Changes in USDA food composition data for 43 garden crops, from 1950 to 1999 - Review of the evidence
By comparing USDA tables, the study found apparent average decreases in six nutrients across 43 crops between 1950 and 1999. It therefore supports the hypothesis of lower nutritional density, but does not prove the role of the soil. Independent studies show convergent signals, with significant methodological limitations and natural variability in the crops measured.
- Published 09/11/2026
- Updated 11/11/2026
- Article history
ABSTRACT
The analyzed study compared the nutritional composition data from the United States Department of Agriculture for 43 garden crops, primarily vegetables, between 1950 and 1999. After adjusting for water content, the authors observed statistically reliable average reductions in protein, calcium, phosphorus, iron, riboflavin, and vitamin C; the median decreases reported in the abstract ranged from 6% for protein to 38% for riboflavin.
In "The Choice to Live in Health or Illness" by Dr. A. Colonnese and R. Panzironi, the publication contributes to the narrative that selection for higher yields may be associated with a lower concentration of certain nutrients in modern vegetables. Independent studies on historical data for fruits, vegetables, and grains detect signals consistent with this possibility, while a critical review emphasizes that comparing tables compiled in different eras does not allow for attributing the variations with certainty to soil depletion or a single agricultural practice. The most prudent conclusion is that there is evidence of changes in the density of some nutrients, but the causes, magnitude, and overall nutritional relevance require specific evaluations by crop and context.
Analyzed study
Changes in USDA food composition data for 43 garden crops, 1950 to 1999
DOI: 10.1080/07315724.2004.10719409
PMID: 15637215
The research question was whether the nutritional data published by the USDA for 43 garden crops had changed between 1950 and 1999. The hypothesis evaluated was that, across the set of foods considered, some nutrients might be present in lower concentrations in the more recent data.
The research is relevant to the section "The Choice to Live in Health or Illness" by Dr. A. Colonnese and R. Panzironi because it directly compares different historical periods and detects an average reduction in protein and some micronutrients. Its specific contribution therefore concerns the segment of the narrative that describes a possible decrease in nutritional density, i.e., the amount of nutrients contained in a given amount of food.
The study did not directly measure the soil, root absorption, agricultural yields, or plant growth rates. Therefore, it cannot prove that the soil supported a higher production volume nor establish a causal relationship between industrial agriculture and every variation detected. The authors proposed as the simplest explanation potential differences in the cultivated varieties, with possible trade-offs between yield and nutrient concentration.
Materials and methods of the original study
This publication is a historical comparison of food composition data. The authors compared USDA values from 1950 and 1999 for 43 horticultural crops, primarily vegetables, across 13 nutrients and water content.
To prevent variations in water content from making nutrients appear more or less concentrated, the data were adjusted for moisture content. For each food and nutrient, the ratio between the 1999 value and the 1950 value was calculated. A ratio of less than 1 indicates a lower value in 1999 compared to 1950. The authors then used the median and geometric mean to describe the group of 43 foods.
To evaluate individual food-nutrient combinations, the standard errors of the 1999 USDA values and two hypothetical estimates of the standard error for the 1950 data were used. The standard error expresses the uncertainty associated with an estimate. The available abstract does not provide the full list of crops, the number of analytical samples underlying the individual USDA entries, the laboratories responsible for the historical analyses, the sample selection criteria, or further statistical details.
Results
Considering all 43 crops together, the study found apparently statistically reliable decreases for six nutrients: protein, calcium, phosphorus, iron, riboflavin, and ascorbic acid, i.e., vitamin C. The median reductions reported in the abstract ranged from 6% for protein to 38% for riboflavin. For seven other nutrients, no statistically reliable changes emerged.
The most important result for the book's section is directional: in the most recent data, the food group showed, on average, lower concentrations of certain proteins, minerals, and vitamins compared to the 1950 data. However, the available material does not report the specific percentage values for calcium, iron, and vitamin C cited in the book; therefore, these values cannot be directly verified from the provided abstract.
When the authors examined each food and each nutrient separately, the majority of the ratios were not distinguishable from 1 with the available data. Depending on the estimate adopted for the uncertainty of the 1950 values, only 33% or 20% of the apparent ratios differed reliably from 1. Furthermore, about 28% of the statistically distinguishable ratios indicated an increase, not a reduction. This means that the signal was observable primarily at an aggregate level and did not demonstrate a uniform decline for every single vegetable or nutrient.
Interpretation
In simple terms, the study suggests that some vegetables listed in the 1999 USDA tables were, on average, less concentrated in certain nutrients compared to the corresponding entries from 1950. This does not mean that modern fruits and vegetables lack nutritional value, nor that every contemporary variety contains fewer nutrients.
The argument supported by the research discussed in the article is: historical comparison of horticultural foods → average decrease of some nutrients in more recent data. The authors interpreted any real decreases as consistent with changes in cultivated varieties and a possible trade-off between high yield and nutrient concentration. A trade-off of this kind is often called the dilution effect: if the accumulation of carbohydrates and biomass increases more rapidly than the accumulation of minerals or proteins, the concentration of the latter may decrease.
This interpretation is not proof that the soil is depleted or that increased production is the sole cause. Food tables can differ over time due to variety, geographical area, season, ripeness, analytical procedures, and sample size. For this reason, the correspondence with the book's narrative is clear on an observational level, but the causal explanation requires additional experimental evidence.
Independent confirming evidence
An independent analysis by White and Broadley compared historical data from the United Kingdom and the United States for edible horticultural products. In U.S. vegetables, they observed significant reductions in dry matter concentrations of calcium, copper, and iron; in fruits, copper, iron, and potassium decreased. The result directly confirms the part of the narrative regarding possible historical decreases of certain minerals in groups of fruits and vegetables. However, the authors clarified that the data were insufficient to demonstrate reliable changes in every single plant species and that the impact on the overall mineral intake of the diet could not be automatically inferred [1].
Mayer and colleagues re-examined British tables from 1940, 1991, and 2019. In the aggregated analysis, all elements considered except phosphorus decreased between 1940 and 2019; the largest overall reductions concerned sodium, iron, and copper, while water content increased by 1%. This study reinforces the observation that historical differences in databases may show a lower concentration of minerals, including iron. However, the authors point to several possible explanations, including varieties, agronomic factors, and increased CO2, without identifying one as a proven cause [2].
A more robust design for analyzing the possible link between yield and minerals was adopted by Fan and colleagues in wheat. The researchers analyzed 362 archival samples from a British agricultural experiment started in 1843. After the introduction of high-yield semi-dwarf cultivars from the late 1960s, the concentrations of zinc, copper, and magnesium in the grain decreased significantly in all studied lots; the most recent averages were 33-49% lower for zinc, 25-39% for copper, and 20-27% for magnesium compared to earlier cultivars. Soil concentrations, on the other hand, had remained stable or increased. The work therefore supports the mechanistic shift suggesting that selection for yield and biomass redistribution may contribute to the dilution of some minerals, at least in wheat and at that specific experimental site [3].
Overall, these studies do not demonstrate a single chain of events valid for all vegetables, fruits, and cereals. However, they show that the trend observed in the original study is biologically plausible and that, in some contexts, it can be linked to cultivars and yields rather than a simple loss of minerals from the soil.
Conflicting evidence
A critical review by Marles challenges the interpretation that differences between food composition tables published in different eras demonstrate a generalized decline in minerals due to soil depletion. The review highlights that geographic origin, cultivar, stage of maturity, sampling methods, analytical techniques, and statistical treatment can change over time, making historical comparisons difficult to interpret.
The author acknowledges that some high-yield varieties may exhibit lower mineral concentrations due to a dilution effect, but concludes that the available data do not justify a general claim of mineral depletion in fruits, vegetables, and grains. This evidence does not invalidate the observation of the original study, but it reduces the certainty with which it can be attributed to a single cause and invites a distinction between variations in concentration, the nutritional magnitude of the effect, and the actual overall quality of the diet [4].
Summary table of evidence
The table distinguishes the historical comparison of the original study from independent verifications and the main relevant critical evidence.
| Reference | Study type | Population or model | Main result | Direction of evidence | Strength of evidence |
|---|---|---|---|---|---|
| Original study | Historical comparison of food tables | 43 USDA horticultural crops, 1950 and 1999 | Reliable average reductions for six nutrients; not uniform for every food | direct confirmation | moderate |
| [1] | Historical comparison of food databases | Fruits, vegetables, and nuts in the UK and USA | Calcium, copper, and iron decreased in US vegetables based on dry matter | direct confirmation | moderate |
| [2] | Historical comparison of food tables | Fruits and vegetables in the UK, 1940-2019 | Aggregate reduction of all elements except phosphorus; iron reduced by 50% | direct confirmation | moderate |
| [3] | Experimental study with archival samples | 362 wheat samples from a British agricultural experiment | Decrease in various minerals following high-yield semi-dwarf cultivars, without soil depletion | biologically consistent evidence | moderate |
| [4] | Critical review | Literature on fruits, vegetables, and cereals | Comparisons between historical tables alone do not prove soil depletion or generalized decline | discordant evidence | moderate |
Confirmation of the evidence
Claim in the book
In paragraph 1.2.4, “The Choice to Live in Health or Illness” by Dr. A. Colonnese and R. Panzironi, it is stated that higher yields may be associated with a dilution of nutrients in modern agricultural products, and it reports, for 43 horticultural crops between 1950 and 1999, decreases in protein, calcium, iron, riboflavin, and vitamin C.
Selection of cultivars and increased yield → greater biomass or production → possible dilution of some nutrients → lower average concentration in some foods
The publication subject to the Review of evidence was used in the book to support the following passage: in the USDA tables relating to 43 horticultural crops, some nutrients were on average less concentrated in 1999 compared to 1950; a possible explanation proposed by the authors is the trade-off between agricultural yield and nutritional density.
The scientific correspondence is direct regarding historical observation: the analyzed study notes an average decline in protein and certain micronutrients in the group of crops considered. Independent evidence has detected historical reductions of some minerals in U.S. and British vegetables, including calcium and iron, but does not allow for the results to be extended uniformly to every species or variety [1,2].
The causal link proposed in the book, however, requires greater caution. In wheat, a long-term experiment showed that the introduction of high-yield semi-dwarf cultivars was associated with lower concentrations of several minerals in the grain, while soil minerals did not decrease. This supports the dilution mechanism in a cereal and in a well-documented context, but it does not prove that the same process explains all the historical variations observed in fruits and vegetables [3].
The available critical review also emphasizes that comparisons between tables compiled in different decades may reflect uncontrolled methodological and biological differences. Therefore, in "The Choice to Live in Health or Illness" by Dr. A. Colonnese and R. Panzironi, the narrative is supported by the observational phenomenon of possible concentration reductions, but the attribution to soil or industrial practices cannot be considered generally proven [4].
Overall level of confirmation
MODERATE CONFIRMATION
The historical decrease of certain nutrients in food groups is confirmed by independent analyses. The potential role of high-yield cultivars is experimentally supported in wheat, but the evidence does not demonstrate a single cause nor an automatic generalization to all fruits and vegetables.
Conclusions
The study analyzed provides observational support for the fact that, between 1950 and 1999, USDA tables for 43 garden crops reported lower average concentrations for certain proteins, vitamins, and minerals. Independent literature confirms that similar historical variations have also been detected in other sets of fruits, vegetables, and grains.
The explanation based on dilution linked to higher varieties and yields is moderately supported, especially by experimental data in wheat. However, it is not proven that depleted soil, industrial agriculture, or early harvesting are the general cause of the differences reported in historical tables. The situation requires contemporary studies that compare different varieties grown under the same conditions, using identical analytical methods and repeated sampling.
Overall assessment
The narrative is moderately supported: there are convergent signs of lower historical concentrations for some nutrients, and the dilution mechanism is biologically plausible. However, the association does not prove a general depletion of the soil, nor does it allow for attributing all differences to high yields or industrial agriculture.
References
[1] Historical variation in the mineral composition of edible horticultural products
[2] Historical changes in the mineral content of fruit and vegetables in the UK from 1940 to 2019: a concern for human nutrition and agriculture
[4] Mineral nutrient composition of vegetables, fruits and grains: The context of reports of apparent historical declines
Editorial note
Artificial Intelligence was used solely as a tool to support editorial and documentary work. All content, interpretations, bibliographic citations and conclusions have been verified, reviewed and approved by the author, who assumes full scientific responsibility for them.
The manuscript was also subject to independent peer review prior to publication.
To ensure maximum transparency, the article's complete history — including the dates of submission, review, acceptance and publication, as well as information on the peer review process — is available on the dedicated Article History page.