You cook with fresh ingredients. You eat vegetables. You limit sugar, fast food and ready-made meals. Perhaps you have already removed several foods, added fibre or tried probiotics.
Yet the bloating continues. Your skin still flares. The acne returns. Your joints remain painful or your energy remains low.
When someone suggests changing your diet, the reaction is understandable: “But I already eat healthy.”
This is not a lack of effort. General healthy-eating advice describes what supports health on average. It does not account for a person whose starting point already includes gut dysbiosis or chronic low-grade inflammation. A diet that helps a stable system remain stable may not be enough to change a system that has been disturbed for years.
When persistent inflammation or microbial imbalance is present, judging the diet only against general healthy-eating standards misses the central issue. What matters is whether the complete dietary pattern changes the conditions that may be sustaining the symptoms. Even substantial improvements may produce little effect if the remaining diet still supports the same microbial state.
The same food meets a different gut
The gut microbiome is the community of bacteria and other microorganisms living in the intestine. These organisms process the components of food that reach the colon. Which microorganisms are present helps determine which metabolites are produced.
This means that the same food does not necessarily produce the same response in every person. In a large human study, identical meals produced widely different post-meal glucose responses, and gut-microbiome features helped predict those differences.[1]
A piece of chocolate is a simple example. The chocolate may be identical, but the gut receiving it is not. In a person with dysbiosis, the microorganisms processing the components that reach the colon may produce a different mixture of metabolites than they would in a more balanced gut ecosystem. The physical response may therefore also differ. This shows why a food cannot be judged entirely in isolation from the person eating it.
A healthier diet does not always shift the microbiome
Changing the gut ecosystem is not simply a matter of adding healthy foods. Microorganisms respond to the full range and quantity of substrates that reach them. Even a diet that has improved substantially can still preserve the existing microbial state.
An extra apple does not correct an otherwise low-fibre week. A probiotic may pass through without becoming established if the surrounding diet does not support it. A fermented food eaten twice a week may have little effect while foods consumed every day continue to support the same microbial ecosystem.
The decisive factor is not how many dietary changes have been made. The intervention must alter the ecological conditions enough to move the system beyond the point at which the existing balance sustains itself. Until that threshold is crossed, microorganisms associated with unwanted effects may retain their position, while those performing useful functions receive too little support to expand.
This explains why several sensible—even substantial—changes can produce little visible result. The diet may be healthier than before, yet still not sufficiently targeted or consistent to shift the existing state.
Your starting point was shaped long before today
The gut does not begin each new diet from a neutral baseline. The present microbiome reflects birth, infant feeding, antibiotics, infections, medication and years of eating habits. Research suggests that this baseline has also shifted at population level in industrialised countries.
Across 14 studies involving more than 300 breastfed babies, average faecal pH rose from about 5.0 in 1926 to 6.5 in 2017, consistent with a decline in acid-producing bifidobacteria.[4]
An international study subsequently found that infant colonisation patterns followed a country’s breastfeeding history more closely than the feeding of one individual baby.[5]
More recent findings reinforce that pattern. In a 2025 US study, 24% of 412 young infants had no detectable Bifidobacterium above the study’s detection limit, while B. infantis was detected in only 8%.[6] In 2026, an analysis of more than 4,000 bacterial genomes from 48 countries found B. infantis to be common in many lower- and middle-income populations but rare in high-income countries.[7]
Together, these findings are consistent with a population-level loss of important infant gut bacteria. The historical replacement of human milk by formula probably contributed, alongside antibiotics, Caesarean birth and other changes. Human milk can feed B. infantis, but it cannot guarantee acquisition when the bacterium is no longer being transmitted.
This does not mean that every health problem begins in infancy. It means that people do not enter adult life with identical microbial capacities. Fibre and HMOs can feed suitable bacteria that are present; they cannot guarantee the return of an organism that was lost or never acquired.
Make the link between food and symptoms visible
The practical goal is to change the recurring dietary inputs enough to test whether the conditions sustaining the symptoms can shift. A person may make several sensible changes and still see little effect when the remaining diet continues to provide enough substrate for the same microbial activities.
The number of foods added or removed is therefore a poor measure of intervention strength. The complete dietary pattern must alter the ecological conditions enough to move the gut beyond the point at which the existing microbial balance continues to sustain itself. A plan that changes from day to day makes that effect difficult to evaluate.
A useful dietary intervention:
- targets foods and substrates consumed regularly rather than isolated products;
- reduces inputs that may sustain unwanted microbial activity while supplying those that support useful functions;
- follows a clearly defined and consistent plan long enough to observe a stable pattern;
- tracks a small number of specific symptoms;
- reintroduces foods systematically to assess individual responses and help identify which changes mattered.
Strength comes from the clarity and ecological relevance of the change. The intervention remains time-limited and is followed by systematic reintroduction and dietary expansion.
A low-FODMAP diet illustrates this approach. Its strict phase can substantially reduce IBS symptoms,[8] after which foods are reintroduced systematically to assess individual responses and broaden the diet.
The same principle applies more widely: create enough change to learn from the response, then build the widest and most varied diet you tolerate.
References
Zeevi D et al. Personalized nutrition by prediction of glycemic responses. Cell. 2015;163:1079–1094.
David LA et al. Diet rapidly and reproducibly alters the human gut microbiome. Nature. 2014;505:559–563.
Wastyk HC et al. Gut-microbiota-targeted diets modulate human immune status. Cell. 2021;184:4137–4153.e14.
Henrick BM et al. Elevated fecal pH indicates a profound change in the breastfed infant gut microbiome due to reduction of Bifidobacterium over the past century. mSphere. 2018;3.
Taft DH et al. Bifidobacterium species colonization in infancy: a global cross-sectional comparison by population history of breastfeeding. Nutrients. 2022;14:1423.
Jarman JB et al. Bifidobacterium deficit in United States infants drives prevalent gut dysbiosis. Communications Biology. 2025;8:867.
Shao Y et al. Genomic atlas of Bifidobacterium infantis and B. longum informs infant probiotic design. Cell. 2026;189:1854–1873.e17.
Halmos EP et al. A diet low in FODMAPs reduces symptoms of irritable bowel syndrome. Gastroenterology. 2014;146:67–75.e5.
