Exposure Risk of Dirofilaria in the Human Population

With any disease it is necessary to access the risk of acquiring that disease, but Dirofilaria has a very unique set of problems that is actually preventing the study of this disease in humans. There is no test for the disease in humans and no obvious visible symptoms. Because of this the disease can not even be studied because there is no way to determine exactly who has the disease and who does not.

The default condition in humans is an occult infection where the damage is being done silently while there is no way to even assess the pathology/morbidity associated with having this disease. So, we can not tell who has the disease but what is possible is to calculate the general probability of being exposed to Dirofilaria infection to some degree. While this certainly does not give us the complete picture it does at least give us an idea of the minimum possible exposure to the disease. So let us start to explore these numbers.

Where do we even start?

We start by looking at the animal population with respect to the exact same disease. What do we know about the exposure in dogs in the US? The numbers are somewhat inconsistent and somewhat inexact references, but we can still work with this.

With the total dog population, some references say approximately 0.24 % of all dogs in the US are diagnosed with Heartworm each year, others say 1.3%, 2.9%, 3.9% [1], so the numbers very widely based on what study you are looking at. So, lets first try to verify this number ourselves using some other referenced numbers.

It is currently estimated and mostly agreed upon that about 1,000,000 dogs become Heartworm positive in the United States each year, but only 30% [2] of them will actually be diagnosed by a licensed veterinarian. This leaves about 70% of all Heartworm infections remain undiagnosed and untreated which then go on to form the pool from which the infected mosquitoes arise, in addition to the wild animal population. This 70% of undiagnosed dogs is likely not contained within the 0.24% statistic above, so its better to logically derive our own numbers for these calculations. So let us instead verify these numbers. There are currently a total of 76,811,305 dogs in US [3] and so from here we can begin to work out the infection rate number ourselves.

1,000,000/76,811,305 = 0.013 or 1.3% dog population is Heartworm positive each year

30 % of 1,000,000 = 300,000 actually diagnosed

70 % of 1,000,000 = 700000 undiagnosed, but assumed to exist via statistical projections

300000/76,811,305 = 0.00390 or 0.39% of canine population actually diagnosed vs the 0.24% number from above.

These two numbers are different but in the the same ballpark. 0.0039-0.0024=0.0015 discrepancy

We will therefor work with the number that we can actually verify which is 0.39%.

Primary assumptions for this analysis:

  • The dog population is fairly evenly distributed among the human population. Their may be fewer dogs in the cities so the distribution is not perfectly homogeneous but this is still usable for a rough estimate. (See Cohabitation below)
  • The human owner of that dog is just as likely of being bitten by an infected mosquito as the dog is. In fact the human is perhaps more likely to be bitten since we do not have a thick fur coat that might sometimes protect a dog against insect bites, while humans almost always have some exposed skin during the prime mosquito season.
  • There are some 70% of dogs that are never brought into a veterinarian for diagnosis. This means that our calculations using 0.39 that will be performed below will be only for the lower threshold of human exposure based on the 30% group. The 70% figure could be used only to give a limit to the maximum exposure. The true answer likely resides somewhere in between these two estimates.
  • Many dogs also receive prophylactic medication to actually prevent Heartworm infection. Because this protection is not perfectly administered by their owners some unknown number of treated dogs will still get infected. The effectiveness of the medication is also decreasing in recent years due to a growing macrocyclic lactoneresistance forming in the Heartworm population. This medication artificially lowers the infection rate and therefor will put our percentage of exposures lower than they would have been without the medication.
  • Had these dogs not been prophylactically treated then the actual infection rate we will be using for humans would naturally be much higher, so again the numbers we calculate for is this lower limit will likely be well below the actual numbers.
  • Humans do not receive Heartworm preventative medication, and do not ever get treated, so the exposure to the disease in the human population is simply accumulative with each passing year of exposure being additive for their entire lifetime.
  • Some mosquitoes may have a general preference for one animal over another. We currently do not have any hard evidence to support any significant bias by all vector arthropods and there are actually many more kinds of insect vectors than “just mosquitoes” that carry this disease. The information found elsewhere on specific Dirofilaria vectors is many years out of date. The total risk of all vectors for Dirofilaria would be very difficult to untangle if they don’t yet even recognize all the proper vectors yet. In fact some studies actually show a clear preference for humans (anthropophagy) over canines [4] while other studies found that feeding patterns are not driven by differences in innate host preference [5], so its better to ignore any bias for now until all these questions are resolved. Our assumption is that feeding is an opportunistic endeavor.
  • With dogs there is a near 100% infection rate [2] upon exposure to an infected insect without any prior preventative medication. Due to the lack of a known infection rate in humans we can only evaluate the lower bound of human exposure, but not the actual risk of infection.

Census Data 2020

The census data taken in 2020 gives us a complete breakdown of the US population by age and sex,but we will ignore the male/female distribution and consider here only individual people, so we simply added these two groups together. The Census data was unfortunately collected in four year increments of age so for out purposes each block of four years was divided evenly into four individual years (¼ weight distributed to each year). This is necessary so that we can begin to calculate exposure totals for each individual year of a persons life. To remove the sharp edges on each block of four years a running 5 year averaging window technique was used to smooth the edges off for graphing purposes. The resulting smoothed data should actually be more representative of the real US population distribution. The Census also clustered all people over 70 into one final group, where we simply distributed those on a curve over the 70-75 year range thus making our data more like the actual age distribution. The few people that are over 75 are still represented but we will be ignoring their additional contribution over the age of 75 as they represent a much smaller fraction of the total population. The numbers would have been even higher if we had a proper distribution out to 105 years of age but it would not have changed the statistical significance by much and we did not have that information for the sake of absolute precision so we compromised using the range of 75 for the upper bound.

For each year of life, a person has a defined % risk of being bitten by an infected insect which is comparable to the known dog population that lives right along beside them. When an insect has many potential sources of a blood meal that insect must choose one, and the probability of being bitten is based on how many potential hosts are present in that locality. There are just 76 million dogs but 333 million humans in the US, so humans are 77% more prevalent than dogs and are therefor more readily available for that chance blood meal, so the number of humans bitten should be much higher than the number of dogs bitten.

The total number of Heartworm infected dogs also naturally decreases as dogs get treated at the vet while the human population percentage can only grow as the population ages because there is never any diagnosis or treatment. We therefor calculate the total number exposed in the human population by adding up each age bracket of the census data multiplied by the known yearly dog infection rate to arrive at the total human exposure in the current US population. And we must do this for each year of a persons life to obtain their total exposure risk.

So we take the age distribution from the 2020 census data and start by calculating the potential exposure rate for year 1. For year two there is an additional exposure risk during that year so we need to add that additional percentage to the exposure rate for the prior year. After calculating the additional exposure rate for each age group in the census population we finally arrive at a total sum of all the exposures across the whole US population. This grand total is the number we are after.

What we currently ignore here is that the likelihood of a person dying is necessarily linked to the health of that individual, so it is of course more likely that an infected person dies first, thus removing them from the general population at a rate somewhat higher than an uninfected person. We have no data that would be required to model this rate of removal so it is simply stated here as a known-unknown that needs to be explored in the future once we have actual research on this disease in humans. If we knew how much quicker infected people died from the general population this disease would be getting a lot more attention.



We have seen that there is a theoretical minimum of 41.1 million people in the US that have currently been exposed to Dirofilaria, though the actual infection rate of humans due to that exposure is still a known unknown, but this analysis should at least give a general idea of just how prevalent the disease is with respect to exposure. Looking at the maximum exposure rate, by the age 39, and over, there would be more people exposed to the disease than not exposed, and by age 75 nearly the entire population would have been exposed to the disease. Both the minimum and maximum numbers are at its peak by the age 44 where the population itself really starts to decline. This early data does not represent a causal relationship but it is certainly something that needs to be investigated if and when we find a way to determine who actually has the disease.

The known infection rate in dogs is virtually 100% as well as with most other animals that have been studied, and there is no evidence that the human infection rate should be 0%. It is very likely that people do contract the disease and it just that they are not knowledgeable enough about the symptoms of the disease to recognize it. Doctors currently have no tools to diagnose the disease and the general population has no clue what to look for. Education is required. You can have this disease and never know it until you accumulate enough parasitic load to make it obvious through the sensations called Formication. Currently if you ask your doctor about this specific set of symptoms you will just be ignored or even mistakenly sent to psychotherapy for what is actually a physical disease. Formication is recognized in many situations such as from drugs (e.g. Ritalin, a stimulant) and immunological conditions (e.g. MCAS, hyper immune response to something) but these may actually just be a form of stimulation to the parasite which then causes the symptoms. We need research.

Once a human is infected they will likely have zero resistance to a subsequent exposure thus compounding that persons eventual deterioration. So not only is exposure accumulative but the parasite load is as well. Each filaria secretes mRNA in episome like particles used to turn off the hosts immunological resistance to the existing parasitic infection, but that also lowers the bar for the second exposure to take hold. With each additional exposure the immune system will become stressed and when immunosenescence starts setting in and their general health will decline with any of a wide variety of symptoms appearing such as as chronic fatigue or general geriatric decline. In fact Dirofilaria/Wolbachia both cause hyperglycemia [6], and they actually feed from that excess blood sugar, and this condition mistakenly labeled as Diabetes Type II. None of this has ever been studied in humans because they still do not know who to study.

Note: The above simulation program and/or CSV file output are available on request.


[1] Self, S.W., Pulaski, C.N., McMahan, C.S. et al. Regional and local temporal trends in the prevalence of canine heartworm infection in the contiguous United States: 2012–2018. Parasites Vectors 12, 380 (2019). https://doi.org/10.1186/s13071-019-3633-2

[2] https://healthresearchfunding.org/21-compelling-heartworm-statistics/

[3] https://www.avma.org/resources-tools/reports-statistics/us-pet-ownership-statistics

[4] Fikrig, K., Martin, E., Dang, S., St Fleur, K., Goldsmith, H., Qu, S., Rosenthal, H., Pitcher, S., & Harrington, L. C. (2021). The Effects of Host Availability and Fitness on Aedes albopictus Blood Feeding Patterns in New York. The American Journal of Tropical Medicine and Hygiene, 106(1), 320–331. https://doi.org/10.4269/ajtmh.21-0157

[5] Fikrig, K., Rose, N., Burkett-Cadena, N., Kamgang, B., Leisnham, P. T., Mangan, J., Ponlawat, A., Rothman, S. E., Stenn, T., Mcbride, C. S., & Harrington, L. C. (2022). Aedes albopictus host odor preference does not drive observed variation in feeding patterns across eld populations. https://doi.org/10.21203/rs.3.rs-2018162/v1

[6] Rajamanickam, A., Munisankar, S., Dolla, C., Menon, P. A., Thiruvengadam, K., Nutman, T. B., & Babu, S. (2020). Helminth infection modulates systemic pro-inflammatory cytokines and chemokines implicated in type 2 diabetes mellitus pathogenesis. PLoS Neglected Tropical Diseases, 14(3). https://doi.org/10.1371/journal.pntd.0008101