Bone Density Test: What the T-Score Says, and What It Leaves Out

Most chronic diseases announce themselves at some point. High blood pressure turns up at a routine measurement, diabetes on a blood test, heart disease often as pain on exertion. Osteoporosis has no such stage. There is no pain, no symptom, nothing to feel — until the fracture. That is the single reason bone density testing exists: to find a disease that shows itself in no other way, while there is still a window in which something can be done.

It is worth being explicit about the stakes, because the word “fracture” is misleading. A hip fracture in a person in their seventies is not an injury you recover from and move on. It means surgery, hospitalisation, weeks of rehabilitation, and often a permanent step down in independence — a substantial share of people never return to their previous function, and mortality in the following year is markedly higher than among their peers. A vertebral fracture, sometimes occurring without any single event and found incidentally on an X-ray, predicts further fractures. This is a quiet disease with a loud ending.

What a bone density scan actually measures

The standard test is DXA — a scan that passes X-rays through the body at two energy levels. Bone and soft tissue absorb the two energies in different proportions, and from the difference the machine calculates mineral per unit area. The scan takes minutes, needs no preparation or injection, and delivers a very low radiation dose — less than a plain chest X-ray, on the order of a few days of ordinary background radiation. On safety grounds, this is not a test that needs agonising over.

Two sites are scanned: the lumbar spine and the proximal hip. Not because they are convenient, but because these are the fractures that decide the outcome. The spine is rich in trabecular bone, which turns over quickly and so responds early to hormonal change or treatment; the hip is the fracture that changes lives, and femoral neck density is the best predictor of it among the sites scanned. In older people this matters: degenerative vertebral changes, aortic calcification and old compression fractures artificially inflate the spine reading, which is why the hip value carries the weight.

One limitation is worth understanding. DXA measures areal, not volumetric, density — it is influenced by bone size, and a small-framed person tends to read lower. Nor does it measure bone quality, the microscopic architecture that determines how much force a bone absorbs before it breaks. The number is good, but it is a proxy.

T-score and Z-score answer two different questions

Every report returns two numbers, and most of the confusion comes from treating them as the same thing. The T-score compares you with the peak density of a healthy young adult of your sex, in standard deviations. It underpins the diagnostic definitions, and applies to postmenopausal women and men over fifty. The Z-score compares you with people your own age. It is not used for diagnosis, but it asks a different question: is this abnormal for this age? A value unusually low for age obliges a search for a secondary cause — vitamin D deficiency, an overactive parathyroid, coeliac disease or another malabsorptive state, long-term steroids, hormonal decline, or a thyroid hormone dose that suppresses TSH over years. In a younger person, or a woman before menopause, the Z-score comes first.

Most fractures happen in the middle

Here is what makes the test genuinely useful, and it runs against intuition. The lower the density, the higher the individual risk — true and clear. Yet most fragility fractures occur in people whose density sits in the osteopenic range, not the osteoporotic one, simply because far more people are in that range. Anyone who treats the T-score as a switch — below the threshold treat, above it relax — misses the group in which most fractures happen.

A proper assessment therefore does not end at the number. Fracture risk calculators, of which FRAX is the best known, estimate the probability of fracture over the next ten years from age, sex, body mass index, previous fracture, parental hip fracture, current smoking, steroid use, rheumatoid arthritis and high alcohol intake — with or without the density value. Calibrated by country, the tool adds what the scan alone cannot: two women with the same T-score, one fifty-five and one seventy-five with a previous wrist fracture, live in two different worlds of risk. They also need two different decisions.

These tools have limitations of their own. They do not price in steroid dose, the number of previous fractures or a tendency to fall, and they underestimate risk in someone who has fallen several times in the past year. They are the start of a conversation, not a verdict.

Bone density is one side of the equation. The other is how often you fall and from what height — and no scan measures that.

When to test, and who should not wait for the age

For someone with no risk factors, the accepted guidelines place the test at around sixty-five for women. For men the guidelines are less uniform, and usually point to seventy and above. Blanket scanning of healthy forty-five-year-olds yields little: most get a normal result that changes nothing, and some get a borderline value that leads to anxiety and unnecessary follow-up. Eligibility for funding through the health funds (kupot holim) likewise depends on age and risk factors, and the exact criteria and approved frequency are not the same for everyone. It is worth checking in advance with your fund, and what your supplementary insurance (bituach mashlim) covers. This sits inside a broader logic of screening by age.

Some people, though, should be tested much earlier, usually for a defined reason:

• A fracture from a fall at standing height after fifty — including a wrist or rib fracture that looked “trivial”
• Early menopause, before forty-five, or surgical removal of the ovaries
• A hip fracture in a parent
• Oral steroid treatment for more than three months, at any age
• Low body mass index, significant weight loss, or a history of an eating disorder
• Coeliac disease, inflammatory bowel disease, or bariatric surgery — any malabsorptive state
• Aromatase inhibitors for breast cancer, or androgen deprivation for prostate cancer
• Current smoking, high alcohol intake, or prolonged inactivity
• An untreated hormonal disorder — see hormones and the thyroid

For women, the natural anchor is the years around menopause. Bone loss accelerates markedly in the years before the final period and the first years after it, then moderates. This is the one window in which a single measurement is not only diagnostic but sets a reference point: a baseline for the next decade, precisely when the slope is steepest.

What actually moves the number

Bone is living tissue that responds to load. The stimulus that builds it is mechanical effort that is relatively intense and brief — resistance training at progressively heavier loads, and jumping or running that puts controlled impact through the skeleton. Swimming and cycling, excellent for cardiorespiratory fitness, barely load the skeleton. Honesty requires saying that the effect of training on the number itself is modest — a few percent at best — and no substitute for drug treatment where there is a clear indication. But preserving bone, muscle and balance together reduces falls, a second and no less important route to lowering fracture risk.

In nutrition, two things recur. Enough protein — the bone matrix is collagen, and a low-protein diet harms bone and muscle alike. And vitamin D and calcium at sufficiency, not megadoses. Very high doses of vitamin D have not shown a density advantage, and in some studies were associated with a worse outcome. Calcium is better obtained from food, with a supplement reserved for those who cannot reach a reasonable intake by diet. Stopping smoking and cutting back on alcohol do more for bone than people tend to assume.

Bone and muscle move together

Bone loss and muscle loss are not two parallel processes that happen to share a decade. Muscle applies the force that bone adapts to; lose muscle mass and you lose the stimulus that maintains the skeleton. So the measures of the movement system belong to the same conversation: grip strength, muscle mass, and the ability to rise from a chair without using your hands. The same DXA machine that scans the hip can also measure body composition, so the separation between a “bone test” and a “muscle test” is technical, not biological. Within a comprehensive assessment the right approach is to look at them together.

The failure is not in diagnosis, it is in persistence

And here is the most important caveat in this article. A scan without a plan changes nothing. Medicine diagnoses osteoporosis well, and holds treatments that substantially reduce fractures in people with a clear indication. The problem begins after the prescription. A large share of patients stop treatment within the first year. And after a hip fracture, when the risk of another is among the highest there is, a substantial proportion never start at all. That is true in Israel and in most developed countries.

The reasons are understandable. The treatment feels like nothing, the disease feels like nothing, and the rare side effects that received wide media coverage frighten more than a fracture that has not yet happened. An honest conversation weighs the two directly: those very rare effects against the fracture risk of someone with an indication. With some drugs a planned pause after several years is possible; with others, stopping without a follow-on treatment raises risk instead — which is why this is a decision made with the physician rather than alone. That conversation should be had properly, once, and not left to the leaflet inside the box.

What to do with the result

If the test has already been done, four things make it actionable. First, ask for the numbers themselves — T-score and Z-score for each site — not the word “normal”. Second, ask for a ten-year risk calculation using the femoral neck value, particularly if the result is osteopenic. Third, before accepting a diagnosis, rule out a secondary cause: vitamin D, calcium, kidney function, urinary calcium, the thyroid and the parathyroid. Fourth, have any repeat scan done at the same facility on the same machine — a small difference between scans on different machines is usually measurement noise, not real change.

Above all: an osteopenic result is an early warning, not a diagnosis. This is the decade in which loading the skeleton, enough protein and orderly follow-up still bend the line. A diagnosis of osteoporosis calls for a treatment decision and someone to make sure it lasts beyond the third month. If you are not sure which group you belong to, you can book a consultation and go through the data you already have.