Measles: the virus that finds every gap

Start here: what measles is

Measles is a negative-strand RNA virus that spreads through the air, causes a systemic illness with fever and rash, and is one of the most contagious human diseases ^{[1] [2] [3]}. That last phrase is not a rhetorical flourish; it is the fact from which everything else in this review follows, including the arithmetic in the centrepiece.

The infection is not confined to the airway. The virus enters via the respiratory route, but its receptors point it straight at the immune system: SLAM (CD150), expressed on lymphocytes, and CD46 were both identified as cellular receptors for measles virus ^{[4] [5] [6]}, and their tissue distribution maps the virus's targets ^[7]. Measles is, mechanically, an infection of immune cells that happens to produce a rash ^[1].

Why it matters beyond the acute illness. This is the finding that reframed the disease. Measles does not merely suppress immunity for a few weeks — it deletes existing immune memory. Using VirScan to track antibodies against thousands of pathogen epitopes in 77 unvaccinated children before and two months after natural infection, investigators found that measles "caused elimination of 11 to 73% of the antibody repertoire across individuals." Those effects were not seen in infants vaccinated against MMR, and were confirmed in measles-infected macaques ^[8]. Mechanistic work during a Dutch outbreak found measles-virus-infected memory CD4+ and CD8+ T cells and B cells in patients, with reduced circulating memory B cells afterwards ^{[9] [10]}.

The population-level consequence had already been visible in the data before anyone could explain it. Non-measles infectious-disease mortality in high-income countries is tightly coupled to measles incidence at a lag "extending over 2 to 3 years," in both the pre- and post-vaccine eras — so the long-term immunological sequelae of measles drive interannual fluctuations in deaths from other infections ^{[11] [12] [13]}. In the authors' phrase: "By preventing measles-associated immune memory loss, vaccination protects polymicrobial herd immunity" ^[11]. Catching measles costs a child not just one illness but part of everything their immune system had already learned.

Why it is in the news. Coverage stalled and then slipped, and measles found the gaps immediately. Globally, first-dose coverage rose from 72% to 84% between 2000 and 2010 and has since plateaued at 84–85%, while reported incidence fell 88% (from 145 to 18 cases per million) before reversing ^[14]. The pattern is the same everywhere it recurs: a population that had eliminated measles, an importation, and an under-vaccinated community. New York City's 2018–19 outbreak began with one unvaccinated child returning from Israel and produced 649 confirmed cases, 93.4% of them in the Orthodox Jewish community and 72.9% in one Brooklyn neighbourhood ^[15]. California in 2014–15 ^[16], Minnesota in 2017 ^[17], Indiana in 2005 ^[18], the United States again in 2019 ^[19] and more recently still ^[20] — same script, different setting.

Three pillars follow — measurements, medicines and prevention, and progress — with the cleanest model in this whole collection in between.

Pillar 1: measurements and diagnosis

Recognising it clinically

Classic measles runs a recognisable course: a prodrome of fever with cough, coryza and conjunctivitis (the "three Cs"), then Koplik spots — small white lesions on the inside of the cheek that appear a day or two before the rash and are close to pathognomonic — then a maculopapular rash spreading downward from the face and hairline, with fever peaking as the rash appears ^{[1] [21]}.

Clinical diagnosis alone is unreliable where measles is rare, because clinicians have never seen it and because other illnesses produce fever-and-rash. This matters directly for elimination: a country that has eliminated measles has, by construction, a workforce that cannot recognise it ^[22].

Confirming it in the laboratory

Two tests do the work. Measles-specific IgM serology detects the acute antibody response, and RT-PCR on a throat or nasopharyngeal swab or urine detects viral RNA. PCR has a second job that serology cannot do: the amplified product can be genotyped, and genotype is what links cases into transmission chains and distinguishes an imported lineage from endemic circulation — the evidence base on which elimination is verified ^{[14] [23]}.

There is an important diagnostic trap in highly vaccinated populations. Secondary vaccine failure — infection in someone previously vaccinated — produces a modified illness that "may present as a modified illness that is unrecognizable as measles outside of the context of a measles outbreak," in which "the immunoglobulin M response ... may be nominal or fleeting, and viral replication may be limited." Such cases are instead characterised by high-avidity antibodies and distinctively high neutralising antibody levels, which have been proposed as biomarkers for classifying them ^{[24] [25]}. The standard test is least reliable exactly where elimination programmes most need it.

Measuring the population, not the patient

For measles the decisive measurement is not made on a patient at all. It is two-dose vaccination coverage, tracked through immunisation registries — New York City monitored MMR uptake through its Citywide Immunization Registry during the outbreak ^[15] — together with case-based surveillance sensitive enough to detect importations. Roughly half of countries did not achieve the surveillance sensitivity indicator of two or more discarded measles and rubella cases per 100,000 population ^{[14] [26]}.

Coverage is the measurement because coverage is the threshold, which is the subject of the next section.

Centerpiece: the herd-immunity threshold

Measles has the simplest and most consequential model in this collection, and it is a single line of algebra.

The basic reproduction number R₀ is the average number of people one infectious person infects in a fully susceptible population. If a fraction p of the population is already immune, only (1 − p) of contacts can be infected, so the effective reproduction number is:

R_eff = R₀ · (1 − p)

Transmission dies out when R_eff < 1, which happens once the immune fraction exceeds the herd-immunity threshold:

H_c = 1 − 1/R₀

And because a vaccine is not perfect, the coverage required is higher than the immune fraction required. With two-dose effectiveness VE, the immune fraction is p = V · VE, so the critical coverage is:

V_c = H_c / VE = (1 − 1/R₀) / VE

Read what R₀ does to that formula. For a pathogen with R₀ = 2, H_c is 50%. For R₀ = 5, it is 80%. For measles — among the most contagious human infections known ^{[2] [1]}, with R₀ estimated higher than for mumps or rubella in pre-vaccination European data ^[27] — the threshold is pushed into the mid-90s. 1 − 1/R₀ approaches 1 fast, and it does so precisely in the range where measles sits.

Grounding, and what is illustrative. The model form is textbook and the herd-immunity concept is grounded ^[28]; measles R₀ has been characterised in its own systematic review ^[29]. That review has no abstract available in the assembled substrate, so no numeric R₀ range is quoted from it, and the specific values used below (R₀ = 12, 15, 18, and VE = 97%) are illustrative and flagged. They are not, however, unconstrained. Elimination programmes are built around a grounded operational target — "maintenance of high (≥95%) 2-dose MCV coverage is crucial for controlling the spread of measles and successfully reaching measles elimination" ^[2], with Europe described as "far from the 95% coverage rates necessary for elimination" ^[30] — and the script asserts that the chosen R₀ values require coverage that brackets that 95% figure. The illustration has to reproduce the real-world target, or it fails.

What the model explains. Three things, and they are the whole public-health story.

First, why measles is the sentinel disease. Because its threshold is the highest of the routine vaccine-preventable infections, measles is the first thing to come back when coverage falls. It is the smoke alarm for an immunisation programme — outbreaks signal gaps that other, less transmissible pathogens have not yet exploited.

Second, why "95%" is not bureaucratic roundness. It falls out of the algebra. With realistic R₀ and a very good but imperfect vaccine, the required coverage lands in a narrow band in the mid-90s, and the target is set at the top of it ^{[2] [30]}.

Third — and this is why small coverage drops make headlines — the response is steep near the threshold. At R₀ = 15, letting coverage slip from 95% to 90% takes R_eff from 1.18 to 1.91: a five-point drop in coverage produces a 62% rise in transmission potential. Coverage statistics that look almost identical describe populations on opposite sides of an epidemic threshold.

Limits, honestly. The model assumes homogeneous mixing, and that assumption is exactly wrong in the way that matters most. Real under-vaccination is clustered — geographically, socially, by community — so a country can hold 95% national coverage and still contain neighbourhoods far below the threshold, which is precisely what the outbreak reports describe ^{[15] [16] [31] [32]}. National coverage is the wrong denominator; the formula applies locally. The model also treats immunity as binary and permanent, ignoring waning and secondary vaccine failure ^{[33] [24]}, and treats R₀ as a constant when it varies with contact patterns and setting ^{[27] [34] [35]}.

Pillar 2: medicines and prevention

There is no antiviral

Nothing on the shelf kills measles virus in a patient. Treatment is supportive: fluids, fever control, nutrition, oxygen where needed, and prompt treatment of the bacterial complications — especially pneumonia — that cause most measles deaths. That is the honest whole of acute therapy.

Vitamin A is the one specific adjunct

Vitamin A deficiency worsens measles outcomes, and supplementation is the exception to the "supportive only" rule. A meta-analysis of randomised trials of vitamin A supplementation in infectious disease, including three trials in children admitted to hospital with measles, found deaths from respiratory disease in the measles studies reduced by 70% (95% CI 15% to 90%, p = 0.02), alongside a 30% reduction in all-cause mortality in community trials ^[36]. Vitamin A deficiency remains a major public-health problem affecting 190 million children under five, and WHO has continued to recommend supplementation for children aged 6 to 59 months on the evidence reviewed in Cochrane ^{[37] [38] [39] [40]}. This is treatment of a nutritional deficiency that measles unmasks and worsens — not an antiviral.

Post-exposure prophylaxis

For a susceptible person already exposed, passive immunisation with immunoglobulin can prevent or attenuate disease. The Cochrane review is candid about the state of the evidence: "estimates of effectiveness have varied and no minimum effective dose has been determined" ^[41]. Vaccination given very soon after exposure is the other option.

The vaccine is the entire prevention story

MMR is a live attenuated vaccine, and its performance is what makes elimination arithmetically possible at all. Antibody persistence has been followed for 20 years in a vaccinated cohort ^[33], and measles-containing vaccination has been associated with substantial reductions in childhood mortality — plausibly beyond the measles deaths it directly prevents, consistent with the immune-amnesia mechanism ^{[42] [43] [44] [45] [46]}.

Two doses, not one, is the design. That reflects both primary vaccine failure in a minority after dose one and the very high coverage the threshold demands ^[2].

And the safety question has been answered. The claim that linked MMR to inflammatory bowel disease and, by extension, to autism has been tested directly and found wanting — measles-mumps-rubella and other measles-containing vaccines do not increase the risk for inflammatory bowel disease ^[47]. The damage that claim did to coverage is a large part of why this review has a resurgence section.

Pillar 3: progress — and what is going backwards

Elimination is being lost in places that had achieved it. The global picture is a programme that made enormous progress and then stalled: MCV1 coverage climbing from 72% to 84% over 2000–2010, then flat at 84–85%; incidence down 88% and then rising again ^{[14] [2]}. The theory of what elimination requires was worked out long ago ^{[34] [23]}, along with the strange epidemiology of a disease after elimination, when susceptibility slowly rebuilds and clinicians forget the presentation ^[22].

The proximate cause is coverage, and the cause of that is more complicated than ignorance. Vaccine hesitancy is now recognised as a threat to programme success in its own right, with its own determinants and decision psychology rather than a simple information deficit ^{[48] [49]}. It has been mapped globally through confidence surveys ^{[50] [51]}, reviewed specifically for measles in Europe ^{[52] [53]}, traced through religious and community barriers ^{[31] [54] [55]}, and addressed — with mixed results — through systematically reviewed strategies ^{[56] [30]}.

Immune amnesia is the most active science. The 2015 population-level finding ^[11] and the 2019 antibody-repertoire measurement ^[8] have been followed by mechanistic work in outbreak settings ^[9], modelling of its effect on concurrent epidemics ^[13], analyses in other populations ^[12], theoretical work on innate as well as adaptive memory ^[57], and early attempts at therapeutic mitigation ^{[58] [59]}. The practical upshot is already clear: the benefit of measles vaccination is larger than the measles cases it prevents.

SSPE is the complication that arrives years later. Subacute sclerosing panencephalitis is a fatal degenerative brain disease caused by persistent measles virus, appearing after a long latency ^{[60] [61] [62]}. A California review of 1998–2015 cases found SSPE diagnosed at a median age of 12 years with a latency of 9.5 years (range 2.5–34), and 71% of cases had a measles-like illness before 15 months of age. Its incidence estimate is the alarming part: among measles cases reported during 1988–1991, SSPE occurred in 1 in 1,367 children infected under age 5, and 1 in 609 of those infected under 12 months — far more common than previously recognised ^[63]. Vaccination has correspondingly changed SSPE epidemiology ^{[64] [65] [66] [67]}, and the virology of brain adaptation is being worked out ^{[68] [69]}. Because infants below vaccination age carry the highest SSPE risk, they are protected only by the herd immunity of everyone around them — which returns the argument to the threshold.