HPV infection linked to 7% higher cardiovascular disease risk in a study of 3 million Swedish women

A letter saying that a cervical screening test has come back abnormal is one of the more common pieces of unwelcome medical news a woman can receive. It nearly always traces back to human papillomavirus, or HPV, and the concern it raises is cancer. Over the past fifteen years a second question has been attached to the virus: whether it also has something to do with heart attacks and strokes.

A study from Karolinska Institutet in Stockholm gives the most complete answer so far. Yunyang Deng, Jiayao Lei and colleagues used Sweden’s national health registers to compare almost half a million women who had a new HPV-related screening result with 2.5 million who did not, and with their own sisters. Their paper in PLOS Medicine reports a real but small excess of cardiovascular disease, and a pattern over time that complicates the idea of a virus harming the heart. The excess was largest early on, with the risk peaking during the first year of follow-up and marginally increased thereafter.

What HPV is, and why researchers link it to heart disease

HPV is a family of viruses passed on mainly through sex, and catching one is close to universal. The Karolinska team cite an estimated lifetime infection rate exceeding 80% among sexually active individuals. Most infections clear without ever being noticed. Some types persist, and persistent infection is the primary cause of nearly all cervical cancers, which is why screening programs look for the virus and for the cell changes it produces.

The suspicion about the heart began with a statistical association. In 2011, Hsu-Ko Kuo and Ken Fujise at the University of Texas Medical Branch analyzed a national survey with data from 2,450 women (age 20 to 59 years) who had sent in swabs for HPV testing. Those who carried the virus were more than twice as likely to say they had been told they had suffered a heart attack or stroke. The study was small, with 60 affected women, and it measured infection and disease at the same moment, so it could not show which came first.

Two larger studies from South Korea followed. Both came from researchers at Kangbuk Samsung Hospital, who took women who underwent a high-risk HPV test and followed them in insurance or death records. The first, of 63,411 women, found new cardiovascular disease about 25% more often in those who tested positive. The second, of 163,250 women who were tracked for up to 17 years, found far higher rates of death from artery disease in the infected group, although it rested on only 134 cardiovascular deaths.

Laboratory work offers ways the virus might do this. Viral proteins that disable a cell’s tumor defenses also disturb the lining of blood vessels in experiments, and infection can provoke low-grade inflammation, which is implicated in the fatty build-up inside arteries. The Karolinska team review these ideas and then add that the mechanisms linking the virus to cardiovascular risk are poorly understood.

How Swedish screening records were used to study HPV and heart disease

Sweden has run organized cervical screening since the 1960s, and the authors describe a well-organized national cervical cancer screening program with high participation rates, close to 80%. Every result since 1995 sits in a national registry, and each can be linked, woman by woman, to hospital records, the cause-of-death register and the population register.

The study did not test anyone for the virus. It defined infection from screening: a woman counted as exposed from the date of her first abnormal cell sample or first positive HPV test between 2006 and 2024, provided she had no earlier abnormal result and no earlier cardiovascular diagnosis. Each of these 497,445 women was matched by age with five women who had no such result at that date. The typical woman was about 32 when her follow-up began, and the cohort was followed for a median of five and a half years.

The researchers adjusted the comparison for education, income, country of birth, obesity, type 2 diabetes, high cholesterol, chronic lung disease, mental disorders, HPV vaccination and a parent’s history of cardiovascular disease.

The family analysis sets this study apart. Sisters share half their genes on average and most of their childhood. By comparing 143,787 women with HPV infection to their 174,637 unaffected full siblings, the researchers could cancel out much of what families have in common, including influences that no register records.

How much higher was cardiovascular risk after an HPV result?

Over the whole follow-up, new cardiovascular diagnoses were about 7% more common in women with an HPV result than in their matched counterparts, after adjustment. Deaths from cardiovascular causes were about 25% more common. Against their own sisters, the figures were 5% and 25%.

Those percentages need the underlying rates to make sense, because the women were young and the events rare. Each year there were about 8.3 new cardiovascular diagnoses for every thousand women with an HPV result and 7.9 for every thousand without. After adjustment, the difference came to roughly one extra diagnosis a year among every 1,900 women. For deaths the excess was about one a year among every 15,000 women.

Looked at by type of disease, the excess appeared for strokes and related conditions, blood clots, high blood pressure and heart rhythm disorders. The authors estimate that if the association were entirely causal, HPV infection would account for a little over 1% of cardiovascular disease among women in the population.

Why the first year after an HPV result stands out

Timing carries most of the information in this study. In the twelve months after the screening result, new cardiovascular diagnoses ran 43% above the comparison group. After that the excess shrank to about 2%. Cardiovascular deaths followed the same shape, falling from 86% above in the opening twelve months to 21% above afterward.

A virus that slowly inflames artery walls would be expected to show its effects over many years. A spike that arrives within months and then fades suggests that something about receiving the result is involved. The paper proposes two candidates.

One is stress. They write that receiving an abnormal cervical screening result may lead to enhanced psychological stress, with effects on blood pressure and blood vessels, and they note that new diagnoses of mental disorders in these women were also concentrated in the opening year or two. The other is attention. A woman called back after an abnormal result sees more clinicians and has more measurements taken, and the paper says that enhanced medical surveillance following an abnormal cervical screening result may facilitate earlier identification of conditions such as high blood pressure that would otherwise have been found later.

The arrow can also point backward. Women experiencing early cardiovascular symptoms may seek medical care more frequently, the authors observe, and so have more chances of being offered a screening test. They repeated the analysis without the first three months of follow-up and the results held.

None of these explanations accounts easily for the 21% excess in cardiovascular deaths that persisted beyond those twelve months, and the researchers do not think closer monitoring explains the more serious outcomes. A small lasting association remains.

What the Swedish HPV study leaves unmeasured

Smoking went largely unmeasured. It raises the risk of heart disease, and the authors treated it as a possible confounder, meaning a factor tied to both the exposure and the outcome. The registers held smoking data for only part of the cohort: the authors state that information on smoking was available only from pregnancy visits. When they restricted the analysis to women who had given birth and adjusted for smoking in pregnancy, the estimates were similar, but that is only a partial check.

The sibling comparison helps with anything sisters share. It does nothing for what they do not. The authors acknowledge that residual confounding not shared within full siblings, such as diet and physical activity, cannot be entirely ruled out.

Infection itself was inferred and never tested for directly. An abnormal cell sample was treated as infection, and the definition does not account for the dynamic nature of HPV infection, in which the virus is often cleared within a year or two. The study could not separate brief infections from persistent ones or identify which virus type was present, and it included no men.

How the result compares with earlier HPV and heart studies

The Swedish estimate is far smaller than its predecessors. The American survey suggested a doubling of risk. The Korean cohorts suggested a 25% rise in new disease and a severalfold rise in deaths. With almost three million women, about eighteen times the size of the largest Korean cohort, and with family comparisons the others lacked, the figure settles at 7%.

Epidemiologists have seen that pattern before: early, small studies of a new risk factor tend to report large effects that shrink as bigger and better-controlled ones arrive. The Karolinska authors offer a more specific reason as well. The Korean studies tested women once and so would have captured long-standing infections, whereas the Swedish design counted only new ones. If persistent infection is what matters, a study of new infections would be expected to find less.

What the finding changes for women with an HPV result

The authors frame their conclusion as reassurance. “Clinicians should be aware of the slightly short-term CVD risk in women with HPV; however, given the small absolute difference, women should not be unduly concerned,” they write. Elsewhere in the paper they call the results comforting, especially once the first year has passed.

The study does not alter cervical screening or the assessment of heart risk, and it was not designed to test whether vaccination against HPV has any bearing on cardiovascular disease.

In three million women, an HPV-related screening result went with a small and mostly short-lived rise in cardiovascular diagnoses, and whether the virus itself plays a part is still unsettled.