📄 medicine

No causal effect of genetically predicted serum 25-hydroxyvitamin D on tuberculosis risk: a bidirectional two-sample Mendelian randomization study

This bidirectional two-sample Mendelian randomization study utilizing European ancestry data found no causal evidence that genetically predicted serum 25-hydroxyvitamin D levels influence tuberculosis risk, suggesting the previously observed strong association is likely driven by confounding and reverse causation rather than a direct causal effect.

Original authors: Kaweesi Calvin Nantalaga

Published 2026-06-28

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Kaweesi Calvin Nantalaga

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Question: Is Vitamin D a Shield Against Tuberculosis?

For a long time, doctors and scientists have noticed a pattern: people who get sick with Tuberculosis (TB) often have very low levels of Vitamin D. It's like walking into a room and seeing that almost everyone who is sick is also wearing a blue hat. Naturally, people assumed the blue hat (low Vitamin D) was causing the sickness, or at least that wearing a red hat (taking Vitamin D) would stop the sickness.

However, there was a problem with this thinking. Maybe the sickness made them lose their blue hats? Or maybe a third factor, like poor nutrition or living in a crowded, dark apartment, caused both the low Vitamin D and the TB? This is the "chicken or the egg" problem.

The Detective Tool: Genetic "Randomization"

To solve this mystery, the researcher used a special detective tool called Mendelian Randomization.

Imagine you are trying to figure out if eating spicy food causes heartburn. If you just ask people, "Do you eat spicy food and get heartburn?", you might get confused because spicy food eaters might also drink a lot of soda.

But what if you could look at people's DNA? Your genes are like a lottery ticket drawn at the moment of conception. Some people are genetically "dealt" a hand that makes their bodies naturally produce less Vitamin D, while others are dealt a hand that makes them produce more. This happens randomly, long before anyone gets sick, and it has nothing to do with their diet, income, or lifestyle.

By looking at these "lottery tickets," the researcher could see: Do the people who were genetically dealt a "low Vitamin D" hand actually get TB more often than those with a "high Vitamin D" hand?

The Investigation: What They Found

The researcher ran a massive digital experiment using data from nearly half a million people of European ancestry.

The Setup: They identified 114 specific genetic "switches" that control how much Vitamin D a person has. These switches were strong and reliable.
The Test: They checked if people with the "low Vitamin D" switches were more likely to have TB.
The Result: Nothing happened.

It was as if they checked the lottery tickets and found that people who were genetically destined to have low Vitamin D had exactly the same chance of getting TB as people destined to have high Vitamin D. The odds were 1.00, which means "no change."

The researcher used five different mathematical methods to double-check, like using five different maps to find a treasure. Every single map said the same thing: There is no causal link.

Why the Confusion? (The "Reverse Causation" Twist)

So, why did the old studies show that TB patients have low Vitamin D?

The paper suggests that the sickness itself is the culprit. When you have active TB, your body goes into a state of high inflammation, and you might stay indoors more because you feel sick. This inflammation and lack of sun exposure drain your Vitamin D levels.

The Analogy: Imagine a fire alarm (TB) going off in a building. The fire alarm causes the sprinklers (low Vitamin D) to turn on. If you see sprinklers running, you might think, "The sprinklers caused the fire!" But actually, the fire caused the sprinklers. The study found that the "fire" (TB) doesn't seem to be caused by the "sprinklers" (Vitamin D) being off; rather, the fire turns the sprinklers off.

The "Reverse" Check

The researcher also asked the opposite question: Does having a genetic tendency to get TB cause your Vitamin D levels to drop?

They tried to check this, but the genetic data for TB was too "fuzzy" (weak) to give a definite answer. However, the results didn't show a strong link in that direction either.

The Catch: Who Was Studied?

There is one very important limit to this study. All the data came from people of European ancestry.

Think of it like testing a car engine only in the snow. You might conclude the engine works perfectly in the cold. But you don't know if it will work in the desert heat.

The Gap: The biggest burden of Tuberculosis is in Sub-Saharan Africa. The genes that control Vitamin D work differently in different populations.
The Warning: The researcher explicitly states that we cannot assume this result applies to African populations yet. We need to run the same "genetic lottery" check on people from high-TB regions before we can say for sure that Vitamin D doesn't help them.

The Bottom Line

The Finding: In the people studied, having low Vitamin D due to your genes does not cause Tuberculosis.
The Implication: The strong link seen in regular observation is likely because getting sick with TB lowers your Vitamin D, or because other factors (like poverty or diet) cause both.
The Caveat: This conclusion is based on European data. We need more research in African and other high-burden populations to see if the rules are different there.

In short: Genetics says Vitamin D isn't the shield against TB that we thought it was, but we need to check if this holds true for the parts of the world where TB is most common.

Technical Summary: No Causal Effect of Genetically Predicted Serum 25-Hydroxyvitamin D on Tuberculosis Risk

Problem Statement
Observational epidemiology has consistently identified a strong association between low vitamin D status and increased risk of tuberculosis (TB), with meta-analyses suggesting vitamin D deficiency is approximately three times more common in TB patients than in controls. However, this association is confounded by factors such as sunlight exposure, nutrition, HIV co-infection, and socioeconomic status. Furthermore, reverse causation is a significant concern, as the systemic inflammation and reduced outdoor activity associated with active TB can lower serum 25-hydroxyvitamin D [25(OH)D] levels. While randomized controlled trials (RCTs) of vitamin D supplementation have generally failed to show benefits for TB treatment outcomes, the causal nature of the relationship remains unresolved. This study aims to determine whether genetically predicted serum 25(OH)D causally influences TB risk, and conversely, whether genetic liability to TB influences 25(OH)D levels, using Mendelian randomization (MR) to overcome the limitations of observational designs.

Methodology
The study employed a bidirectional two-sample Mendelian randomization design using summary statistics from genome-wide association studies (GWAS) restricted to populations of European ancestry to minimize population stratification.

Data Sources:
- Exposure (Forward): Serum 25(OH)D data from a GWAS of 496,946 individuals (Revez et al.).
- Outcome (Forward): Tuberculosis data comprising 2,473 cases and 482,125 controls.
- Reverse Direction: TB as the exposure and 25(OH)D as the outcome.
Instrument Selection:
- Forward Analysis: 114 independent single-nucleotide polymorphisms (SNPs) associated with 25(OH)D at genome-wide significance ( $p < 5 \times 10^{-8}$ ) were selected. These instruments were located in canonical genes involved in vitamin D metabolism (GC, CYP2R1, DHCR7-NADSYN1, CYP24A1). The mean F-statistic was 198.3, indicating strong instruments free of weak-instrument bias.
- Reverse Analysis: Due to a lack of genome-wide significant SNPs for TB, the threshold was relaxed to $p < 5 \times 10^{-6}$ , yielding 11 SNPs. These results are treated as exploratory due to weaker instrument strength.
Statistical Analysis:
- Primary Estimator: Inverse-variance weighted (IVW) random-effects regression.
- Sensitivity Analyses: MR-Egger, weighted median, simple mode, and weighted mode estimators were used to test robustness against different assumptions regarding instrument validity.
- Bias Assessment: Cochran's Q test assessed heterogeneity; MR-Egger intercept and MR-PRESSO global tests assessed directional horizontal pleiotropy. Leave-one-out analysis and funnel plots were used to detect influential outliers.
- Power: A priori power calculations indicated the study had 80% power to detect an odds ratio (OR) of 1.30 and near-complete power (>99%) to detect an effect size matching the observational literature (OR $\approx$ 3.23).

Key Results

Forward Direction (25(OH)D $\rightarrow$ TB): Genetically predicted serum 25(OH)D showed no causal association with TB risk.
- IVW Estimate: OR 1.00 (95% CI: 0.999 to 1.001; $p = 0.88$ ) per standard-deviation increase in 25(OH)D.
- Consistency: All five analytical methods (IVW, MR-Egger, weighted median, simple mode, weighted mode) yielded concordant null estimates.
- Validity Checks: No evidence of heterogeneity (Cochran's Q $p = 0.63$ ) or directional pleiotropy (MR-Egger intercept $p = 0.55$ ; MR-PRESSO $p = 0.61$ ).
Reverse Direction (TB $\rightarrow$ 25(OH)D): The analysis found no detectable reverse causal effect, though the results were limited by weak instruments.
- IVW Estimate: OR 1.30 (95% CI: 0.09 to 19.3; $p = 0.85$ ).
- The wide confidence intervals reflect the underpowered nature of the reverse analysis due to the relaxed significance threshold for instrument selection.

Significance and Claims
The paper concludes that the strong observational association between vitamin D deficiency and tuberculosis is likely driven by confounding and reverse causation rather than a causal effect of vitamin D.

Reconciliation with Biology: The authors suggest that while vitamin D has known antimicrobial properties (e.g., inducing cathelicidin), these mechanisms may operate through short-term, local, or threshold-dependent pathways that lifelong genetic exposure does not capture. Alternatively, low 25(OH)D in TB patients may be a consequence of the disease state rather than a cause.
Limitations and Scope: The authors explicitly state that these findings are specific to European ancestry populations. They identify the absence of data from African and other high-burden populations as a critical gap, cautioning against generalizing these null results to regions with the highest TB burden without dedicated ancestry-matched studies.
Future Directions: The study calls for GWAS of 25(OH)D and TB in under-represented populations to enable valid MR in high-burden settings and suggests exploring non-linear MR to detect potential threshold effects.

In summary, this study provides robust genetic evidence that lifelong elevated serum 25(OH)D levels do not causally reduce the risk of tuberculosis in European populations, challenging the rationale for vitamin D supplementation as a primary preventative strategy for TB based solely on observational associations.