The Next Frontier in Dairy · The Cost of the Gap

Two months on her own reserves carries a price far bigger than the vet bills you see.

The fresh-cow energy gap doesn't just cost you litres in the tank. It runs up six separate bills across her lactation — most never itemised on any invoice. Here they are, on one honest basis: per 1,000 cows, per year.

See the itemised bill ↓

A dairy farmer at a desk reviewing herd production figures on a tablet, the farm visible through the window — The bill for the energy gap lands here — in the figures, long after the three weeks are over.

~1,451

MJ NEL — the energy deficit behind every bill

800+ L

milk forgone per cow, at the conservative floor

£272k+

lost milk alone, per 1,000 cows / year

34 ppl

current AHDB farmgate price we cost it at

separate, individually costable bills

Where the bill comes from — in one line of biologySee the full biology →

~1,451 MJ

She opens an energy deficit in the first three weeks — demand outruns intake.

0.9 BCS

She covers it by burning her own body — fat and muscle — then rebuilds it at a loss.

survival first

The body funds maintenance and defence before milk or getting back in calf.

6 bills

So the shortfall surfaces later, in six separate places on your P&L.

The itemised bill

The invoice the fresh cow can't send you.

One energy deficit, six separate bills. Lost milk is the one line we measure directly. Three more are costed from peer-reviewed research at the lowest published incidence in each case, so the figures lean low on purpose. Two we name and deliberately leave unpriced — not because they are small, but because the study that would price them honestly for a UK herd has not been done. We would like to help do it.

Illustrative, not a prediction. Every line is a separate area of study, costed at the lowest published incidence and converted to sterling. Read any single value — or the sum — as the range of risks a transition cow faces, not as a forecast for your herd. Figures are per 1,000 cows a year, and every source named is one we hold and can send you.

Bill

What low transition intake does

Evidence

Per 1,000 cows/yr

BILL 01Lost milk

Less intake means less milk — and she never makes back the litres missed in the fresh period. The measured floor of the whole bill.800,000+ L not sold8,000 L average UK lactation × 10% = 800 L per cow × 1,000 cows = 800,000 L × £0.34/L

DMI Impact commercial trials, 6 farms, 10+ yearsOur own commercial data, not peer-reviewed. The 10% floor is the conservative end of what we measured. We do not offer a published study in support of that specific figure, and we will say so until we hold one.

£272k+directly measured

BILL 02Body condition

She burns ~0.9 of a condition score (~1,451 MJ) of her own fat and muscle to cover the gap, then buys it back in feed at a 35% penalty.~1,451 MJ mobilised/cow; +35% to replace

NASEM, 2021 · DMI Impact calculationNASEM supplies the constants: one condition score for a 650 kg cow holds 385 Mcal, and putting it back costs 520 Mcal of feed to return 343. That is the 35% penalty, and it is NASEM’s own arithmetic. The per-MJ feed price is ours.

~£15k+est. — replacement penalty*

BILL 03Disease

The deficit tips a share of cows into subclinical ketosis, displaced abomasum and metritis. We count only what she costs to treat — vet, medicine, labour. The milk, fertility and culling inside those same papers are already billed above and below, and we will not charge them twice.Ketosis 11.2% × €28.50 + metritis 6.9% × £79Read per case, not per cow: a ketosis case costs €150 and a metritis case $176 all-in. We charge only the treatment share — €28.50 and £79 — because the milk, fertility and culling inside those figures are already on Bills 01, 04 and 06. Spread the 181 affected cows across the whole 1,000, and the herd average lands at ~£8 a cow.

Suthar et al., 2013 · Steeneveld et al., 2020 · USDA NAHMS, 2014 · Liang, 201311.2% is a prevalence, not an incidence — one blood test between 2 and 15 days in milk, and the lowest of ten countries (Turkey). A convenience sample of 528 herds, May–October 2011, no UK farms. The Dutch cost model behind €150 assumes an 11% subclinical rate of its own, so the rate and the price come from the same level of disease. Of that €150, 62% is milk and 12% fertility and 7% culling — our Bills 01, 04 and 06. We take the 19% that is treatment and other disease. Metritis the same way: of Liang’s $175.77, we take the $99.41 that is vet, medicine and labour. Metritis incidence is US — no UK metritis incidence has been published, and UK figures describe endometritis, a different condition, so they are not substituted here.

~£8kest. — treatment only, overlap removed

BILL 04Fertility

Delayed cycling and lower conception push her days-open out. Cows with a low number of days open had better energy balance through the first seven weeks and needed fewer services.Named, not priced — an open question, below

Ma et al., 2020 · Bedere et al., 2018The biology is well established and we hold both papers. What does not exist is a transferable price: the cost of an extra day open is not a flat rate — it depends entirely on where a herd already sits. So we name this bill and leave it unpriced.

not pricedopen question — see below

BILL 05Emissions

Lower intake and more disease raise the carbon carried by every litre. In herd simulation, the strategy that reduced health problems cut farm emissions per litre by 3.2% and lifted gross margin by 5.0% before the cost of doing it.Named, not priced — an open question, below

Chen, Thorup & Østergaard, 2026Ten Danish herds, cradle-to-farm-gate. The 3.2% and 5.0% belong specifically to the improved-health strategy. Converting a percentage of emissions into pounds needs a UK herd carbon footprint we do not yet hold, so we do not price this line.

not pricedopen question — see below

BILL 06Early culling

Cows that don't recover, or don't get back in calf, leave the herd before they ever repay the cost of rearing them.~1% extra involuntary culls × ~£712 net

DMI Impact assumptionThe 1% figure is ours, not a published finding. Net replacement is rearing cost less cull value, worked at £1,200 − £487.50 = £712.50 on published UK figures.

~£7kest. — our own assumption

Conservative total — four bills priced, two namedLost milk measured directly; the priced bills use the lowest published incidence in each case. Disease is counted at treatment cost only, because the published per-case figures also contain milk, fertility and culling — which are billed on their own lines. Fertility and emissions are named but deliberately left unpriced. Deliberately conservative, and deliberately incomplete.

~£302k+

At the high end of the same research

The one line we measure directly has a ceiling as well as a floor. At the 16% upper figure from the same commercial trials, lost milk alone reaches:

~£435klost milk alone, per 1,000 cows / year (16% ceiling, commercial data)

The upper end of each priced line, taken separately: ketosis — 36.6% prevalence (Suthar et al. 2013, highest of the ten countries) × €422/case (Steeneveld et al. 2020, 95th percentile) ≈ £131k. Metritis — 8.0% on large operations (USDA NAHMS 2014) × $191.22 (Liang 2013, multiparous) ≈ £12k. Both of those per-case figures are contradicted inside their own papers — Steeneveld’s discussion gives €418 where its results give €422, and Liang’s Table 4.3 gives $186.33 where its text gives $191.22. We show the ceiling with that on the record, and we do not use either figure anywhere that carries a cost. These are individual, different areas of study. Any single value, or a sum of them, should be viewed as the range of risks a transition cow faces as quantified by peer-reviewed research — the ranges offered are aggregations of the low and high ends across separate fields, not a validated combined figure for any herd.

Two bills we can name but will not price — and why that matters

Fertility and emissions are real costs of the fresh-cow energy gap. The biology is established and we hold the papers. What does not exist is a study that turns either into a defensible pound figure for a British dairy herd. Rather than borrow a number from a different question, we leave these lines open — and we would like to help close them.

What an extra day open actually costs a UK herd

The only model we hold gives no per-day price to apply. Its daily figures are averages measured against a fixed 60-day baseline, not the cost of one more day; and its cheapest conception points — 113 days in first lactation, 105 in second — are where the cost is at its lowest, not where it turns positive. Under that model’s own 2013 prices the cost stayed negative all the way out to 300 days. Apply a single per-day figure to every herd and you will be wrong in both directions.

What is known: better energy balance in weeks 1–7 goes with fewer days open, fewer services and earlier cycling (Ma et al. 2020). Condition at service is among the strongest predictors of holding to first service (Bedere et al. 2018). Cost curve: Liang, 2013.

What better transition health is worth in carbon, in Britain

Danish herd simulation puts the improved-health strategy at a 3.2% cut in emissions per kilogram of milk and 5.0% more gross margin before implementation costs. Turning that into pounds needs a UK herd carbon footprint, a UK carbon price and UK herd structure. We hold none of the three.

What is known: Chen, Thorup & Østergaard, 2026 — ten Danish Holstein herds, cradle-to-farm-gate life-cycle assessment.

How common metritis actually is in UK herds

There is no published UK metritis incidence. We use a US national figure and say so on the line. UK data describes endometritis — a different condition, at a different point in the cycle — and multiplying it by a metritis cost would be wrong.

What is known: USDA NAHMS Dairy 2014 — 6.9% nationally, 8.0% on large operations, the herd size we work with.

What a case of transition disease costs a UK farm

The per-case costs on this page are Dutch and American, converted to sterling. British labour rates, vet fees, milk price and cull values are all different. A UK costing study would change every disease line here, in one direction or the other.

What is known: Steeneveld et al. 2020 (Netherlands); Liang, 2013 (United States).

These are the four studies we would most like to see run — and, as trial numbers and funding allow, to contribute to. A UK transition-cow dataset that priced fertility and carbon properly would sharpen every number on this page, including ours. We would rather help build that evidence than quietly borrow a figure from somewhere it does not belong.

The one line we cost today, in the open

800+ L × £0.34/L × 1,000 cows = £272,000+ / year

Assumptions, stated: a conservative 10%+ milk floor from DMI Impact's commercial US trials (6 farms, 10+ years — commercial data, not peer-reviewed) on a ~8,000 L lactation gives ~800 L/cow; costed at the current AHDB/Defra UK farmgate price of 34 ppl (April 2026); per 1,000 cows/year. This line rests on our own commercial data. We are not offering a peer-reviewed study in support of the 10% figure, and we will say so until we hold one.

Basis (lowest published incidence in each case; per 1,000 cows/year): Lost milk — 10%+ floor from DMI Impact commercial trials (6 farms, 10+ years), 16% at the ceiling; commercial data, not peer-reviewed. Body condition — 35% replacement penalty on ~1,451 MJ of reserves at ~£0.03/MJ NEL feed; NASEM (2021, Ch. 3) gives 385 Mcal per condition score for a 650 kg cow and 520 Mcal of feed to replace it. Disease — subclinical ketosis prevalence 11.2% (Suthar et al. 2013, lowest of ten European countries, Turkey; study mean 21.8%) × €28.50/case, being the 19% of Steeneveld et al. 2020’s €150 mean that is treatment and other disease (Fig 1B, p.9); + metritis at 6.9% (USDA APHIS NAHMS 2014, producer-identified, 17 US states) × ~£79/case, being the $89.09 vet and treatment plus $10.32 labour of Liang 2013’s $175.77 primiparous total (p.109). The remaining components of both per-case figures are milk, fertility and culling, which are billed on their own lines. Cows with subclinical ketosis had 1.5, 9.5 and 5.0 times greater odds of metritis, clinical ketosis and displaced abomasum respectively (Suthar et al. 2013). Fertility — named, not priced. Emissions — named, not priced. Early culling — ~1% extra involuntary culls (a DMI Impact assumption, not a published finding) × ~£712.50 net replacement (£1,200 rearing less £487.50 cull value). Currencies converted at ~$1=£0.79, €1=£0.85. Illustrative, not a prediction: these are separate areas of study and are not additive as a validated combined figure. Every source named here is one we hold a full copy of and can send you. Author + year shown; internal library IDs are not.

Why this page is a calculator in waiting

Six bills, one root — and one lever that reaches all of them.

Costed conservatively — and with two of the six bills left unpriced because we will not put a number on the page we cannot source — the fresh-cow energy gap runs to ~£302k+ per 1,000 cows. Most of it is money you are already spending, in six places at once. The reason it matters: if a single lever raises transition-cow intake, it doesn't ease one of these bills. It reaches all six.

The evidence behind this page — 8 sources

Every figure on this page traces to a named page in one of these documents. We hold a copy of each. Where a source contradicts itself, or where a number is ours rather than published, we say so on the line itself rather than here.

National Academies of Sciences, Engineering, and Medicine (2021) Nutrient Requirements of Dairy Cattle, 8th revised edition. Chapter 3, Energy. National Academies Press, Washington DC · www.ncbi.nlm.nih.gov/books/NBK600603/ — free full text
Suthar VS, Canelas-Raposo J, Deniz A, Heuwieser W (2013) Prevalence of subclinical ketosis and relationships with postpartum diseases in European dairy cows. Journal of Dairy Science 96(5):2925–2938 · doi.org/10.3168/jds.2012-6035
Steeneveld W, Amuta P, van Soest FJS, Jorritsma R, Hogeveen H (2020) Estimating the combined costs of clinical and subclinical ketosis in dairy cows. PLoS ONE 15(4):e0230448 · doi.org/10.1371/journal.pone.0230448
USDA APHIS National Animal Health Monitoring System (2018) Health and Management Practices on U.S. Dairy Operations, 2014. Report 3, #696.0218. United States Department of Agriculture · www.aphis.usda.gov/sites/default/files/dairy — US national data
Liang D (2013) Estimating the economic losses from diseases and extended days open with a farm-level stochastic model. MSc thesis, University of Kentucky · uknowledge.uky.edu/animalsci_etds/23/ — US model, 2013 prices
Ma J, van Hoeij RJ, Bruckmaier RM, Kok A, Lam TJGM, Kemp B, van Knegsel ATM (2020) Relationships between energy balance and metabolic and reproductive performance of dairy cows. Animals 10(6):1100 · doi.org/10.3390/ani10061100
Bedere N, Cutullic E, Delaby L, Garcia-Launay F, Disenhaus C (2018) Meta-analysis of the relationships between reproduction, milk yield and body condition score in dairy cows. Livestock Science 210:73–84 · doi.org/10.1016/j.livsci.2018.01.017
Chen L, Thorup VM, Østergaard S (2026) Herd management strategies for greenhouse gas emission mitigation in dairy production: a simulation study. Journal of Dairy Science 109(5):5334–5346 · doi.org/10.3168/jds.2025-26699

Our own data. The lost-milk figure comes from DMI Impact commercial trials rather than from any paper listed above. It is not peer-reviewed and we do not present it as though it were. A UK trial is being arranged, and when it reports we will replace the figure and say what changed.

There is one lever behind all six bills.

See how the wearable neurostimulation device supports fresh-cow intake.

See the Solution →See the proof →