Inside the homes of elderly residents living within 100 metres of powerloom weaving sheds in Kumarapalayam, Tamil Nadu, the average sound level was 77.6 decibels. Not in the factory. In the living room, the bedroom, the kitchen. Two kilometres away, in the control neighbourhoods, the same meters read 52.3 decibels, roughly the difference between standing beside a busy road and sitting in a quiet library.
That gap is the heart of a study by S Chitra Thara, K Vidhya Lekshmi and N Venkateswaramurthy of J.K.K. Nattraja College of Pharmacy, published in the Indian Journal of Occupational and Environmental Medicine. Between May and July 2024 the team recruited 436 residents aged 60 and over, half living beside the looms and half in quieter areas, matched on age, socioeconomic status and general health. They excluded anyone whose hearing loss had another explanation: ear infections, head trauma, ototoxic drugs, severe cognitive impairment.
The researchers measured noise with calibrated sound meters, six 15-minute readings per household, three indoors and three outdoors, taken at 8am, noon, 4pm and 8pm across three days including a weekend. Hearing was tested with a validated smartphone app running on a calibrated phone, an approach the authors note has shown roughly 93 percent sensitivity against standard audiometry in earlier work.
The exposed group heard worse. Mean thresholds came in at 37.1 decibels in the right ear and 39.4 in the left, against 25.4 and 26.8 in the control group. Anything above 25 decibels counts as impaired under WHO criteria, so the average person living near the looms was already over the line while the average control resident sat just under it. Severe hearing loss affected 11 to 13 percent of loom-side residents depending on the ear, compared with 2 to 4 percent of controls. The odds of any impairment were about four times higher.
Ageing on a steeper slope
The more interesting result is not that noise hurts hearing. It is how noise appears to interact with getting old. In the control group, each additional year of age was associated with about half a decibel of threshold loss. Among loom-side residents the slope was more than twice as steep, around 1.25 to 1.32 decibels per year. Age and hearing loss correlated moderately in the exposed group (r values of 0.52 and 0.56) and only weakly in the controls (0.24 and 0.17). The authors read this as noise amplifying the ordinary decline of ageing rather than simply adding to it.
Length of residence mattered too. Residents exposed for three to ten years averaged around 30 decibels; those who had lived beside the looms for 41 to 50 years averaged nearly 54. A regression adjusting for age found each extra year of exposure independently predicted a further 0.19 decibels of loss in the right ear and 0.23 in the left.
One finding cuts against the simple story, and the authors report it plainly. Current measured noise level at a house barely correlated with that resident's hearing at all (r = 0.034, not statistically significant). Their reading: what damages hearing is cumulative dose over decades, not whatever the meter says on a Tuesday afternoon. A single spot measurement is a poor proxy for a lifetime.
Health conditions layered on top. Diabetics in both groups had significantly worse thresholds. Smokers did too. Alcohol alone showed no effect, but smoking combined with drinking did. Hypertension on its own showed nothing, which the authors attribute partly to the small number of hypertensive participants in the sample.
Why it matters
Occupational noise rules are built around 85 decibels, the threshold at which workplaces must act. The homes in this study averaged 77.6, comfortably below that line and therefore invisible to the regulations that protect the people working inside the looms. Yet residents are exposed continuously, not for a shift. The authors' receiver operating characteristic analysis put the useful warning threshold at 74 decibels, with moderate discriminatory power (area under the curve 0.70), which suggests the protective bar for homes should sit well below the industrial one.
The study cannot prove causation, and the authors say so. It is cross-sectional, a single snapshot with no follow-up, and residence duration was self-reported, which invites recall error among elderly participants. Noise was sampled over three summer months, missing seasonal variation. Genetics and non-industrial noise went unmeasured. Some subgroups, notably those over 80 and those with comorbidities, were small enough to limit statistical power. The findings come from one rural district and may not transfer to cities.
What survives those caveats is a mapping problem. Hearing conservation has been framed as a workplace issue, with earplugs, monitoring and legal thresholds all aimed at workers. The elderly woman living across the lane from the loom shed gets none of that, and by the authors' figures she may be losing hearing faster than the man operating the machine.