Dinse2022 - Increasing ANA Prevalence in United States

Full citation: Dinse GE, Parks CG, Weinberg CR, et al. Increasing Prevalence of Antinuclear Antibodies in the United States. Arthritis & Rheumatology 2022. DOI: 10.1002/art.42330.

Raw file: [[raw/dinse2022.pdf]]

Note on publication history: This paper was originally published as Dinse et al. Arthritis & Rheumatology 2020;72(6):1026–35 (DOI: 10.1002/art.41214). It was subsequently retracted and republished as art.42330 because the CDC removed and revised some of the NHANES data used in the original study. The retraction was not due to any author wrongdoing or error; conclusions are unchanged.

Summary

This study uses NHANES serum repositories across three time periods — 1988–1991, 1999–2004, and 2011–2012 — to investigate whether ANA prevalence in the US has changed over a 25-year span. All 13,519 serum samples were tested using identical IIF methods in a single laboratory, providing methodological consistency unavailable in cross-study comparisons. The study finds a clear and statistically robust increase in ANA prevalence over time, from ~11% in the late 1980s to ~16% by 2011–2012, representing approximately 41.5 million Americans. The increase was not explained by concurrent changes in BMI, smoking, or alcohol consumption, and was most pronounced in adolescents, men, and non-Hispanic White individuals.

Study Design

• Type: Longitudinal cross-sectional trend analysis using archived NHANES serum samples
• Sample size: 13,519 participants ≥12 years across 3 time periods: 1988–1991 (n=4,727), 1999–2004 (n=4,527), 2011–2012 (n=4,265)
• Setting: Nationally representative US population (NHANES); all ANA assays performed in a single laboratory
• Population: Civilian, non-institutionalised US population ≥12 years; complex multistage sampling with population-representative weights

Key Findings

• ANA prevalence at 1:80: 11.0% (95% CI 9.7–12.6%) in 1988–1991; 11.4% (95% CI 10.2–12.8%) in 1999–2004; 16.1% (95% CI 14.4–18.0%) in 2011–2012 — P for trend <0.0001
• Estimated ANA-positive Americans: ~22.3 million (1988–91), ~26.6 million (1999–2004), ~41.5 million (2011–12)
• Sex: Increase significant in both men (P=0.0001) and women (P=0.008); OR relative to 1988–91 for 2011–12: men OR 1.76 (95% CI 1.32–2.35), women OR 1.37 (95% CI 1.09–1.73)
• Adolescents (12–19 years): Largest relative increase — 5.0% → 9.7% → 12.4%; ORs 1.00 → 2.07 → 2.77 (P for trend = 0.0004). Most concerning subgroup given potential harbinger role for future autoimmune disease
• Older adults (≥50 years): Significant increase (P=0.002); no significant trend in adults 20–49 years
• Race: Increase primarily in non-Hispanic White participants (P<0.0001); no significant trend in non-Hispanic Black or Mexican American participants
• Not explained by: Changes in BMI, smoking exposure (cotinine-confirmed), or alcohol consumption — adjusting for these had minimal impact on observed trends
• Thyroid disease association: Concurrent increase in self-reported thyroid disease; ANA rates higher in thyroid disease patients (21–24%) than without (12–16%) across all time periods
• The BMI–ANA relationship shifted: inverse association (overweight/obese had lower ANA) in earlier time periods weakened or reversed by 2011–2012

Methods Used

• Indirect Immunofluorescence ANA Test (HEp-2 substrate; NOVA Lite HEp-2 ANA slide with DAPI kit, Inova Diagnostics; 1:80 dilution; grades 1–4 = positive; grades 3–4 further titrated to 1:1,280; single laboratory; automated NOVA View image capture; dual-reader blinded review)

Entities Mentioned

(None specific to dengue)

Concepts Addressed

• Antinuclear Antibodies (central topic: temporal trend in US population 1988–2012)
• Autoimmunity in Dengue (temporal context: the denominator for ANA positivity in any US or global population has been rising; the 23.1% ANA rate in post-dengue patients must be interpreted against contemporary, not 1990s, population baselines)

Relevance & Notes

This paper is the most important temporal dataset for ANA prevalence. It establishes that healthy-population ANA positivity at 1:80 has increased from ~11% to ~16% in the US between 1988 and 2012, independent of known lifestyle confounders. This has two direct implications for dengue research:

1. When interpreting ANA positivity rates in dengue studies, the appropriate comparison is the contemporary population prevalence (~16% at 1:80 by 2011–12), not the older ~13.8% estimate from Satoh2012 - ANA Prevalence in United States. The 23.1% rate in symptomatic post-dengue patients in Garcia2009 - Long-term Clinical Symptoms Post-Dengue may be less elevated above background than initially appears — though Garcia2009 was conducted in Cuba with a Cuban population, and the specific dilution used is not clearly comparable to the 1:80 NHANES standard.

2. The dramatic increase in adolescents (5% → 12.4%) is particularly relevant for dengue research, as dengue disproportionately affects children and adolescents in endemic settings.

Connection to Satoh2012 - ANA Prevalence in United States: That study provided the 1999–2004 snapshot (13.8%); Dinse2022 uses the same NHANES data for that period (11.4% — a slight discrepancy potentially reflecting different positivity thresholds between studies: Satoh2012 used intensity ≥3 as positive, while Dinse2022 used grades 1–4).

Limitations: Cross-sectional; oldest samples are 30+ years old (though antibodies are stable in frozen storage); children <12 years not assessed; institutionalised population excluded; causes of the trend remain unknown.

Questions Raised

• What environmental or immune-dysregulation triggers are driving the increase in ANA in adolescents specifically?
• Does the rising baseline ANA prevalence reflect increasing subclinical autoimmune susceptibility — and if so, could infectious triggers like dengue be superimposed on a population with already-heightened autoimmune potential?
• How does the temporal trend in ANA prevalence compare in dengue-endemic regions such as Cuba, Southeast Asia, and South Asia?