil Y, Kuyumcu ME, Neyal M, et al. Is increased red cell distribution

width (RDW) indicating the inflammation in Alzheimer's disease (AD)? Arch Gerontol Geriatr. 2013; 56

(1):50–4. doi: 10.1016/j.archger.2012.10.002 PMID: 23103090

70. Malandrino N, Wu W, Taveira T, Whitlatch H, Smith R. Association between red blood cell distribution

width and macrovascular and microvascular complications in diabetes. Diabetologia. 2012; 55(1):226–

35. doi: 10.1007/s00125-011-2331-1 PMID: 22002006

71. Spell DW, Jones DV, Harper WF, Bessman JD. The value of a complete blood count in predicting can-
cer of the colon. Cancer Detect Prev. 2004; 28(1):37–42. PMID: 15041076

72. Yeşil A, Şenateş E, Bayoğlu İV, Erdem ED, Demirtunç R, Övünç AOK. Red cell distribution width: a

novel marker of activity in inflammatory bowel disease. Gut and liver. 2011; 5(4):460. doi: 10.5009/gnl.

2011.5.4.460 PMID: 22195244

73. Stewart JC, Rand KL, Muldoon MF, Kamarck TW. A prospective evaluation of the directionality of the

depression–inflammation relationship. Brain Behav Immun. 2009; 23(7):936–44. doi: 10.1016/j.bbi.

2009.04.011 PMID: 19416750

74. Miller AH, Maletic V, Raison CL. Inflammation and its discontents: the role of cytokines in the patho-
physiology of major depression. Biol Psychiatry. 2009; 65(9):732–41. doi: 10.1016/j.biopsych.2008.11.

029 PMID: 19150053

75. Mezuk B, Eaton WW, Albrecht S, Golden SH. Depression and type 2 diabetes over the lifespan a meta-
analysis. Diabetes Care. 2008; 31(12):2383–90. doi: 10.2337/dc08-0985 PMID: 19033418

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 22 / 23

76. Van der Kooy K, van Hout H, Marwijk H, Marten H, Stehouwer C, Beekman A. Depression and the risk

for cardiovascular diseases: systematic review and meta analysis. Int J Geriatr Psychiatry. 2007; 22

(7):613–26. PMID: 17236251

77. Ani C, Ovbiagele B. Elevated red blood cell distribution width predicts mortality in persons with known

stroke. J Neurol Sci. 2009; 277(1):103–8.

78. Anderson RJ, Freedland KE, Clouse RE, Lustman PJ. The prevalence of comorbid depression in adults

with diabetes a meta-analysis. Diabetes Care. 2001; 24(6):1069–78. PMID: 11375373

79. Egede LE, Ellis C. Diabetes and depression: global perspectives. Diabetes Res Clin Pract. 2010; 87

(3):302–12. doi: 10.1016/j.diabres.2010.01.024 PMID: 20181405

80. Winokur A, Maislin G, Phillips JL, Amsterdam JD. Insulin resistance after oral glucose tolerance testing

in patients with major depression. Am J Psychiatry. 1988; 145(3):325–30. PMID: 2894176

81. Stocker R, Yamamoto Y, McDonagh AF, Glazer AN, Ames BN. Bilirubin is an antioxidant of possible

physiological importance. Science. 1987; 235(4792):1043–6. PMID: 3029864

82. Sedlak TW, Snyder SH. Bilirubin benefits: cellular protection by a biliverdin reductase antioxidant cycle.

Pediatrics. 2004; 113(6):1776–82. PMID: 15173506

83. Maes M, Galecki P, Chang YS, Berk M. A review on the oxidative and nitrosative stress (O&NS) path-
ways in major depression and their possible contribution to the (neuro) degenerative processes in that

illness. Prog Neuropsychopharmacol Biol Psychiatry. 2011; 35(3):676–92. doi: 10.1016/j.pnpbp.2010.

05.004 PMID: 20471444

84. Schwertner HA. Association of smoking and low serum bilirubin antioxidant concentrations. Atheroscle-
rosis. 1998; 136(2):383–7. PMID: 9543110

85. Van Hoydonck PG, Temme EH, Schouten EG. Serum bilirubin concentration in a Belgian population:

the association with smoking status and type of cigarettes. International journal of epidemiology. 2001;

30(6):1465–72. PMID: 11821365

86. Longadge R, Dongre S. Class Imbalance Problem in Data Mining Review. arXiv preprint

arXiv:13051707. 2013.

Data Mine & Machine Learn Biomarkers of Depression

PLOS ONE | DOI:10.1371/journal.pone.0148195 February 5, 2016 23 / 23

Reproduced with permission of the copyright owner. Further reproduction prohibited without

permission.


Tidak ada komentar:

Posting Komentar