A US blood-chemistry trend has been presented as evidence that rising atmospheric carbon dioxide is entering human bodies. The underlying study found that average serum bicarbonate increased across National Health and Nutrition Examination Survey cycles from 1999 to 2020. It did not measure carbon dioxide in participants' breathing air or prove that the atmosphere caused the change.
The distinction is medically important. Serum bicarbonate, often reported as total CO2 on a chemistry panel, helps clinicians assess acid-base balance. It is not the same measurement as the partial pressure of carbon dioxide on an arterial blood gas, and a higher population average does not show that healthy people have chronic respiratory acidosis.
Alexander Larcombe and Phil Bierwirth published the analysis in Air Quality, Atmosphere & Health. They compared NHANES cycle averages for serum bicarbonate, calcium and phosphorus with the Mauna Loa atmospheric CO2 record and proposed rising exposure as one possible explanation for parallel trends.
The Study Averaged Survey Cycles, Not Individual Exposure
NHANES is a nationally representative US programme combining interviews, examinations and laboratory tests. The researchers used roughly 7,000 records per two-year cycle, covering participants from birth through age 80 and older, and calculated a mean for each blood marker at each time point.
The latest mean serum bicarbonate in the analysis was 25.3 milliequivalents per litre in 2019-2020, about 7% higher than in 1999-2000. Mean serum calcium declined by about 2%, while phosphorus declined by about 7% after the authors excluded a potentially spurious early phosphorus measurement.
Atmospheric CO2 rose over the same period. NOAA's Mauna Loa series is the longest direct record of atmospheric carbon dioxide and documents the continuing increase. That makes the comparison a legitimate question to investigate, but the two datasets operate at different levels: one is a remote atmospheric time series, and the other consists of changing samples of US residents.
No participant was followed for 21 years. The analysis did not connect a person's blood result to a monitor in that person's home, school, workplace or outdoor environment. It therefore cannot show whether people with greater CO2 exposure had larger bicarbonate changes than people with lower exposure.
The Average Remained Within the Cited Venous Range
The paper cited 30 mEq/L as an upper reference value for venous bicarbonate and projected that a linear continuation of the observed trend could reach that level around 2076. It similarly projected lower calcium and phosphorus reference limits later in the century.
Those projections are not diagnoses. The 2019-2020 bicarbonate mean of 25.3 mEq/L was below the cited venous threshold, and a reference interval is not a cliff at which an entire population suddenly becomes ill. Individual interpretation depends on the laboratory method, sample type, other electrolytes and the patient's clinical condition.
The forecast assumes that the relationship continues linearly for decades. The authors acknowledged uncertainty from changing participant numbers, environments before sampling, laboratory procedures and the possibility that the relationships are not linear. A trend-line crossing is therefore a scenario generated by the model, not a measured future outcome.
The study also did not report arterial pH or arterial carbon-dioxide pressure. Claims that blood is becoming acidic, lungs are working harder or kidneys are under constant strain require those physiological measurements and a design capable of connecting them to exposure.
Major Alternative Explanations Were Not Adjusted
In a later explanation of the research, the authors explicitly said correlation is not causation. They noted that the analysis did not adjust for diet, kidney function, diuretic use or obesity, all of which can influence the measurements. Physical activity, other medications and changes in sample collection or processing could also contribute.
Averaging every participant in a survey cycle can hide changes in the age, health and demographic composition of the sample. NHANES provides survey weights and extensive clinical variables for population analysis, but the paper reports raw cycle means rather than an adjusted individual-level model testing whether atmospheric CO2 independently predicts bicarbonate.
Calcium and phosphorus are regulated by several systems, including bone, kidneys, hormones, diet and vitamin status. Their decline alongside bicarbonate does not establish that minerals were being diverted to buffer atmospheric carbon dioxide. The mechanisms discussed in the paper draw partly on animal or human experiments at concentrations far above current outdoor air.
The original article went further still, linking the trend to obesity, diabetes, vascular change, cognitive decline and permanent renal load. The NHANES comparison did not test those outcomes, so it cannot identify rising CO2 as a hidden cause of any of them.
Indoor Air Is a Separate Hypothesis
Poorly ventilated rooms can exceed outdoor CO2 concentrations because occupants exhale the gas. Short-term indoor studies have reported effects on some cognitive and physiological measures at roughly 1,000 to 2,500 parts per million, although findings and mechanisms remain debated.
The NHANES dataset used here contained no personal indoor-air measurements. It cannot determine how long participants spent in a given building, whether ventilation changed, or what concentration they inhaled before a blood draw. Indoor exposure could be relevant to future research, but it cannot be inserted retrospectively as the cause of the observed trend.
A stronger study would link repeated personal CO2 monitoring with individual blood measurements, respiratory data, kidney function, diet, medications, physical activity and building ventilation. It would use NHANES survey design correctly, test sensitivity to laboratory changes and compare results across independent populations.
Atmospheric CO2 is rising, and reducing emissions remains essential because of its established role in climate change. That established climate evidence does not lower the standard for a new medical claim. A plausible mechanism and two parallel trend lines are the beginning of a causal investigation, not its conclusion.
The responsible reading is deliberately narrower than the headline. US serum bicarbonate averages rose while calcium and phosphorus averages fell, and the reasons remain unresolved. Until exposure is measured and confounders are addressed, the data do not show that outdoor CO2 has reset healthy blood chemistry, strained kidneys or driven chronic disease.