We read with interest the report Correlation Between Mean Temperature and Incidence of Tick-borne Diseases Among Active Duty Service Members in the Contiguous U.S., 2000–2023 by Denagamage and Mabila in the March 2025 issue of MSMR.¹

We were pleased to see the Department of Defense’s data systems being used to derive information for public health purposes. We noticed three items, however, that we think warrant attention because of how they might affect the interpretation of results. They include the potential over-diagnosis of Lyme disease, the findings between temperature and tick-borne diseases that are inconsistent with literature, and calculations based upon degrees Celsius rather than a temperature ratio scale with a true zero.

The potential for Lyme disease overdiagnosis is insufficiently addressed. The methods are based on encounter diagnoses, which may be based on clinical appearance with or without laboratory results. It is common for clinicians to over-diagnose Lyme disease based on symptoms and clinical findings alone, since diagnosis based solely on clinical presentation is unreliable. Several tick-borne illnesses have similar symptoms. A reaction to any tick bite, infected or not, may result in a skin response that could be mis-diagnosed as erythema migrans, the distinctive bullseye rash typically associated with Lyme disease. Southern tick-associated rash illness is a tick-borne disease also characterized by a rash very similar to erythema migrans of Lyme disease, though it occurs most often in the southern U.S., where Lyme disease is considered rare.² The main vector of STARI is Amblyomma americanum, an aggressive species and most abundant tick in the mid-Atlantic and Southeast. The report does not adequately address how over-diagnosis limits the interpretation of results.

The geographical distribution of Lyme disease and Ixodes scapularis activity peaks appear to be inadequately addressed. Active surveillance data in southern Virginia show that adult I. scapularis, the vector of the pathogen that causes Lyme disease, is most active in the Southeast in late winter and early fall (i.e., not hot summer months).³ Historically, Lyme disease is more prevalent in the northern and midwestern U.S.; increased Lyme disease incidence in service members in the South is inconsistent with existing literature. Even though the tick vector is seen in the South, multiple factors account for its lower Lyme disease incidence. The more notable factors include that host-seeking or questing behavior differs between I. scapularis in the North compared to the South, resulting in dramatic differences in Lyme disease prevalence. In the Southeast, ticks tend to feed on reptiles more than mammals. Since reptiles are inefficient reservoirs for Lyme disease spirochetes, there is lower incidence of the spirochetes in reptiles and ticks.⁴ Climate change may be responsible for habitat expansion but alone is insufficient to account for the changes in Lyme disease distribution.

Additionally, we had concerns about the impact of certain assumptions on the overall conclusions and statements regarding temperature. When using a ratio scale, “equality of ratios as well as equality of intervals may be determined. Fundamental to the ratio scale is a true zero point.”⁵ For example, a 6-foot-tall adult is 2 times the height of a 3-foot-tall child, which matches the ratio of 2 between the values. If these 2 were to stand in a hole 2 feet deep, the top of the adult head would be 4 feet above ground level and the child’s head would be 1 foot above ground level. Having them stand in a hole does not make the adult 4 times taller than the child.

Turning to the report, the text reports a 0.6°C increase (5.3%) in overall annual mean temperature from 2000 to 2023. We assume the starting temperature is 11.3°C, increasing to an end temperature of 11.9°C (starting and ending temperatures were not reported). The temperature scale affects the results, however. Consider the following, where the first bullet captures what was reported, the second uses Fahrenheit, and the third uses the Kelvin scale: Celsius: the range from 11.3°C to 11.9°C resulted in a 5.3% increase; Fahrenheit: the range from 52.3°F to 55.2°F would result in a 5.5% increase; Kelvin: the range from 284.5 K to 285.1 K would result in a 0.2% increase. Referring to the example, using Celsius and Fahrenheit is like standing in a hole: It misrepresents the reality. A calculation like this must be based upon a ratio scale with a true 0, such as the Kelvin scale, where a change of 0 K is identical to a change of 1°C; also, 0 K equals -273.2°C and 0°C equals 273.2 K. This affects other areas of the report, as well. In the caption of Figure 3a, the 9.9% increase of annual mean temperature from 2011 to 2012 should be reported as about 0.4%. The 2.3% increase from 2015 to 2016 should be reported as about 0.1%.

Thank you for your consideration.

Author Affiliations

Defense Centers for Public Health–Portsmouth, VA: Dr. Rockswold, Ms. Riehl

Disclaimer

The views expressed in this letter are those of the authors and do not necessarily reflect official policy of the Department of Defense, Defense Health Agency, nor the U.S. Government. The mention of any non-federal entity or product is for informational purposes only and is not to be construed, implied, nor interpreted, in any manner, as federal endorsement.

References

1. Denagamage P, Mabila SL. Correlation between mean temperature and incidence of tick-borne diseases among active duty service members in the contiguous U.S., 2000-2023. MSMR. 2025;32(3):11-19. Accessed May 23, 2025. https://www.health.mil/news/articles/2025/03/01/msmr-tick-diseases 
2. Centers for Disease Control and Prevention. About Southern Tick-Associated Rash Illness. U.S. Dept. of Health and Human Services. April 17, 2025. Updated May 15, 2024. Accessed Apr. 17, 2025. https://www.cdc.gov/lyme/about/about-southern-tick-associated-rash-illness.html 
3. Morris CN, Gaff HD, Berghaus RD, Wilson CM, Gleim ER. Tick species composition, collection rates, and phenology provide insights into tickborne disease ecology in Virginia. J Med Entomol. 2022;59(6):1993-2005. doi:10.1093/jme/tjac121 
4. Ginsberg HS, Hickling GJ, Burke RL, et al. Why Lyme disease is common in the northern US, but rare in the south: the roles of host choice, host-seeking behavior, and tick density. PLoS Biol. 2021;19(1):1-20. doi:10.1371/journal.pbio.3001066 
5. Daniel WW. Biostatistics: A Foundation for Analysis in the Health Sciences. 7th ed. Wiley Series in Probability and Statistics. John Wiley & Sons;1998.

In Reply:

We thank Dr. Rockswold and Ms. Riehl for their interest in our article, "Correlation Between Mean Temperature and Incidence of Tick-borne Diseases Among Active Duty Service Members in the Contiguous U.S., 2000–2023", published in the March 2025 issue of MSMR.

We agree that Lyme disease may have been over-diagnosed in our study, particularly in areas where erythema migrans or similar rashes are prevalent. Following the Armed Forces Health Surveillance Division Lyme disease case definition, our study sample includes unconfirmed cases from medical encounters as well as reportable medical events confirmed by laboratory data per the 2022 Armed Forces RME Guidelines and Case Definitions.¹ Despite potential over-reporting due to inclusion of unconfirmed cases from medical encounters, our study’s large sample size and lengthy surveillance period allowed identification of temporal and regional patterns that can guide efforts to protect service members from Lyme disease. Future studies using exclusively laboratory data may provide more accurate incidence estimates.

We appreciate the note on how Ixodes scapularis behavior and host preferences affect Lyme disease distribution. As stated, Lyme disease is historically more prevalent in the northern and midwestern U.S., and our crude incidence rates reflect this pattern, as shown in Table 1 (Northeast, 52.2 cases per 100,000 person-years; Upper Midwest, 13.3 cases per 100,000 person-years). Our adjusted incidence rate ratios indicate, however, a relatively higher burden in the Southeast after controlling for climatic and demographic variables. We agree that this requires further investigation to discern if adjusting for temperature and precipitation measures alongside demographic variables accounts for such a discrepancy. Analysis of entomologic and ecological factors was beyond the scope of our study, but future research including these data may clarify how tick behavior, host preferences, and habitat influence regional variation in incidence of tick-borne diseases.

While Celsius is widely used in public health literature, we recognize that percentage changes based on a scale without a true zero, such as Celsius and Fahrenheit, can be misleading. Our intent in citing percentage increases was to contextualize changes in temperature over time, while reporting temperature in Celsius was intended to make it easier to compare our results with those of similar studies. We agree, however, that expressing changes in absolute degrees Celsius, or percent change using Kelvin, are more appropriate in this context.

Reference

Armed Forces Health Surveillance Branch. Armed Forces Reportable Medical Events Guidelines and Case Definitions. Defense Health Agency, U.S. Dept. of Defense. Accessed May 22, 2025. https://www.health.mil/reference-center/publications/2022/11/01/armed-forces-reportable-medical-events-guidelines