Personalized Profiling of Lipoprotein and Lipid Metabolism Based
on 1018 Measures from Combined Quantitative NMR and LC-MS/MS
Platforms
Siyu Zhao, Corey Giles, Kevin Huynh, Johannes Kettunen, Marjo-Riitta Järvelin, Mika Kähönen,
Jorma Viikari, Terho Lehtimäki, Olli T. Raitakari, Peter J. Meikle, Ville-Petteri Mäkinen,
and Mika Ala-Korpela*
Cite This: Anal. Chem. 2024, 96, 20362−20370 Read Online
ACCESS Metrics & More Article Recommendations *sı Supporting Information
ABSTRACT: Applications of advanced omics methodologies are
increasingly popular in biomedicine. However, large-scale studies
aiming at clinical translation are typically siloed to single technologies.
Here, we present the first comprehensive large-scale population data
combining 209 lipoprotein measures from a quantitative NMR
spectroscopy platform and 809 lipid classes and species from a
quantitative LC-MS/MS platform. These data with 1018 molecular
measures were analyzed in two population cohorts totaling 7830
participants. The association and cluster analyses revealed excellent
coherence between the methodologically independent data domains
and confirmed their quantitative compatibility and suitability for large-
scale studies. The analyses elucidated the detailed molecular
characteristics of the heterogeneous circulatory macromolecular lipid transport system and the underlying structural and
compositional relationships. Unsupervised neural network analysis�the so-called self-organizing maps (SOMs)�revealed that
these deep molecular and metabolic data are inherently related to key physiological and clinical population characteristics. The data-
driven population subgroups uncovered marked differences in the population distribution of multiple cardiometabolic risk factors.
These include, e.g., multiple lipoprotein lipids, apolipoprotein B, ceramides, and oxidized lipids. All 79 structurally unique
triglyceride species showed similar associations over the entire lipoprotein cascade and indicated systematically increased risk for
carotid intima media thickening and other atherosclerosis risk factors, including obesity and inflammation. The metabolic attributes
for 27 individual cholesteryl ester species, which formed six distinct clusters, were more intricate with associations both with
higher�e.g., CE(16:1)�and lower�e.g., CE(20:4)�cardiometabolic risk. The molecular details provided by these combined data
are unprecedented for molecular epidemiology and demonstrate a new potential avenue for population studies.
Lipoprotein lipids have had a central role in clinical riskassessment for cardiometabolic diseases since the 1950s.1
The realization that 1H NMR spectroscopy would be
inherently suited for lipoprotein analytics took place in the
early 1990s.2 In addition to the fundamental advantages of 1H
NMR spectroscopy in producing quantitative data, the
spherical monolayered structure of lipoprotein particles
provides a compelling attribute that the chemical shifts of
lipid molecules transported by these particles are dependent on
the particle diameter. This phenomenon arises from the
orientational order of the surface lipids and the resulting
anisotropy of the magnetic susceptibility, and leads to
distinctive particle size-dependent frequency shifts. This
facilitates comprehensive lipoprotein analytics that is pertinent
to the metabolic constituents of the lipoprotein cascade.
Practical applications and statistical considerations have
indicated that 14 lipoprotein subclasses and their main lipid
constituents can be robustly quantified.3,4 Multiple commercial
methodologies have also been developed.5−7 Here we applied a
platform that has been widely applied in large-scale
epidemiology and genetics over the last 15 years with some
500 publications and 2 M samples analyzed, including all the
0.5 M serum samples in the UK Biobank.3,5,8 This particular
platform has also played a key role in yielding open access
genetic data, with the most recent genome-wide association
study meta-analysis being carried out in 136,000 individuals.9
Mass spectrometry based lipidomics methodologies have
been developed independent of the above and are technolog-
Received: June 24, 2024
Revised: November 18, 2024
Accepted: November 26, 2024
Published: December 16, 2024
Articlepubs.acs.org/ac
© 2024 The Authors. Published by
American Chemical Society
20362
https://doi.org/10.1021/acs.analchem.4c03229
Anal. Chem. 2024, 96, 20362−20370
This article is licensed under CC-BY 4.0
D
ow
nl
oa
de
d 
vi
a 
U
N
IV
 O
F 
TU
RK
U
 o
n 
Ja
nu
ar
y 
17
, 2
02
5 
at
 1
1:
10
:0
4 
(U
TC
).
Se
e 
ht
tp
s:/
/p
ub
s.a
cs
.o
rg
/sh
ar
in
gg
ui
de
lin
es
 fo
r o
pt
io
ns
 o
n 
ho
w
 to
 le
gi
tim
at
el
y 
sh
ar
e 
pu
bl
ish
ed
 a
rti
cl
es
.
ically more challenging than the applications of 1H NMR
spectroscopy.10−12 Nevertheless, epidemiological applications
have become feasible over the recent years and clinical
translation has been suggested, for example, with respect to
assessing cardiometabolic risk13,14 and cognitive impairment.15
Initial genetic studies have also been published.16,17
The NMR spectroscopy and MS lipidomics approaches to
lipid characterization and quantification are fundamentally
different (Figure 1). In NMR the serum sample is directly
analyzed with the lipoprotein particles intact; this allows
quantification of the comprehensive and specific data on the
lipoprotein subclasses and lipids.18 Unlike NMR, LC-MS/MS
lipidomics calls for lipid extraction of the serum samples as an
integral prerequisite for the analyses.19 The lipidomics sample
preparations thus represent pooled mixtures of all the lipid
molecules from all the circulating lipoprotein particles. All the
information from the specific lipoprotein origin of a certain
lipid in the circulation is lost in the extraction phase. However,
with LC-MS/MS hundreds of unique lipid species can be
analyzed from dozens of different molecular lipid classes.12,20
Lipoprotein particles are the sole transport vehicles for all
lipid molecules in serum,18,21,22 though albumin contains the
bulk of free fatty acids and acylcarnitines, and transports also
various lyso-type of lipid molecules.23,24 The NMR and the
LC-MS/MS data on serum essentially represent the same lipids
as noted by Ghorasaini et al.25 The NMR provides the size-
resolution and metabolic information on lipoprotein particles
(209 measures) and the LC-MS/MS lipidomics opens up a
prolific and detailed molecular view on all the circulating lipids
(809 measures).
These different methodologies and molecular views to study
serum lipids have independently advanced toward large-scale
population studies. Here, we integrate these state-of-the-art
data�over 1000 quantitative molecular lipid measures per
individual�in two large-scale population cohorts of over 7800
participants. These data reveal the fundamental relationships
between the lipoprotein subclass cascade and metabolism
(NMR) and the plethora of specific molecular information
(LC-MS/MS). The molecular and metabolic corollaries are
numerous, and we elaborate on some pertinent findings.
Among the various analyses we also demonstrate the admirable
coherence within the independent experiments and data
domains. It is also uncovered�via data-driven, unsupervised
neural network analysis�that these molecular lipid measures
per se can be utilized for metabolically and clinically
meaningful subgrouping of populations. The results illustrate
that these combined comprehensive data lead to improved
molecular characterization of population-level metabolic health
and cardiometabolic risk.
■ MATERIALS AND METHODS
Population Cohorts. Two cohorts were studied: the
Northern Finland Birth Cohort 1966 (NFBC66) with 5657
participants (median age 46 years, 56% women) and the
Young Finns Study (YFS) with 2036 participants (43 years,
55% women). Details on these cohorts are given in the
Supporting Information, including the clinical characteristics in
Table S1. NFBC66 is one of the biggest epidemiological
cohorts with LC-MS/MS data available and these cohorts
comprise a unique combination of large-scale molecular data
for comprehensive lipoprotein panels and lipidomics measures.
1H NMR Spectroscopy.We applied an NMR platform that
has been widely used in epidemiology and for which the
general methodological issues have been published.3,5,8,9,18 The
separation of lipoprotein subclasses by proton NMR spectros-
copy is based on particle size,2 the platform resolution being 14
subclasses (Figure 2).3,18 Independent verification for the
number of subclasses has been published by Mihaleva and co-
workers with an in-depth handling of statistical grounds.4
Particle (P), phospholipid (PL), free cholesterol (FC),
cholesteryl ester (CE), and triglyceride (TG) concentrations
are quantified for all 14 subclasses, also allowing calculation of
the relative lipid composition for each lipoprotein subclass.
The platform also provides all key clinically used lipid
measures (e.g., total TG, cholesterol, and HDL-C) as well as
apolipoprotein B (apoB) and A-I (apoA-I) concentrations. In
total 209 lipoprotein-related measures were analyzed (98
concentration and 70 compositional measures from 14
lipoprotein subclasses, 19 clinical lipid measures, and 22
measures related to fatty acids, average particle sizes, and
apoA-I and apoB).
Lipoprotein Subclasses. The lipoprotein subclass in-
formation can be partitioned into three key categories of
variables: (1) circulating particle concentrations, (2) circulat-
ing lipid concentrations, and (3) particle lipid compositions as
the percentage of a certain lipid class of total lipids (mol %).
All the measures are listed in Supporting Information Table S2
together with their median concentrations for both cohorts.
The relative circulating concentrations for TG and cholesterol
in the 14 lipoprotein subclasses are illustrated in Figure 2.
LC-MS/MS Mass Spectrometry. Serum lipids were
extracted as previously described.19 The LC-MS/MS profiling
was performed using the standardized methodology described
recently by Huynh et al.12 and updated with a larger set of
Figure 1. Schematic illustration of the methodological differences and
analytical characteristics between 1H NMR and LC-MS/MS
lipidomics and their molecular views into serum lipids. (A) The
lipoprotein analysis with NMR is performed directly from the intact
serum sample. (B) For the LC-MS/MS analysis all the lipids need to
be extracted to a homogeneous lipid-soluble mixture. This allows LC-
MS/MS to reach detailed molecular quantification, however, as a total
serum concentration without any information on the originating
lipoproteins. (C,D) The NMR analysis gives molecular information
on 14 lipoprotein subclasses with the abundant core and surface lipids
quantified. The percentages depict the compositional variation of
these lipids in different lipoprotein particles.
Analytical Chemistry pubs.acs.org/ac Article
https://doi.org/10.1021/acs.analchem.4c03229
Anal. Chem. 2024, 96, 20362−20370
20363
targets. The mass spectrometric details and parameters are
given in the Supporting Information.
Lipid Classes and Species. The LC-MS/MS platform
resulted in quantitative data for 809 measures: 38 lipid classes
with 767 individual lipid species plus four classes with a single
detectable molecular species, namely free cholesterol, ubiq-
uinone, ceramide-1-phosphate, and the GM1 ganglioside.
Relative concentrations for the 16 most abundant lipid classes
in the circulation are shown as a summary in Figure 2. All the
measures are listed in Supporting Information Tables S3 and
S4 together with their median concentrations for both cohorts.
These data provide valuable reference concentrations for key
lipid measures at population level (individual variation is
depicted in Supporting Information Figure S1). Notably only
eight lipid classes constitute over 97% of circulating lipids. In
general, within each lipid class, a few molecular species are
clearly more abundant than the rest. The most abundant lipid
classes are also rich in molecular variety, for example, 79
individual molecular species are identified in the TG class, 64
for phosphatidylcholines, and 44 for sphingomyelins. This
detailed panel of individual lipid species provides a novel and
compelling opportunity to study the molecular intricacies of
lipoprotein metabolism and extends the molecular epidemiol-
ogy approach to a mostly unknown area.
Statistical Analyses. Partial Spearman’s correlations were
calculated between all the lipoprotein and lipidomics measures.
Both cohorts were analyzed separately and then combined via
inverse variance weighted meta-analysis. All correlations were
adjusted for sex in both cohorts and age in YFS. Hierarchical
clustering was applied to the lipidomics data domain to
facilitate visual viewing of the results from the correlation
analyses. Two-hundred principal components explained over
95% of the variation in the 1018 lipoprotein and lipid measures
in both cohorts. Therefore, we set the 5% Bonferroni-adjusted
type 1 error threshold at p < 0.05/200 = 0.00025. Extreme
values for all measures were truncated in all analyses to third
quartile +8 × interquartile range. Extreme values were rare and
had negligible effects on the results. All analyses were done
with the R software (version 4.2.1).
Self-Organizing Maps. The Numero R software pack-
age26 was used to derive the data-driven metabolic subgroups.
The NMR and LC-MS/MS data were combined to generate
1018 quantitative molecular inputs for the 7830 participants.
The overall process and principles of SOM analysis are
depicted in Figure 3. This framework has been used in
numerous metabolic studies27 and here we followed the de
facto standard procedure as instructed in the software
documentation (see also Supporting Information).26
■ RESULTS AND DISCUSSION
Characteristics of Integrated Data. The NMR and LC-
MS/MS data domains with key measures are depicted in
Figures 1 and 2. The two essential background issues, namely
the serum sample and the related measurements as well as the
biological system and lipoprotein and lipid metabolism are
briefly introduced above, and more details are given in the
Supporting Information.
NMR is unique in the ability to comprehensively quantify a
multitude of lipoprotein subclasses.2,3,5,8,9,18 Figure 1 illustrates
the ten apoB-containing and four HDL particles that can be
analyzed together with their main lipid constituents, namely
PL, FC, CE, total cholesterol (TC), and TG. The VLDL
particles are large and buoyant with TG-enriched cores (up to
Figure 2. Illustration of the key data from the 1H NMR and LC-MS/
MS analyses. (A) NMR quantifies the most abundant lipid classes in
14 size-specific lipoprotein subclasses and results in a comprehensive
view on systemic lipoprotein metabolism. For example, the main
transport of TG takes place in VLDL particles (over 60% of
circulating amount) and that of cholesterol in IDL and LDL particles
(over 50%). (B−D) LC-MS/MS quantifies also low abundance lipid
classes and related molecular species. (B) Relative concentrations for
the 16 most abundant lipid classes in the circulation. The five most
abundant species for the most abundant core (C) and surface lipids
(D) in lipoprotein particles. Figure 3. Schematic illustration of the self-organizing map (SOM)
framework. (A) First, the SOM algorithm enables the representation
of the multidimensional data set as a two-dimensional map in which
proximity between participants (black dots) indicates similarity in the
metabolic profile. (B) Second, the map is colored by each measure to
compare the regions, i.e., how the average values of these measures for
the participants compare between the map districts. (C) Lastly, the
map is divided into (population) subgroups of participants; here this
was based on the lipoprotein and lipidomics measures used as inputs.
Clinical and biochemical data, not used in the training of the SOM,
can then be visualized similarly to the metabolic inputs.
Analytical Chemistry pubs.acs.org/ac Article
https://doi.org/10.1021/acs.analchem.4c03229
Anal. Chem. 2024, 96, 20362−20370
20364
70% of all particle lipids). The IDL and LDL particles are
smaller, at around 20−30 nm in diameter with CE-enriched
cores (up to 70% of all particle lipids). Consistent with these
compositional features, VLDL particles are the main carriers of
TG (some 60% of circulating TG) and IDL and LDL particles
the cholesterol (some 50% of circulating TC) in the
bloodstream. All the particles have a functional monolayer
that allows the particles to be soluble in water and controls the
particle−particle and particle−cell interactions.18,21,22 The
main molecular constituents in the surface monolayer are
apolipoproteins, PL, and FC. HDL particles represent over
90% of circulating particles but apoB particles carry 65% of
circulating lipids.
LC-MS/MS is unique in the ability to comprehensively
identify and quantify hundreds of individual lipid species in
dozens of different lipid classes.10−12,19 In Figure 2, the relative
percentages of the 16 most abundant lipid classes in serum are
depicted (data for all the lipid classes are given in Supporting
Information Table S3 and a literature comparison for the most
abundant lipid classes in Table S5). CE, TG, phosphatidylcho-
lines (PC), and FC are the most abundant lipid classes
amounting to almost 90% of all circulating lipids. Sphingo-
myelins (SM), lyso-PC, and diacylglycerols (DG) together
contribute around 5%. The rest of the lipid classes contribute
less than 1% each of all the circulating lipid molecules, except
free fatty acids (FFA) that account for some 3% and are
transported by albumin. It is also noticeable that within each
lipid class, there is a tendency that the three to five most
abundant species account for over 50% of the concentration of
the entire class. For example, CE(18:2) alone contributes over
50% and CE(18:1) almost 20% of all CE. TG(52:2),
TG(52:3), and TG(50:2) contribute around 40% of all TG
(Figure 2).
Associations between the Lipoprotein Subclasses
and Lipid Classes. Partial Spearman’s correlations between
the NMR-based lipoprotein and the LC-MS/MS-based lipid
class measures are illustrated in Figure 4. The NMR axis
represents lipoprotein subclass particle (denoted by P) and
main lipid concentrations (total lipids denoted by L). The LC-
MS/MS axis shows concentrations for 40 specific molecular
lipid classes. The clustered heatmap is organized for the LC-
MS/MS data, while the NMR-based lipoprotein data are
presented in the order of decreasing particle size from the
largest VLDL to the smallest HDL particles, since these data
blocks intrinsically correspond to the two dominant metabolic
pathways.18,25
The most pronounced positive correlation block is noted
between the five largest VLDL subclasses and the LC-MS/MS
lipidomics cluster no. 5 that consists of di-, tri-, and
phosphatidylglycerols. This is anticipated due to the key role
of VLDL particles in the transport of these molecular
components in the bloodstream. The related strong negative
correlation block between cluster no. 5 and the XL-, L-, and M-
HDL subclasses arises from the strong systemic negative
metabolic association between triglycerides and HDL-C. This
is an important notion in the sense that, even though we have
specific molecular measures in both data domains, the
circulating lipoprotein particles are metabolically constrained
for both absolute and relative concentrations.18 The
population-level correlation between triglycerides and HDL-
C is typically up to −0.7:28 for the NFBC66 and YFS it is
−0.64 and −0.56, respectively.
Figure 4. Associations between the concentrations of 40 lipid classes (LC-MS/MS) and the particle and lipid concentrations of 14 lipoprotein
subclasses (NMR) as indicated by partial Spearman’s rank correlations. The data are from two independent large population studies, NFBC66
(5657 participants) and YFS (2173 participants). All correlations were adjusted for sex in both cohorts and age in YFS. Both cohorts were analyzed
separately and then meta-analyzed. The LC-MS/MS data are organized to clusters and the NMR-based lipoprotein data are presented in the order
of decreasing particle size, since these data blocks intrinsically correspond the dominant metabolic pathways. The general association characteristics
for the lipid classes are summarized via 14 metabolic clusters. The percentages shown depict the contribution of each cluster to the circulating total
lipid concentration. P-value <0.00025 is marked with an asterisk to indicate a multiple testing corrected association. Abbreviations are in the
Supporting Information.
Analytical Chemistry pubs.acs.org/ac Article
https://doi.org/10.1021/acs.analchem.4c03229
Anal. Chem. 2024, 96, 20362−20370
20365
Another demonstration of the excellent coherence within
the independent data domains is the strong positive correlation
block between lipidomics cluster no. 6, the main components
of which are FC and SM, and the XS-VLDL, IDL, and LDL
particles, which transport some 60% of circulating cholesterol.
The associations of CE (lipidomics cluster no. 11) are similar
but show weak negative associations for the XXL- to S-VLDL
subclasses and positive associations to all HDL subclasses (that
transport some 30% of circulating cholesterol).
The molecular information in Figure 4 is substantial and
shows many attributes that have not been elucidated before.
Since ceramides have been suggested as independent
biomarkers for cardiovascular disease risk,29 it is of interest
to note that the lipidomics cluster no. 2, consisting of
circulating ceramides (Cer) and dihydroceramides (dhCer), is
strongly positively associated with the entire metabolic cascade
of apoB-containing lipoprotein particles. Several components
of this pathway, including TG, LDL-C, and apoB, are known to
be causal for the development of coronary heart disease.1,30,31
It is also notable that this ceramide cluster constitutes only
0.1% of the total circulating lipid concentration. Nonetheless,
the low circulating concentration per se does not prevent
ceramides (or any low abundance lipid class or species) to be
relevant or causal for the atherosclerotic process, particularly
since it has been demonstrated that the interindividual
differences in LDL particle lipidome can affect the suscepti-
bility of these particles for enzymatic modifications and
influence the development and progression of atheroscle-
rosis.32 Based on abundant genetic epidemiology evidence, it is
however more probable that the key mediating factor is
intrinsically the apoB.31 The minor compositional differences
in the LDL particles may not matter much over decades of
apoB overload.33
Notwithstanding, complementary studies on the (potentially
causal) role of ceramides are needed to put these findings on
the population associations of ceramides and cardiometabolic
outcomes under scrutiny together with comprehensive lip-
oprotein data. This notion is not relevant only for ceramides,
but to all lipidomics measures. Coming to robust conclusions
on these individual constituents of lipoprotein particles is
demanding as recently elaborated by Borges et al.34 in the case
of circulating polyunsaturated fatty acids and calling attention
to the potential bias in the analyses of individual lipid measures
due to the lipoprotein-related mediation.
The molecular characteristics of the various subclass
particles are heterogeneous with many lipid classes and species
being distributed unevenly (see also Figure 1C).18,23 Within
these fundamental impediments, the overall view from Figure 4
makes perfect sense between the NMR-based lipoprotein
profiling and the LC-MS/MS-based comprehensive lipid
analyses. This unique outlook would also serve well as a
detailed guide for the potential key lipoprotein mediators for
various lipidomics associations in epidemiology and biomedical
studies when comprehensive lipoprotein profiles are unavail-
able (Spearman’s rank correlations between all the NMR and
LC-MS/MS measures are given in a numerical form in
Supporting Information Table S6).
The associations of the lipidomics cluster no. 9 with the
lipoprotein cascade are very weak. The main component of this
cluster is FFA, that are known to be almost entirely transported
by albumin. It is therefore likely that, at least up to a noticeable
portion, the other lipids, namely lysoalkenylphosphatidyletha-
nolamines (plasmalogens; LPE(P)), acylcarnitines (AC), and
phosphatidylinositol monophosphate (PIP1), are transported
by albumin, or other proteins than lipoproteins, in the
bloodstream. These molecules exemplify the unique benefits
of the lipidomics analyses in providing specific biological
information beyond the lipoprotein cascade.
Associations between the Lipoprotein Subclasses
and Lipid Species. The LC-MS/MS lipidomics data for the
lipid classes (Figures 2 and 4), represent concentrations of all
the individual lipid species summed up for a total
concentration for each lipid class. Therefore, the most
abundant lipid species in each lipid class are those that
contribute nearly all to these total concentrations and can
potentially overwhelm the associations of less abundant
molecular species. Triglycerides are a representative example
of a lipid class within which all the analyzed individual
molecular species appear to associate in a very similar manner
with the entire lipoprotein cascade (Supporting Information
Figure S2). Thereby, the individual TG species are likely to
behave similarly to the total TG class associations. However, as
Figure 5. Associations between the concentrations of 27 individual CE species (LC-MS/MS) and the particle and lipid concentrations of 14
lipoprotein subclasses (NMR) as indicated by partial Spearman’s rank correlations. The general association characteristics for the CE species are
summarized via six metabolic clusters. The percentages shown depict the contribution of each cluster to the circulating total CE concentration.
Otherwise, the data and analyses are as detailed in the caption for Figure 4.
Analytical Chemistry pubs.acs.org/ac Article
https://doi.org/10.1021/acs.analchem.4c03229
Anal. Chem. 2024, 96, 20362−20370
20366
depicted in Figure 5, the metabolic attributes for the analyzed
27 cholesteryl ester species are more intricate with six distinct
molecular clusters that substantially differ in their associations
with the lipoprotein cascade. CE clusters no. 1 (contributing
26.4% of total CE concentration and including CE(18:1), the
second most abundant species), nos. 2, 3, and 4 (solely
CE(16:1)) have positive associations for the XXL- to S-VLDL
subclasses. CE clusters no. 5 (with CE(18:2), the most
abundant CE species) and particularly 6 (with CE(20:4), the
third most abundant species) manifest negative corresponding
associations as also the total CE lipid class in Figure 4. This is
logical since CE clusters no. 5 and 6 contain almost 70% of the
total circulating CE concentration.
Thus, with any lipid class within which the lipid species
differ in their associations with the lipoproteins, it should be
noted that the class association reflects only the most abundant
species. The associations of the lipid species also vary for
different lipoprotein subclasses: e.g., all individual CE species
associate positively with all the LDL subclasses, but there are
strongly positively as well as strongly negatively associated CE
species with respect to various VLDL subclasses.
In general, there are multiple lipid classes that have
heterogeneous lipoprotein associations with their constituent
lipid species, e.g., lysophosphatidylcholines and GM3 ganglio-
sides. Like with the lipid classes (Figure 4), these novel
association maps between the hundreds of individual lipid
species and the lipoprotein subclasses can assist in under-
standing the intertwined metabolic issues and help in
interpreting the specific lipidomics data (Supporting Informa-
tion Figures S2−S11).
Self-Organizing Maps and Population Stratification.
We demonstrated in the UK Biobank, with clinical
biochemistry data from 329 908 participants, that cross-
sectional metabolic subgroups relate to future disease burden
and multimorbidity.27 Here, we have extensive molecular data
(over 1000 measures) on lipoprotein and lipid metabolism for
two large populations (over 7800 participants) and we are
interested in understanding how these lipid measures−
intrinsically−relate to the population health characteristics
and disease risk, and−concomitantly−how the lipoprotein
metabolism and the related attributes (from NMR) associate
with the molecular details of multiple lipid classes and species
(from LC-MS/MS). With the SOM analysis both objectives
were achieved.
Figure 6 depicts representative SOM biomarker planes to
visually tie together clinical lipid measures, lipoprotein subclass
Figure 6. SOM-based subgrouping with the combined lipoprotein and lipid data from NMR and LC-MS/MS (209 and 809 quantitative measures,
respectively) in two population cohorts, NFBC66 with 5657 and YFS with 2173 participants. The SOM was applied with all 7830 participants and
all 1018 molecular inputs. In each plot, the same participants reside in the same district. The district colors indicate the regional deviation from the
global mean and the numbers indicate local mean values. The biomarker planes for all the inputs are given in Supporting Information Figure S12.
The histograms are to clarify and emphasize the inherent capability of the comprehensive lipoprotein and lipid data to reveal key aspects of
population health and disease risk.
Analytical Chemistry pubs.acs.org/ac Article
https://doi.org/10.1021/acs.analchem.4c03229
Anal. Chem. 2024, 96, 20362−20370
20367
metabolism, lipidomics, and how these molecular measures
associate with population health and disease aspects for the
resultant five population subgroups. One of the first messages
conveyed by an overall look is the consistently high
triglycerides (biomarker plane A) for subgroups I and II.
This pattern is almost identical with the one for body mass
index (BMI, R) and C-reactive protein (CRP, S), indicating
that, at the population level, obesity, inflammation, and
triglycerides are strongly positively associated. This is a well-
known metabolic characteristic of obesity35 and serves here as
a confirmation to support the SOM analysis. The composi-
tional information on the lipoprotein subclass particles (L, M,
and N) reveals that high circulating triglycerides (A) are
associated with TG-enrichment of all lipoprotein particles,
including HDL, and possibly reflecting higher risk of coronary
heart disease, independent of total cholesterol and triglycer-
ides.36
The SOM analysis connotate that high circulating
concentrations of ceramides (Figure 6H) are strongly
associated with high triglycerides (A), particularly for the
population subgroup II, and with high LDL-C (B), for the
population subgroups II and III. The associations of Cer with
apoB (C) are like with LDL-C. The subgroups I and II are
characterized by high BMI, high CRP, and TG-enriched
lipoprotein particles. In addition, the individuals with these
metabolic profiles are at higher risk for kidney (T) and liver
dysfunction (U) as well as for coronary heart disease (W) than
individuals characterized by the metabolic profiles in
subgroups III, IV, and V. The population subgroup III is the
only one characterized also by high lipoprotein (a) (Lp(a))
(Figure 6V), a causal cardiometabolic risk marker that is highly
genetically controlled and has only minor links to other
lipoprotein measures.37,38
The highest ceramide concentrations at the population level
therefore coincide with the highest apoB (subgroup II) as well
as with the highest Lp(a) concentrations (subgroup III) but
are not clearly elevated within subgroups I with elevated TG.
Thereby, even though highly correlated with the TG-related
lipoprotein metabolism (Figure 4), its associations with
cardiometabolic outcomes are likely to be more related to
cholesterol- than TG-metabolism. Notably, this is a finding
arising solely based on the combined comprehensive lip-
oprotein and lipid data, without any additional information on
cardiometabolic risk or disease outcomes. This does not
confirm an independent role for ceramides in the cardiome-
tabolic risk assessment, but it does give support for further
studies on its potential additive role as a cardiometabolic
biomarker.
As discussed in relation to Figure 5, the individual
cholesteryl ester species form six distinct molecular clusters
that differ in their associations with the lipoprotein cascade.
The integration of lipoprotein and individual CE species data
gives a unique perspective also for the SOM analysis. The
highest concentrations for the total CE class (Figure 6O)
coincide with population subgroups II, III and IV. The highest
concentrations for the individual CE species CE(16:1) (P)
coincide with subgroups I, II and III, but not IV. On the
contrary, CE(20:4) (Q) coincide with subgroups III, IV and V,
but not II.
In one of the first population level lipidomics studies,
CE(16:1) was found to be associated with the development of
cardiovascular disease.39 In a related editorial,40 mechanistic
insights were proposed relating the, e.g., diet originating
saturated fatty acids, to the generation of monounsaturated
CEs and their integration into the VLDL particles in the liver,
eventually resulting in LDL particles enriched in monounsa-
turated CEs, that might be more prone to advance the intimal
lipid accumulation and the progression of atherosclerosis.
While this still remains a hypothesis, the current results
reflect higher cardiometabolic risk for CE(16:1) than for
CE(20:4). Since the LC-MS/MS measures reflect total
circulating concentrations, both the lipoprotein subclass
distribution and particle compositions affect these measures.
It is the subclass distribution that contributes markedly more
for the circulating lipid concentrations,18 so the primary
interpretation on the heterogeneity regarding CE(16:1) and
CE(20:4) is that it is due to differences in lipoprotein subclass
profile (Figure 6E−G) with different lipoprotein subclasses
being inhomogeneous regarding the transport of different CE
species. Apparently CE(20:4) concentrations (Q) are strongly
associated with large HDL particles (G), known to be
associated negatively with cardiometabolic risk, as also
indicated by various biomarker planes in Figure 6. The
participants in both population studies are rather young, so
there is a very limited number of disease outcomes. However,
the carotid intima media thickness (cIMT, W), an indicator of
vasculopathy associated with the development of atheroscler-
otic plaques, does show thickest values for subgroups I and II.
All CE-measures appear high for subgroup III with also
elevated LDL (F) and Lp(a) (V) concentrations. High
CE(16:1) concentrations also mostly coincide with high
oxidized (J and K) and monounsaturated lipids (I), all linked
to increased risk of atherosclerosis. The current results
therefore support the earlier findings related to CE(16:1)
and the role of monounsaturated CEs as molecular risk factors
for atherosclerosis.39,40 However, how much, if any, of this
increased risk is related to the specific molecular species (e.g.,
monounsaturated CE(16:1) and CE(18:1)) or relative lipid
compositions (e.g., enrichment of certain lipoprotein particles
in monounsaturated lipids or triglycerides) and how much is
mediated by the metabolic issues directly in relation to the
lipoprotein subclass concentrations, remains to be determined.
■ CONCLUSIONS
The presented combination of lipoprotein and lipid data from
NMR spectroscopy and LC-MS/MS lipidomics led to an
unprecedented metabolic and molecular detail with over 1000
measures per person in a population setting of over 7800
individuals. The analyses demonstrated notable coherence
between the data from inherently different spectroscopic
techniques. The data-driven analyses revealed multiple
confirmatory and novel molecular results in relation to
systemic lipid metabolism and illustrated the power of all-
inclusive lipid data in understanding cardiometabolic pop-
ulation health, related clinical factors, and disease risk. The
analyses emphasized the augmented value of interpreting
lipidomics data (deep molecular phenotyping) together with
lipoprotein data (deep metabolic phenotyping) and association
heatmaps between all the lipidomics measures and the key
measures characterizing lipoprotein subclass metabolism are
provided. These unique motifs are useful as guidelines for the
key lipoprotein mediators for various lipidomics associations in
biomedical studies when comprehensive lipoprotein profiles
are not available.
Analytical Chemistry pubs.acs.org/ac Article
https://doi.org/10.1021/acs.analchem.4c03229
Anal. Chem. 2024, 96, 20362−20370
20368
■ ASSOCIATED CONTENT
*sı Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.analchem.4c03229.
Detailed information regarding study populations, data,
methods, funding, and supplementary results (PDF)
LC-MS/MS lipidomics measures for the concentrations
of the circulating individual lipid species in NFBC66 and
YFS (XLSX)
Spearman’s rank correlations (adjusted for sex) (XLSX)
■ AUTHOR INFORMATION
Corresponding Author
Mika Ala-Korpela − Systems Epidemiology, Faculty of
Medicine and Research Unit of Population Health, Faculty of
Medicine, University of Oulu, 90014 Oulu, Finland;
Biocenter Oulu, 90014 Oulu, Finland; Monash University,
Melbourne 3004, Australia; NMR Metabolomics Laboratory,
School of Pharmacy, University of Eastern Finland, 70210
Kuopio, Finland; orcid.org/0000-0001-5905-1206;
Email: mika.ala-korpela@oulu.fi
Authors
Siyu Zhao − Systems Epidemiology, Faculty of Medicine and
Research Unit of Population Health, Faculty of Medicine,
University of Oulu, 90014 Oulu, Finland; Biocenter Oulu,
90014 Oulu, Finland
Corey Giles − Baker Heart and Diabetes Institute, Melbourne
3004, Australia; Baker Department of Cardiometabolic
Health, University of Melbourne, Melbourne 3004,
Australia; orcid.org/0000-0002-6050-1259
Kevin Huynh − Baker Heart and Diabetes Institute,
Melbourne 3004, Australia; Baker Department of
Cardiometabolic Health, University of Melbourne, Melbourne
3004, Australia; orcid.org/0000-0001-6170-2207
Johannes Kettunen − Systems Epidemiology, Faculty of
Medicine and Research Unit of Population Health, Faculty of
Medicine, University of Oulu, 90014 Oulu, Finland;
Biocenter Oulu, 90014 Oulu, Finland; Department of Public
Health and Welfare, Finnish Institute for Health and
Welfare, 00271 Helsinki, Finland
Marjo-Riitta Järvelin − Research Unit of Population Health,
Faculty of Medicine, University of Oulu, 90014 Oulu,
Finland; Department of Epidemiology and Biostatistics, MRC
Centre for Environment and Health, School of Public Health,
Imperial College London, London W12 0BZ, U.K.;
Department of Life Sciences, College of Health and Life
Sciences, Brunel University London, London UB8 3PH, U.K.
Mika Kähönen − Department of Clinical Physiology, Tampere
University Hospital, and Finnish Cardiovascular Research
Center Tampere, Faculty of Medicine and Health Technology,
Tampere University, 33270 Tampere, Finland
Jorma Viikari − Department of Medicine, University of Turku,
20014 Turku, Finland; Division of Medicine, Turku
University Hospital, 20014 Turku, Finland
Terho Lehtimäki − Department of Clinical Chemistry, Fimlab
Laboratories, and Finnish Cardiovascular Research Center
Tampere, Faculty of Medicine and Health Technology,
Tampere University, 33270 Tampere, Finland
Olli T. Raitakari − Research Centre of Applied and Preventive
Cardiovascular Medicine and InFLAMES Research Flagship,
University of Turku, 20014 Turku, Finland; Centre for
Population Health Research, University of Turku and Turku
University Hospital, 20014 Turku, Finland; Department of
Clinical Physiology and Nuclear Medicine, Turku University
Hospital, 20014 Turku, Finland
Peter J. Meikle − Baker Heart and Diabetes Institute,
Melbourne 3004, Australia; Baker Department of
Cardiometabolic Health, University of Melbourne, Melbourne
3004, Australia; Monash University, Melbourne 3004,
Australia
Ville-Petteri Mäkinen − Systems Epidemiology, Faculty of
Medicine and Research Unit of Population Health, Faculty of
Medicine, University of Oulu, 90014 Oulu, Finland;
Biocenter Oulu, 90014 Oulu, Finland
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.analchem.4c03229
Author Contributions
S.Z. performed statistical analyses, prepared figures and tables,
and contributed to the concept. C.G., K.H., and P.J.M.
provided the LC-MS/MS data. J.K., M.-R.J., M.K., J.V., T.L.,
O.T.R., and M.A.-K. provided the clinical and NMR data. V.-
P.M. contributed to the concept and figures and supervised
statistical analyses. M.A.-K. conceived, designed, and led the
study, and wrote the manuscript. All authors interpreted and
discussed the results, contributed to the manuscript, and
approved the final version.
Funding
The funding details are given in the Supporting Information.
Notes
The authors declare no competing financial interest.
The abbreviations are given in the Supporting Information.
■ REFERENCES
(1) Goldstein, J. L.; Brown, M. S. Cell 2015, 161, 161−172.
(2) Ala-Korpela, M. Prog. Nucl. Magn. Reson. Spectrosc. 1995, 27,
475−554.
(3) Soininen, P.; Kangas, A. J.; Würtz, P.; Suna, T.; Ala-Korpela, M.
Circ. Cardiovasc. Genet. 2015, 8, 192−206.
(4) Mihaleva, V. V.; van Schalkwijk, D. B.; de Graaf, A. A.; van
Duynhoven, J.; van Dorsten, F. A.; Vervoort, J.; Smilde, A.;
Westerhuis, J. A.; Jacobs, D. M. Anal. Chem. 2014, 86, 543−550.
(5) Würtz, P.; Kangas, A. J.; Soininen, P.; Lawlor, D. A.; Davey
Smith, G.; Ala-Korpela, M. Am. J. Epidemiol. 2017, 186, 1084−1096.
(6) Garcia, E.; Bennett, D. W.; Connelly, M. A.; Jeyarajah, E. J.;
Warf, F. C.; Shalaurova, I.; Matyus, S. P.; Wolak-Dinsmore, J.;
Oskardmay, D. N.; Young, R. M.; Sampson, M.; Remaley, A. T.;
Otvos, J. D. Lipids Health Dis. 2020, 19, 247.
(7) Debik, J.; Isaksen, S. H.; Stro̷mmen, M.; Spraul, M.; Schäfer, H.;
Bathen, T. F.; Giskeo̷degård, G. F. Anal. Chem. 2022, 94, 17003−
17010.
(8) Julkunen, H.; Cichonśka, A.; Tiainen, M.; Koskela, H.; Nybo, K.;
Mäkelä, V.; Nokso-Koivisto, J.; Kristiansson, K.; Perola, M.; Salomaa,
V.; Jousilahti, P.; Lundqvist, A.; Kangas, A. J.; Soininen, P.; Barrett, J.
C.; Würtz, P. Nat. Commun. 2023, 14, 604.
(9) Karjalainen, M. K.; Karthikeyan, S.; Oliver-Williams, C.; Sliz, E.;
Allara, E.; Fung, W. T.; Surendran, P.; Zhang, W.; Jousilahti, P.;
Kristiansson, K.; Salomaa, V.; Goodwin, M.; Hughes, D. A.; Boehnke,
M.; Fernandes Silva, L.; Yin, X.; Mahajan, A.; Neville, M. J.; van
Zuydam, N. R.; de Mutsert, R.; Li-Gao, R.; Mook-Kanamori, D. O.;
Demirkan, A.; Liu, J.; Noordam, R.; Trompet, S.; Chen, Z.;
Kartsonaki, C.; Li, L.; Lin, K.; Hagenbeek, F. A.; Hottenga, J. J.;
Pool, R.; Ikram, M. A.; van Meurs, J.; Haller, T.; Milaneschi, Y.;
Kähönen, M.; Mishra, P. P.; Joshi, P. K.; Macdonald-Dunlop, E.;
Mangino, M.; Zierer, J.; Acar, I. E.; Hoyng, C. B.; Lechanteur, Y. T.
E.; Franke, L.; Kurilshikov, A.; Zhernakova, A.; Beekman, M.; van den
Analytical Chemistry pubs.acs.org/ac Article
https://doi.org/10.1021/acs.analchem.4c03229
Anal. Chem. 2024, 96, 20362−20370
20369
Akker, E. B.; Kolcic, I.; Polasek, O.; Rudan, I.; Gieger, C.;
Waldenberger, M.; Asselbergs, F. W.; China Kadoorie Biobank
Collaborative Group; Estonian Biobank Research Team; FinnGen;
Hayward, C.; Fu, J.; den Hollander, A. I.; Menni, C.; Spector, T. D.;
Wilson, J. F.; Lehtimäki, T.; Raitakari, O. T.; Penninx, B. W. J. H.;
Esko, T.; Walters, R. G.; Jukema, J. W.; Sattar, N.; Ghanbari, M.;
Willems van Dijk, K.; Karpe, F.; McCarthy, M. I.; Laakso, M.; Järvelin,
M.-R.; Timpson, N. J.; Perola, M.; Kooner, J. S.; Chambers, J. C.; van
Duijn, C.; Slagboom, P. E.; Boomsma, D. I.; Danesh, J.; Ala-Korpela,
M.; Butterworth, A. S.; Kettunen, J. Nature 2024, 628, 130−138.
(10) Weir, J. M.; Wong, G.; Barlow, C. K.; Greeve, M. A.;
Kowalczyk, A.; Almasy, L.; Comuzzie, A. G.; Mahaney, M. C.; Jowett,
J. B. M.; Shaw, J.; Curran, J. E.; Blangero, J.; Meikle, P. J. J. Lipid Res.
2013, 54, 2898−2908.
(11) Mundra, P. A.; Shaw, J. E.; Meikle, P. J. Int. J. Epidemiol. 2016,
45, 1329−1338.
(12) Huynh, K.; Barlow, C. K.; Jayawardana, K. S.; Weir, J. M.;
Mellett, N. A.; Cinel, M.; Magliano, D. J.; Shaw, J. E.; Drew, B. G.;
Meikle, P. J. Cell Chem. Biol. 2019, 26, 71−84.e4.
(13) Alshehry, Z. H.; Mundra, P. A.; Barlow, C. K.; Mellett, N. A.;
Wong, G.; McConville, M. J.; Simes, J.; Tonkin, A. M.; Sullivan, D. R.;
Barnes, E. H.; Nestel, P. J.; Kingwell, B. A.; Marre, M.; Neal, B.;
Poulter, N. R.; Rodgers, A.; Williams, B.; Zoungas, S.; Hillis, G. S.;
Chalmers, J.; Woodward, M.; Meikle, P. J. Circulation 2016, 134,
1637−1650.
(14) Meikle, T. G.; Huynh, K.; Giles, C.; Meikle, P. J. J. Lipid Res.
2021, 62, No. 100127.
(15) Wang, T.; Huynh, K.; Giles, C.; Mellett, N. A.; Duong, T.;
Nguyen, A.; Lim, W. L. F.; Smith, A. A.; Olshansky, G.; Cadby, G.;
Hung, J.; Hui, J.; Beilby, J.; Watts, G. F.; Chatterjee, P.; Martins, I.;
Laws, S. M.; Bush, A. I.; Rowe, C. C.; Villemagne, V. L.; Ames, D.;
Masters, C. L.; Taddei, K.; Doré, V.; Fripp, J.; Arnold, M.;
Kastenmüller, G.; Nho, K.; Saykin, A. J.; Baillie, R.; Han, X.;
Martins, R. N.; Moses, E. K.; Kaddurah-Daouk, R.; Meikle, P. J.
Alzheimers Dement. 2022, 18, 2151−2166.
(16) Cadby, G.; Giles, C.; Melton, P. E.; Huynh, K.; Mellett, N. A.;
Duong, T.; Nguyen, A.; Cinel, M.; Smith, A.; Olshansky, G.; Wang,
T.; Brozynska, M.; Inouye, M.; McCarthy, N. S.; Ariff, A.; Hung, J.;
Hui, J.; Beilby, J.; Dubé, M.-P.; Watts, G. F.; Shah, S.; Wray, N. R.;
Lim, W. L. F.; Chatterjee, P.; Martins, I.; Laws, S. M.; Porter, T.;
Vacher, M.; Bush, A. I.; Rowe, C. C.; Villemagne, V. L.; Ames, D.;
Masters, C. L.; Taddei, K.; Arnold, M.; Kastenmüller, G.; Nho, K.;
Saykin, A. J.; Han, X.; Kaddurah-Daouk, R.; Martins, R. N.; Blangero,
J.; Meikle, P. J.; Moses, E. K. Nat. Commun. 2022, 13, 3124.
(17) Ottensmann, L.; Tabassum, R.; Ruotsalainen, S. E.; Gerl, M. J.;
Klose, C.; Widén, E.; Simons, K.; Ripatti, S.; Pirinen, M. Nat.
Commun. 2023, 14, 6934.
(18) Ala-Korpela, M.; Zhao, S.; Järvelin, M.-R.; Mäkinen, V.-P.;
Ohukainen, P. Int. J. Epidemiol. 2022, 51, 996−1011.
(19) Alshehry, Z. H.; Barlow, C. K.; Weir, J. M.; Zhou, Y.;
McConville, M. J.; Meikle, P. J. Metabolites 2015, 5, 389−403.
(20) Arnold, M.; Nho, K.; Kueider-Paisley, A.; Massaro, T.; Huynh,
K.; Brauner, B.; MahmoudianDehkordi, S.; Louie, G.; Moseley, M. A.;
Thompson, J. W.; John-Williams, L. S.; Tenenbaum, J. D.; Blach, C.;
Chang, R.; Brinton, R. D.; Baillie, R.; Han, X.; Trojanowski, J. Q.;
Shaw, L. M.; Martins, R.; Weiner, M. W.; Trushina, E.; Toledo, J. B.;
Meikle, P. J.; Bennett, D. A.; Krumsiek, J.; Doraiswamy, P. M.; Saykin,
A. J.; Kaddurah-Daouk, R.; Kastenmüller, G. Nat. Commun. 2020, 11,
1−12.
(21) Hevonoja, T.; Pentikäinen, M. O.; Hyvönen, M. T.; Kovanen,
P. T.; Ala-Korpela, M. Biochim. Biophys. Acta 2000, 1488, 189−210.
(22) Borén, J.; Taskinen, M.-R.; Björnson, E.; Packard, C. J. Nat.
Rev. Cardiol. 2022, 19, 577−592.
(23) Wiesner, P.; Leidl, K.; Boettcher, A.; Schmitz, G.; Liebisch, G. J.
Lipid Res. 2009, 50, 574−585.
(24) Dambrova, M.; Makrecka-Kuka, M.; Kuka, J.; Vilskersts, R.;
Nordberg, D.; Attwood, M. M.; Smesny, S.; Sen, Z. D.; Guo, A. C.;
Oler, E.; Tian, S.; Zheng, J.; Wishart, D. S.; Liepinsh, E.; Schiöth, H.
B. Pharmacol. Rev. 2022, 74, 506−551.
(25) Ghorasaini, M.; Tsezou, K. I.; Verhoeven, A.; Mohammed, Y.;
Vlachoyiannopoulos, P.; Mikros, E.; Giera, M. Metabolites 2022, 12,
1030.
(26) Gao, S.; Mutter, S.; Casey, A.; Mäkinen, V.-P. Int. J. Epidemiol.
2019, 48, 369−374.
(27) Mulugeta, A.; Hyppönen, E.; Ala-Korpela, M.; Mäkinen, V.-P.
Sci. Rep. 2022, 12, 8590.
(28) Davey Smith, G.; Phillips, A. N. Int. J. Epidemiol. 2020, 49, 4−
14.
(29) Hilvo, M.; Meikle, P. J.; Pedersen, E. R.; Tell, G. S.; Dhar, I.;
Brenner, H.; Schöttker, B.; Lääperi, M.; Kauhanen, D.; Koistinen, K.
M.; Jylhä, A.; Huynh, K.; Mellett, N. A.; Tonkin, A. M.; Sullivan, D.
R.; Simes, J.; Nestel, P.; Koenig, W.; Rothenbacher, D.; Nygård, O.;
Laaksonen, R. Eur. Heart J. 2020, 41, 371−380.
(30) Schmidt, A. F.; Joshi, R.; Gordillo-Marañón, M.; Drenos, F.;
Charoen, P.; Giambartolomei, C.; Bis, J. C.; Gaunt, T. R.; Hughes, A.
D.; Lawlor, D. A.; Wong, A.; Price, J. F.; Chaturvedi, N.;
Wannamethee, G.; Franceschini, N.; Kivimaki, M.; Hingorani, A.
D.; Finan, C. Commun. Med. 2023, 3, 9.
(31) Ala-Korpela, M. Int. J. Epidemiol. 2019, 48, 1389−1392.
(32) Ruuth, M.; Nguyen, S. D.; Vihervaara, T.; Hilvo, M.; Laajala, T.
D.; Kondadi, P. K.; Gisterå, A.; Lähteenmäki, H.; Kittilä, T.; Huusko,
J.; Uusitupa, M.; Schwab, U.; Savolainen, M. J.; Sinisalo, J.; Lokki, M.-
L.; Nieminen, M. S.; Jula, A.; Perola, M.; Ylä-Herttula, S.; Rudel, L.;
Öörni, A.; Baumann, M.; Baruch, A.; Laaksonen, R.; Ketelhuth, D. F.
J.; Aittokallio, T.; Jauhiainen, M.; Käkelä, R.; Borén, J.; Williams, K. J.;
Kovanen, P. T.; Öörni, K. Eur. Heart J. 2018, 39, 2562−2573.
(33) Toth, P. P.; Sniderman, A. D. JACC Adv. 2024, 3, No. 100756.
(34) Borges, M. C.; Haycock, P. C.; Zheng, J.; Hemani, G.; Holmes,
M. V.; Davey Smith, G.; Hingorani, A. D.; Lawlor, D. A. BMC Med.
2022, 20, 210.
(35) Würtz, P.; Wang, Q.; Kangas, A. J.; Richmond, R. C.; Skarp, J.;
Tiainen, M.; Tynkkynen, T.; Soininen, P.; Havulinna, A. S.; Kaakinen,
M.; Viikari, J. S.; Savolainen, M. J.; Kähönen, M.; Lehtimäki, T.;
Männistö, S.; Blankenberg, S.; Zeller, T.; Laitinen, J.; Pouta, A.;
Mäntyselkä, P.; Vanhala, M.; Elliott, P.; Pietiläinen, K. H.; Ripatti, S.;
Salomaa, V.; Raitakari, O. T.; Järvelin, M.-R.; Smith, G. D.; Ala-
Korpela, M. PLoS Med. 2014, 11, No. e1001765.
(36) Kettunen, J.; Holmes, M. V.; Allara, E.; Anufrieva, O.;
Ohukainen, P.; Oliver-Williams, C.; Wang, Q.; Tillin, T.; Hughes,
A. D.; Kähönen, M.; Lehtimäki, T.; Viikari, J.; Raitakari, O. T.;
Salomaa, V.; Järvelin, M.-R.; Perola, M.; Smith, G. D.; Chaturvedi, N.;
Danesh, J.; Angelantonio, E. D.; Butterworth, A. S.; Ala-Korpela, M.
PLoS Biol. 2019, 17, No. e3000572.
(37) Kettunen, J.; Demirkan, A.; Würtz, P.; Draisma, H. H. M.;
Haller, T.; Rawal, R.; Vaarhorst, A.; Kangas, A. J.; Lyytikäinen, L.-P.;
Pirinen, M.; Pool, R.; Sarin, A.-P.; Soininen, P.; Tukiainen, T.; Wang,
Q.; Tiainen, M.; Tynkkynen, T.; Amin, N.; Zeller, T.; Beekman, M.;
Deelen, J.; van Dijk, K. W.; Esko, T.; Hottenga, J.-J.; van Leeuwen, E.
M.; Lehtimäki, T.; Mihailov, E.; Rose, R. J.; de Craen, A. J. M.;
Gieger, C.; Kähönen, M.; Perola, M.; Blankenberg, S.; Savolainen, M.
J.; Verhoeven, A.; Viikari, J.; Willemsen, G.; Boomsma, D. I.; van
Duijn, C. M.; Eriksson, J.; Jula, A.; Järvelin, M.-R.; Kaprio, J.;
Metspalu, A.; Raitakari, O.; Salomaa, V.; Slagboom, P. E.;
Waldenberger, M.; Ripatti, S.; Ala-Korpela, M. Nat. Commun. 2016,
7, 11122.
(38) Small, A. M.; Pournamdari, A.; Melloni, G. E. M.; Scirica, B.
M.; Bhatt, D. L.; Raz, I.; Braunwald, E.; Giugliano, R. P.; Sabatine, M.
S.; Peloso, G. M.; Marston, N. A.; Natarajan, P. JAMA Cardiol. 2024,
9, 385−391.
(39) Stegemann, C.; Pechlaner, R.; Willeit, P.; Langley, S. R.;
Mangino, M.; Mayr, U.; Menni, C.; Moayyeri, A.; Santer, P.; Rungger,
G.; Spector, T. D.; Willeit, J.; Kiechl, S.; Mayr, M. Circulation 2014,
129, 1821−1831.
(40) Brown, J. M.; Hazen, S. L. Circulation 2014, 129, 1799−1803.
Analytical Chemistry pubs.acs.org/ac Article
https://doi.org/10.1021/acs.analchem.4c03229
Anal. Chem. 2024, 96, 20362−20370
20370