# Using Metrics of a Mixture Effect and Nutrition from an Observational Study for Consideration towards Causal Inference

^{1}

^{2}

^{3}

^{4}

^{5}

^{6}

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Methods

#### 2.1. Study Population and Cognitive Assessment

#### 2.2. Collection of Prenatal Blood and Urine Samples

#### 2.3. My Nutrition Index (MNI) and Covariates

#### 2.4. Statistical Analysis

#### 2.5. G-Computation

## 3. Results

## 4. Discussion

_{2}or PM

_{2.5}and childhood asthma incidence. They asked, “How would the incidence rate of asthma in our participants change if we could modify their exposure to regional NO

_{2}(or PM

_{2.5})?” Instead of focusing on beta coefficients relating conditional incidence rate ratios for a one-unit change in air pollution exposure, they presented a population intervention measure that estimates asthma incidence rates had exposure been, for example, no higher than 20 ppb NO

_{2}. Their motivation, similar to ours, is to move beyond the report of point estimates to potentially improve the translation of the study to policymakers. That is, instead of only interpreting the results of the study in terms of the significance of a beta coefficient, the g methods which use an index to represent multi-dimensional components (e.g., environmental chemicals for the WQS index and dozens of nutrients for the MNI) allow researchers and decision makers to take an additional step using the metric to provide targets for changes. We considered several scenarios for reducing the WQS index to a target level—i.e., an overall cut to all non-persistent chemicals and cuts based on chemical classes. Persistent chemicals such as PFAS (called the “forever chemicals”) should also be considered for reduction strategies; however, their remediation is often ineffective but is attracting intensive research seeking effective technologies for their removal from the environment [27,28]. The illustration indicated that a severe cut is necessary to reduce environmental exposures to the “what if” target value. A cut of 70% of the non-persistent chemicals would no doubt be difficult to achieve, and removing all plasticizers alone is not enough to reach the target value. Use of a weighted index in such counterfactuals may complement current strategies for risk management which generally focus on single chemicals, even though risk increases when mixtures are considered. Further, chemical combinations are generally considered based on convenient groupings (e.g., assuming additivity) and not based on human exposure patterns or unintentional mixtures [29].

## 5. Conclusions

## Author Contributions

## Funding

## Institutional Review Board Statement

## Informed Consent Statement

## Data Availability Statement

## Conflicts of Interest

## Appendix A

Matrix | Chemical Type | Compound (Further Description) | Abbreviation | LOD/LOQ ^{a} | % ≥LOD | GM | GSD | WQS ^{f} |
---|---|---|---|---|---|---|---|---|

Urine | Phenols | 2,4,4′-trichloro-2′-hydroxydiphenyl ether | Triclosan | 0.100 | 92 | 1.34 | 9.9 | x |

bisphenol A | BPA | 0.050 | 100 | 1.52 | 2.4 | x | ||

4,4-bisphenol F (BPA replacement analogue) | BPF | 0.024 | 92 | 0.16 | 5.3 | x | ||

bisphenol S (BPA replacement analogue) | BPS | 0.009 | 98 | 0.07 | 3.0 | x | ||

Plasticizers (Phthalate & non-phthalate) | monoethyl phthalate | MEP | 0.010 | 100 | 63.6 | 3.0 | x | |

monobutyl phthalate | MBP | 0.100 | 100 | 67.6 | 2.2 | x | ||

monobenzyl phthalate | MBzP | 0.040 | 100 | 15.5 | 3.0 | x | ||

mono(2-ethylhexyl) phthalate | MEHP | 0.100 | 100 | - | - | |||

mono(2-ethyl-5-hydroxyhexyl) phthalate | MEHHP | 0.020 | 100 | - | - | |||

mono(2-ethyl-5-oxohexyl) phthalate | MEOHP | 0.030 | 100 | - | - | |||

mono(2-ethyl-5-carboxypentyl) phthalate | MECPP | 0.020 | 100 | - | - | |||

di-(2-ethylhexyl) phthalate (parent compound) | DEHP ^{b} | - | - | 63.6 | 2.4 | x | ||

mono(hydroxy-iso-nonyl) phthalate | MHiNP | 0.020 | 100 | - | - | |||

mono(oxo-iso-nonyl) phthalate | MOiNP | 0.010 | 100 | - | - | |||

mono(carboxy-iso-octyl) phthalate | MCiOP | 0.020 | 100 | - | - | |||

diisononyl phthalate (parent compound) | DINP ^{c} | - | - | 42.4 | 2.6 | x | ||

monohydroxyisodecyl phthalate | MHiDP | 0.031 | 100 | 1.24 | 2.7 | x | ||

monocarboxyisononyl phthalate | MCiNP | 0.031 | 100 | 0.68 | 2.4 | x | ||

2-4-methyl-7-oxyooctyl-oxycarbonyl-cyclohexane carboxylic acid (phthalate replacement) | MOiNCH | 0.023 | 99 | 0.30 | 4.0 | x | ||

diphenylphosphate (organophosphate flame retardant) | DPHP | 0.042 | 100 | 1.34 | 2.5 | x | ||

Other Short-Lived | 3,5,6-trichloro-2-pyridinol (organophosphate pesticide) | TCP | 0.035 | 100 | 1.25 | 2.5 | x | |

3-phenoxybenzoic acid (pyrethroid pesticide) | PBA | 0.017 | 99 | 0.16 | 2.7 | x | ||

2-hydroxyphenanthrene (polycyclic aromatic hydrocarbon) | 2OHPH | 0.003 | 100 | 0.20 | 2.3 | x | ||

Serum | Perfluoro-alkyl Substances (PFAS) | perfluorooctanoic acid | PFOA | 0.020 | 100 | 1.58 | 1.8 | x |

perfluorooctane sulfonate | PFOS | 0.060 | 100 | 5.37 | 1.7 | x | ||

perfluorononanoic acid | PFNA | 0.010 | 100 | 0.53 | 1.7 | x | ||

perfluorodecanoic acid | PFDA | 0.020 | 100 | 0.26 | 1.6 | x | ||

perfluoroundecanoic acid | PFUnDA | 0.020 | 99 | 0.22 | 1.9 | x | ||

perfluorohexanesulfonic acid | PFHxS | 0.030 | 100 | 1.32 | 1.8 | x | ||

Plasma | Persistent Chlorinated | hexachlorobenzene | HCB | 0.005 | 100 | 0.05 | 1.4 | x |

trans-nonachlor | Nonachlor | 0.005 | 78 | 0.01 | 2.7 | x | ||

dichlorodiphenyltrichloroethane alone | DDTa | 0.015 | 99 | - | - | |||

dichlorodiphenyldichloroethylene | DDE | 0.040 | 8 | - | - | |||

total dichlorodiphenyltrichloroethane | DDT ^{d} | - | - | 0.18 | 2.0 | x | ||

polychlorinated biphenyl 74 | PCB 74 | 0.005 | 73 | - | - | |||

polychlorinated biphenyl 99 | PCB 99 | 0.005 | 81 | - | - | |||

polychlorinated biphenyl 118 | PCB 118 | 0.005 | 99 | - | - | |||

polychlorinated biphenyl 138 | PCB 138 | 0.005 | 100 | - | - | |||

polychlorinated biphenyl 153 | PCB 153 | 0.005 | 100 | - | - | |||

polychlorinated biphenyl 156 | PCB 156 | 0.005 | 90 | - | - | |||

polychlorinated biphenyl 170 | PCB 170 | 0.005 | 100 | - | - | |||

polychlorinated biphenyl 180 | PCB 180 | 0.005 | 100 | - | - | |||

polychlorinated biphenyl 183 | PCB 183 | 0.005 | 76 | - | - | |||

polychlorinated biphenyl 187 | PCB 187 | 0.005 | 98 | - | - | |||

total polychlorinated biphenyls | PCB ^{e} | - | - | 0.37 | 1.7 | x |

^{a}LOD reported for all urine and serum compounds, LOQ reported for plasma compounds.

^{b}Molar sum of metabolites: mono-2-ethylhexyl, mono(2-ethyl-5-hydroxyhexyl), mono(2-carboxymethylhexyl), mono(2-ethyl-5-oxohexyl), and mono(2-ethyl-5-carboxypentyl) phthalates.

^{c}Molar sum of metabolites: mono(hydroxyisononyl), mono(oxoisononyl), and mono(carboxyisooctyl) phthalates.

^{d}Sum of DDT and its metabolite dichlorodiphenyldichloroethylene.

^{e}Sum of PCB congeners 74, 99, 118, 138, 153, 156, 170, 180, 183, 187.

^{f}Included in WQS analysis.

## References

- Maki, K.C.; Slavin, J.L.; Rains, T.M.; Kris-Etherton, P.M. Limitations of observational evidence: Implications for evidence-based dietary recommendations. Adv. Nutr.
**2014**, 5, 7–15. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Robins, J.M.; Hernan, M.A.; Brumback, B. Marginal structural models and causal inference in epidemiology. Epidemiology
**2000**, 11, 550–560. [Google Scholar] [CrossRef] [PubMed] - Snowden, J.M.; Rose, S.; Mortimer, K.M. Implementation of g-computation on a simulated data set: Demonstration of a causal inference technique. Am. J. Epidemiol.
**2011**, 173, 731–738. [Google Scholar] [CrossRef] [Green Version] - Naimi, A.I.; Cole, S.R.; Cole, E.H. An introduction to g methods. Int. J. Epidemiol.
**2017**, 46, 756–762. [Google Scholar] [CrossRef] - Daniel, R.M.; Cousens, S.N.; De Stavola, B.L.; Kenward, M.G.; Sterne, J.A.C. Methods for dealing with time-dependent confounding. Stat. Med.
**2013**, 32, 1584–1618. [Google Scholar] [CrossRef] - Wang, A.; Nianogo, R.A.; Arah, O.A. g-computation of average treatment effects on the treated and the untreated. BMC Med. Res. Methodol.
**2017**, 17, 1–5. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Gennings, C.; Wolk, A.; Hakansson, N.; Lindh, C.; Bornehag, C.G. Contrasting prenatal nutrition and environmental exposures in association with birth weight and cognitive function in children at 7 years. BMJ Nutr. Prev. Health
**2020**, 3, 162. [Google Scholar] [CrossRef] [PubMed] - Gennings, C.; Manning, L.; Keefer, L. Evaulation of My Nutrition Index in an IBD Patient Population. J. Nutr. Med. Diet Care
**2021**, 7, 49. [Google Scholar] - Busgang, S.A.; Malin, A.J.; Gennings, C. My Nutrition Index: A method for measuring optimal daily nutrient intake. BMC Nutr. 2022; in press. [Google Scholar]
- Tanner, E.M.; Hallerback, M.U.; Wikstrom, S.; Lindh, C.; Kiviranta, H.; Gennings, C.; Bornehag, C.G. Early prenatal exposure to suspected endocrine disruptor mixtures is associated with lower IQ at age seven. Environ. Int.
**2020**, 134, 105185. [Google Scholar] [CrossRef] - Bornehag, C.G.; Moniruzzaman, S.; Larsson, M.; Lindstrom, C.B.; Hasselgren, M.; Bodin, A.; Kobyletzkic, L.B.V.; Carlstedt, F.; Lundin, F.; Nånberg, E.; et al. The SELMA study: A birth cohort study in Sweden following more than 2000 mother-child pairs. Paediatr. Perinat. Epidemiol.
**2012**, 26, 456–467. [Google Scholar] [CrossRef] - Wechsler, D. Wechsler Intellignece Scale in Children, 4th ed.; The Psychological Corporation: San Antonio, TX, USA, 2003. [Google Scholar]
- Marinopoulou, M.; Billstedt, E.; Lin, P.I.; Hallerback, M.; Bornehag, C.G. Number of words at age 2.5 years is associated with intellectual functioning at age 7 years in the SELMA study. Acta Paediatr.
**2021**, 110, 2134–2141. [Google Scholar] [CrossRef] [PubMed] - Van der Elst, W.; Ouwehand, C.; van Rijn, P.; Lee, N.; Van Boxtel, M.; Jolles, J. The shortened Raven Standard Progressive Matrices: Item response theory-based psychometric analyses and normative. Assessment
**2013**, 20, 48–59. [Google Scholar] [CrossRef] - Bouma, A.; Mulder, J.; Lindeboom, J. Neuropsychologische Diagnostiek: Handboek [Neuropsychological Assessment: Manual]; Swets & Zeitlinger: Lisse, The Netherlands, 1996. [Google Scholar]
- Raven, J.; Raven, J.; Court, J. Manual for Raven’s Progressive Matrices and Vocabulary; Oxford Psychologists Press: Oxford, UK, 1998. [Google Scholar]
- Carrico, C.; Gennings, C.; Wheeler, D.C.; Factor-Litvak, P. Characterization of weighted quantile sum regression for highly correlated data in a risk analysis setting. J. Agric. Biol. Environ. Stat.
**2015**, 20, 100–120. [Google Scholar] [CrossRef] [PubMed] - Bello, G. Application and Extension of Weighted Quantile Sum Regresion for the Development of a Clinical Risk Prediction Tool. Ph.D. Thesis, Virginia Commonwealth University-Graduate School at VCU Scholars Compass, Richmond, VA, USA, 2014. [Google Scholar]
- Curtin, P.; Kellogg, J.J.; Cech, N.B.; Gennings, C. A random subset implementation of weighted quantile sum (WQSRS) regression for analysis of high-dimensional mixtures. Commun. Stat.-Simul. Comput.
**2019**, 50, 1119–1134. [Google Scholar] [CrossRef] - Tanner, E.M.; Bornehag, C.G.; Gennings, C. Repeated holdout validation for weighted quantile sum regression. MethodsX
**2019**, 6, 2855–2860. [Google Scholar] [CrossRef] - Zou, H.; Hastie, T. Regularization and variable selection via the elastic net. J. R. Soc. Ser. B
**2005**, 67, 301–320. [Google Scholar] [CrossRef] [Green Version] - Renzetti, S.; Curtin, P.; Just, A.C.; Bello, G.; Gennings, C. gWQS: Generalized Weighted Quantile Sum Regression. R Package, Version 2.0.0. 2019. Available online: https://rdrr.io/cran/gWQS/ (accessed on 1 June 2021).
- Cowell, W.; Colicino, E.; Levin-Schwartz, Y.; Enlow, M.B.; Amarasiriwardena, C.; Andra, S.S.; Wright, R.J. Prenatal metal mixtures and sex-specific infant negative affectivity. Env. Epidemiol.
**2021**, 5, e147. [Google Scholar] [CrossRef] - Svensson, K.; Tanner, E.; Gennings, C.; Lindh, C.; Kiviranta, H.; Wikstrom, S.; Bornehag, C.G. Prenatal exposures to mixtures of endocrine disrupting chemicals and children’s weight trajectory up to age 5.5 in the SELMA study. Sci. Rep.
**2021**, 11, 11036. [Google Scholar] [CrossRef] - Chatton, A.; Borgne, F.L.; Leyrat, C.; Gillaizeau, F.; Rousseau, C.; Barbin, L.; Laplaud, D.; Léger, M.; Giraudeau, B.; Foucher, Y. g-computation, propensity score-based methods, and targeted maximum likelihood estimator for causal inference with different covariates sets: A comparative simulation study. Sci. Rep.
**2020**, 10, 9219. [Google Scholar] [CrossRef] - Garcia, E.; Urman, R.; Berhane, K.; McConnell, R.; Gilliland, F. Effects of policy-driven hypothetical air pollutant interventions on childhood asthma incidence in southern California. Proc. Natl. Acad. Sci. USA
**2019**, 116, 15883–15888. [Google Scholar] [CrossRef] [Green Version] - Dhore, R.; Murthy, G.S. Per/polyfluoroalkyl substances production, applications and environmental impacts. Bioresour. Technol.
**2021**, 341, 125808. [Google Scholar] [CrossRef] [PubMed] - Zhang, Z.; Sarkar, D.; Biswas, J.K.; Datta, R. Biodegradation of per- and polyfluoroalkyl substances (PFAS): A review. Bioresour. Technol.
**2022**, 344, 126223. [Google Scholar] [CrossRef] [PubMed] - Ruden, C. Future chemical risk management: Accounting for combination effects and assessing chemicals in groups. In Swedish Government Official Reports; SOU 2019:45; Ministry Publications Series; 2019. Available online: https://www.government.se/4adb1a/contentassets/ee36b3e49c354bb8967f3a33a2d5ca50/future-chemical-risk-management---accounting-for-combination-effects-and-assessing-chemicals-in-groups-sou-201945 (accessed on 7 December 2021).
- Woodruff, T.J.; Zota, A.R.; Schwartz, J.M. Environmental chemicals in pregnant women in the United States: NHANES 2003–2004. Environ. Health Perspect.
**2011**, 119, 878–885. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Levin-Schwartz, Y.; Gennings, C.; Schnaas, L.; del Carmen Hernandez Chavez, M.; Bellinger, D.C.; Tellez-Rojo, M.M.; Baccarelli, A.A.; Wright, R.O. Time-varying associations between prenatal metal mixtures and rapid visual processing in children. Environ. Health
**2019**, 18, 92. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Bornehag, C.G.; Engdahl, E.; Hallerback, M.U.; Wikstrom, S.; Lindh, C.; Ruegg, J.; Tanner, E.; Gennings, C. Prenatal exposure to bisphenols and cognitive function in children at 7 years of age in the Swedish SELMA study. Environ. Int.
**2021**, 150, 106433. [Google Scholar] [CrossRef] [PubMed] - Braun, J.M.; Gennings, C.; Hauser, R.; Webster, T.F. What Can Epidemiological Studies Tell Us about the Impact of Chemical Mixtures on Human Health? Environ. Health Perspect.
**2016**, 124, A6–A9. [Google Scholar] [CrossRef] [Green Version] - Silva, E.; Rajapakse, N.; Kortenkamp, A. Something from “nothing”--eight weak estrogenic chemicals combined at concentrations below NOECs produce significant mixture effects. Environ. Sci. Technol.
**2002**, 36, 1751–1756. [Google Scholar] [CrossRef] - Neter, J.; Kutner, M.; Nachtsheim, C.; Wasserman, W. Applied Linear Statistical Models, 4th ed.; Irwin: Chicago, IL, USA, 1996. [Google Scholar]
- Tu, Y.; Gunnell, D.; Gilthorpe, M. Simpson’s Paradox, Lord’s Paradox, and Suppression Effects are the same phenomenon--the reversal paradox. Emerg. Themes. Epidemiol.
**2008**, 5, 2. [Google Scholar] [CrossRef] [Green Version] - Czarnota, J.; Gennings, C.; Wheeler, D.C. Assessment of weighted quantile sum regression for modeling chemical mixtures and cancer risk. Cancer Inform.
**2015**, 14, 159–171. [Google Scholar] [CrossRef] [Green Version] - Arora, M.; Reichenberg, A.; Willfors, C.; Austin, C.; Gennings, C.; Berggren, S.; Lichtenstein, P.; Anckarsäter, H.; Tammimies, K.; Bolte, S. Fetal and postnatal metal dysregulation in autism. Nat. Commun.
**2017**, 8, 1–10. [Google Scholar] [CrossRef] - Keil, A.P.; Buckley, J.P.; O’Brien, K.M.; Ferguson, K.K.; Zhao, S.; White, A.J. A Quantile-Based g-Computation Approach to Addressing the Effects of Exposure Mixtures. Environ. Health Perspect.
**2020**, 128, 47004. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Gennings, C. Comment on “A Quantile-Based g-Computation Approach to Addressing the Effects of Exposure Mixtures”. Environ. Health Perspect.
**2021**, 129, 38001. [Google Scholar] [CrossRef] [PubMed]

**Figure 1.**Estimated weight distribution in a WQS stratified regression model for 26 prenatal chemicals and 7-year IQ, using 100 repeated holdout validation datasets, for (

**A**) boys; (

**B**) girls; and (

**C**) a divergent plot comparing the mean estimated sex-specific weights. Notes (

**A**,

**B**) Bars correspond to the right axis and indicate the percent of times a chemical exceeded the concern threshold in 100 repeated holdouts. Data points, boxplots, and diamonds correspond to the left axis. Data points indicate weights for each of the 100 holdouts. Box plots show 25th, 50th, and 75th percentiles, and whiskers show 10th and 90th percentiles of weights for the 100 holdouts. Closed diamonds show mean weights for the 100 holdouts; (

**C**) The dotted lines represent the threshold guideline from the equi-weighted index (i.e., 1/(2c)), where c is the number of components.

**Figure 2.**(

**A**) Histogram of the stratified interaction WQS index in the SELMA cohort with reference line at selected target value (i.e., one SD below the WQS mean), (

**B**) percentage of the WQS index per subject due to persistent chemicals (i.e., PFAS and persistent chlorinated), phenols, plasticizers, and other short-lived chemicals; (

**C**) the distribution of the WQS index due to the persistent and 30% of the non-persistent chemicals, i.e., a 70% cut to non-persistent chemicals; the distribution of the WQS index (

**D**) eliminating the plasticizers, (

**E**) eliminating the plasticizers and the other short-lived compounds, and (

**F**) eliminating the plasticizers and the phenols.

**Table 1.**Summary statistics of population characteristics using the SELMA pregnancy cohort (N = 678). (*) The WQS index is derived from a WQS sex-stratified interaction model of 26 EDCs associated with child IQ at 7 years of age, adjusted by covariates.

Mean | SD | ||
---|---|---|---|

Exposure | WQS index associated with 7-year IQ (sex-stratified, decile-scaled) * | 2.24 | 0.50 |

Maternal characteristics | Graduated college n (%) | 467 (69) | |

My Nutrition Index (MNI) | 66.8 | 14.0 | |

Energy (kcals) | 1895 | 545 | |

Age at birth (years) | 31.3 | 4.6 | |

Weight in 1st trimester of pregnancy (kg) | 68.8 | 13.5 | |

IQ (Raven) | 114.8 | 14.9 | |

Parity | 1.8 | 0.86 | |

Smoked in 1st trimester pregnancy n (%) | 74 (11) | ||

Creatinine (mmol/L) | 10.4 | 4.7 | |

Child characteristics | Female n (%) | 346 (51) | |

Premature birth n (%) | 25 (3.7) | ||

Full Scale WISC IQ at 7 years | 99.9 | 12.7 |

**Table 2.**Parameter estimates (mean, standard error, 2.5 percentile, 97.5 percentile) from WQS sex-stratified interaction regression across 100 holdout datasets. The slope associated with WQS is for males; the interaction between WQS and sex is the difference in slopes between boys and girls.

Parameter | Estimate | Std. Error | 2.5% | 97.5% |
---|---|---|---|---|

(Intercept) | 88.700 | 4.530 | 80.700 | 96.700 |

WQS | −2.130 | 1.110 | −4.270 | −0.359 |

Female | −0.622 | 3.190 | −5.740 | 5.630 |

MNI | 0.073 | 0.027 | 0.018 | 0.118 |

Energy | 0.000 | 0.001 | −0.001 | 0.001 |

Mom Age (at birth) | −0.158 | 0.089 | −0.324 | 0.004 |

Mom Weight | −0.098 | 0.029 | −0.156 | −0.040 |

Mom Educ | 4.790 | 0.862 | 3.090 | 6.470 |

Mom IQ | 0.158 | 0.026 | 0.104 | 0.205 |

Smoker | −2.100 | 1.320 | −4.420 | 0.536 |

WQS:Female | 1.980 | 1.640 | −1.280 | 5.130 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Gennings, C.; Svensson, K.; Wolk, A.; Lindh, C.; Kiviranta, H.; Bornehag, C.-G.
Using Metrics of a Mixture Effect and Nutrition from an Observational Study for Consideration towards Causal Inference. *Int. J. Environ. Res. Public Health* **2022**, *19*, 2273.
https://doi.org/10.3390/ijerph19042273

**AMA Style**

Gennings C, Svensson K, Wolk A, Lindh C, Kiviranta H, Bornehag C-G.
Using Metrics of a Mixture Effect and Nutrition from an Observational Study for Consideration towards Causal Inference. *International Journal of Environmental Research and Public Health*. 2022; 19(4):2273.
https://doi.org/10.3390/ijerph19042273

**Chicago/Turabian Style**

Gennings, Chris, Katherine Svensson, Alicja Wolk, Christian Lindh, Hannu Kiviranta, and Carl-Gustaf Bornehag.
2022. "Using Metrics of a Mixture Effect and Nutrition from an Observational Study for Consideration towards Causal Inference" *International Journal of Environmental Research and Public Health* 19, no. 4: 2273.
https://doi.org/10.3390/ijerph19042273