Higher plasma concentrations of the vitamin B-6 marker pyridoxal 5'-phosphate (PLP) have been associated with reduced colorectal cancer (CRC) risk. Inflammatory processes, including vitamin B-6 catabolism, could explain such findings. We investigated 3 biomarkers of vitamin B-6 status in relation to CRC risk. This was a prospective case-control study of 613 CRC cases and 1190 matched controls nested within the Northern Sweden Health and Disease Study ( = 114,679). Participants were followed from 1985 to 2009, and the median follow-up from baseline to CRC diagnosis was 8.2 y. PLP, pyridoxal, pyridoxic acid (PA), 3-hydroxykynurenine, and xanthurenic acids (XAs) were measured in plasma with the use of liquid chromatography-tandem mass spectrometry. We calculated relative and absolute risks of CRC for PLP and the ratios 3-hydroxykynurenine:XA (HK:XA), an inverse marker of functional vitamin B-6 status, and PA:(PLP + pyridoxal) (PAr), a marker of inflammation and oxidative stress and an inverse marker of vitamin B-6 status. Plasma PLP concentrations were associated with a reduced CRC risk for the third compared with the first quartile and for PLP sufficiency compared with deficiency [OR: 0.60 (95% CI: 0.44, 0.81) and OR: 0.55 (95% CI: 0.37, 0.81), respectively]. HK:XA and PAr were both associated with increased CRC risk [OR: 1.48 (95% CI: 1.08, 2.02) and OR: 1.50 (95% CI: 1.10, 2.04), respectively] for the fourth compared with the first quartile. For HK:XA and PAr, the findings were mainly observed in study participants with <10.5 y of follow-up between sampling and diagnosis. Vitamin B-6 deficiency as measured by plasma PLP is associated with a clear increase in CRC risk. Furthermore, our analyses of novel markers of functional vitamin B-6 status and vitamin B-6-associated oxidative stress and inflammation suggest a role in tumor progression rather than initiation.

Propensity score-based estimators are increasingly used for causal inference in observational studies. However, model selection for propensity score estimation in high-dimensional data has received little attention. In these settings, propensity score models have traditionally been selected based on the goodness-of-fit for the treatment mechanism itself, without consideration of the causal parameter of interest. Collaborative minimum loss-based estimation is a novel methodology for causal inference that takes into account information on the causal parameter of interest when selecting a propensity score model. This ‘‘collaborative learning’’ considers variable associations with both treatment and outcome when selecting a propensity score model in order to minimize a bias-variance tradeoff in the estimated treatment effect. In this study, we introduce a novel approach for collaborative model selection when using the LASSO estimator for propensity score estimation in high-dimensional covariate settings. To demonstrate the importance of selecting the propensity score model collaboratively, we designed quasi-experiments based on a real electronic healthcare database, where only the potential outcomes were manually generated, and the treatment and baseline covariates remained unchanged. Results showed that the collaborative minimum loss-based estimation algorithm outperformed other competing estimators for both point estimation and confidence interval coverage. In addition, the propensity score model selected by collaborative minimum loss-based estimation could be applied to other propensity score-based estimators, which also resulted in substantive improvement for both point estimation and confidence interval coverage. We illustrate the discussed concepts through an empirical example comparing the effects of non-selective nonsteroidal anti-inflammatory drugs with selective COX-2 inhibitors on gastrointestinal complications in a population of Medicare beneficiaries.

To unbiasedly estimate a causal effect on an outcome unconfoundedness is often assumed. If there is sufficient knowledge on the underlying causal structure then existing confounder selection criteria can be used to select subsets of the observed pretreatment covariates, X, sufficient for unconfoundedness, if such subsets exist. Here, estimation of these target subsets is considered when the underlying causal structure is unknown. The proposed method is to model the causal structure by a probabilistic graphical model, for example, a Markov or Bayesian network, estimate this graph from observed data and select the target subsets given the estimated graph. The approach is evaluated by simulation both in a high-dimensional setting where unconfoundedness holds given X and in a setting where unconfoundedness only holds given subsets of X. Several common target subsets are investigated and the selected subsets are compared with respect to accuracy in estimating the average causal effect. The proposed method is implemented with existing software that can easily handle high-dimensional data, in terms of large samples and large number of covariates. The results from the simulation study show that, if unconfoundedness holds given X, this approach is very successful in selecting the target subsets, outperforming alternative approaches based on random forests and LASSO, and that the subset estimating the target subset containing all causes of outcome yields smallest MSE in the average causal effect estimation.

Supplemental Digital Content is available in the text

The anti‐inflammatory agent palmitoylethanolamide (PEA) reduces cyclooxygenase (COX) activity in vivo in a model of inflammatory pain. It is not known whether the compound reduces prostaglandin production in RAW264.7 cells, whether such an action is affected by compounds preventing the breakdown of endogenous PEA, whether other oxylipins are affected, or whether PEA produces direct effects upon the COX‐2 enzyme. RAW264.7 cells were treated with lipopolysaccharide and interferon‐γ to induce COX‐2. At the level of mRNA, COX‐2 was induced >1000‐fold following 24 h of the treatment. Coincubation with PEA (10 μmol/L) did not affect the levels of COX‐2, but reduced the levels of prostaglandins D2 and E2 as well as 11‐ and 15‐hydroxyeicosatetraenoic acid, which can also be synthesised by a COX‐2 pathway in macrophages. These effects were retained when hydrolysis of PEA to palmitic acid was blocked. Linoleic acid‐derived oxylipin levels were not affected by PEA. No direct effects of PEA upon the oxygenation of either arachidonic acid or 2‐arachidonoylglycerol by COX‐2 were found. It is concluded that in lipopolysaccharide and interferon‐γ‐stimulated RAW264.7 cells, PEA reduces the production of COX‐2‐derived oxylipins in a manner that is retained when its metabolism to palmitic acid is inhibited.