Amish: The Missing Control Group
Old Order Amish have minimal electricity, no personal electronics, and dramatically lower rates of every BERM-predicted condition: obesity -89%, T2D -75%, hypertension -66%, cancer -40%. However, massive lifestyle confounders (diet, exercise, smoking, community) make direct attribution impossible. The critical test is the Amish-Mennonite EMF gradient.
This page presents Amish health data as a natural experiment. The health differences are real and well-documented, but lifestyle confounders are massive. This analysis explicitly acknowledges that Amish data alone CANNOT prove EMF causation. This applies equally to conventional explanations. The gradient test is needed.
The data
Obesity: -89%
Compared to US average (STAT 2025i). Old Order Amish obesity prevalence is dramatically lower than the general US population despite similar genetic background.
Type 2 Diabetes: -75%
Anderson & Potts 2022i review of 126 studies. Amish T2D rates are approximately one-quarter of the US average.
Hypertension: -66%
Amish hypertension rates are roughly one-third of the general US population, consistent across multiple community studies.
Cancer: -40%
Overall cancer incidence is approximately 40% lower in Amish populations. Some cancers (lung, cervical) are even more dramatically reduced.
The confounders
These differences are real — but honest analysis demands acknowledging the massive lifestyle confounders.
Physical activity
Amish are far more physically active: 10,000–18,000 steps/day vs ~4,000 for average Americans. Farming, walking, manual labor are daily norms. This alone could explain substantial metabolic differences.
Diet
Less processed food, more home-grown produce, less refined sugar. Amish diets are closer to pre-industrial patterns. Dietary differences are a well-established driver of metabolic health.
Smoking
Very low smoking rates in Amish communities. This directly reduces cancer, cardiovascular disease, and respiratory conditions. A major confounder for cancer and hypertension data.
Community
Strong social bonds, low social isolation, multigenerational households, shared purpose. Social connectedness is independently associated with lower mortality, better mental health, and reduced chronic disease.
Cross-validation: Why lifestyle alone may be insufficient
Independent lines of evidence reduce — but do not eliminate — the lifestyle-only explanation.
8 species gaining weight in controlled environments (p=10⁻⁷). Lab animals with fixed diets and exercise are also gaining weight. Diet and exercise alone cannot explain this cross-species trend.
Weight-stable men still show testosterone decline. If obesity were the sole driver of T decline, weight-stable men should have stable T. They do not.
LH + T declining after controlling for BMI. The decline persists even when body mass is statistically removed. Something beyond weight is driving hormonal changes.
These independent lines REDUCE the lifestyle-only explanation. They do not prove EMF causation, but they establish that lifestyle factors alone are insufficient to explain all observed trends. This applies equally to conventional explanations.
The critical test: Amish-Mennonite EMF gradient
Old Order Amish (no electricity) → lowest EMF exposure
Conservative Amish (some electricity, limited electronics) → low EMF exposure
Mennonite (modern technology, full electricity) → moderate EMF exposure
General US population → highest EMF exposure
SAME religion, similar genetics, graduated EMF exposure. These four groups share cultural roots, religious values, and substantial genetic overlap — but differ systematically in electromagnetic field exposure. If disease rates follow the EMF gradient even within this cultural continuum, EMF becomes an independent factor that cannot be explained by diet, exercise, or community alone.
This is BERM's most important proposed population test. It is the single study design most likely to separate EMF effects from lifestyle confounders. It has not been conducted.
Derived prediction · L* level
This section describes predictions derived from the BERM framework that have not yet been directly tested. They are presented as testable hypotheses, not established findings.
Prediction E-NEW-37: Disease rates across the Amish–Mennonite gradient correlate with EMF exposure levels after controlling for diet, exercise, and other lifestyle factors.
See predictions →