Can Team Handball Build Bone After 60? The 2-Day Test

For inactive men in their late 60s, one weekly game of recreational team handball did not change bone-turnover markers. Two weekly sessions did. Three added further marker changes. That is a useful dose signal—but it is not proof that handball increased bone density or prevented fractures.

The distinction matters. Bone is slow tissue. A 16-week change in blood markers can show that bone metabolism responded to exercise; it cannot tell us whether the skeleton became denser, stronger, or less likely to break in a fall. Still, the trial offers unusually direct evidence that frequency matters when older, inactive men take up a high-intensity team sport.

Two sessions weekly crossed the marker threshold

In a 2023 randomized controlled trial, Carneiro et al. assigned 54 inactive men, average age 68 ± 4 years, to no exercise or to 60 minutes of recreational team handball one, two, or three times each week for 16 weeks. The exercise groups were small: 13 men in the once-weekly group, 15 in the twice-weekly group, and 12 in the three-times-weekly group.

This was not gentle catch-and-throw. Sessions averaged 78–80% of maximum heart rate, reached peaks of 86–89%, and covered roughly 4,676 to 5,202 meters. The sport combines running, rapid changes of direction, jumping, stopping, and physical play—loads that differ sharply from steady cycling or walking.

The investigators measured four bone-turnover markers: P1NP and osteocalcin, which relate to bone formation; CTX, a marker of bone resorption; and sclerostin, a protein that restrains bone formation. They found statistically significant time-by-group interactions for all four markers (all p ≤ 0.047).

The practical pattern was straightforward within this study. The twice-weekly group was the lowest-frequency group with a significant rise in P1NP versus the inactive control group. The once-weekly group did not differ from control on any measured marker. The three-times-weekly group showed a significant osteocalcin increase versus control and lower CTX than the once-weekly group. Sclerostin fell across the handball groups, most noticeably with two and three sessions weekly.

That makes two sessions per week the minimum effective frequency for this trial's P1NP result, in this population, over 16 weeks. It does not establish a universal bone prescription. A p-value near 0.047 in groups of 12 to 15 people deserves replication before anyone treats it like a law of physiology.

What the trial measured—and what it did not

P1NP, osteocalcin, CTX, and sclerostin can help researchers detect changes in bone metabolism before a DXA scan can reasonably show a structural difference. They are useful biological signals. They are not patient outcomes.

Carneiro and colleagues did not measure DXA-assessed bone mineral density (BMD), bone geometry, falls, or fractures. Sixteen weeks is also shorter than the roughly 12 months commonly needed to detect reliable densitometric change. So the accurate conclusion is narrow: recreational handball changed bone-turnover markers in inactive older men, with a frequency pattern favoring two or three sessions over one. The study did not show that handball built measurable bone mass.

This gap is more than technical bookkeeping. Areal BMD captures only part of skeletal strength. A 2023 systematic review and meta-analysis by Ng et al. found that moderate- to high-impact exercise improves BMD across the lifespan, while noting that evidence for bone structure—geometry and microarchitecture that also influence fracture risk—remains less certain. A better DXA number is encouraging; it is not a fracture trial in disguise.

Why impact and resistance still set the broader standard

The handball result fits a wider exercise literature. Impact loading can provide the high-magnitude, rapid strains thought to stimulate bone adaptation. But the most favorable programs in people at fracture risk generally combine impact with progressive resistance training rather than relying on impact alone.

A systematic review and meta-analysis by Rodrigues et al. reported better outcomes for multicomponent programs than for impact exercise alone. That does not mean a handball game equals a resistance program. Handball includes collisions and upper-body work, but no study here directly compares those game demands with planned, progressively loaded resistance training. Calling recreational sport a substitute for strength training would be an inference, not a finding.

Frequency also appears to matter for structural outcomes. A comparative review by Zitzmann et al. examined how training frequency relates to lumbar-spine and femoral-neck BMD in older adults and identified frequency as a meaningful moderator. The available abstract does not provide extractable effect sizes, so it cannot tell us that any particular number of sessions is optimal.

Nor does the longer evidence settle the issue. In an 18-month randomized trial analysis of progressive resistance plus weight-bearing impact exercise in middle-aged and older men, Daly et al. concluded that the optimal training frequency and dose for BMD remain uncertain. The studies ask related but different questions: Carneiro's trial tracked short-term blood markers during team sport, while Daly's analysis examined longer-term bone changes during a structured multicomponent program. Those endpoints and interventions should not be blended into one confident rule.

Age changes the odds

Older bone can respond to loading, but it may respond less than younger bone. A 2026 systematic review and meta-analysis by Miao et al. discusses reduced BMD responses to impact and jumping exercise after age 50, citing impaired osteoblast differentiation and deterioration of the lacunar-canalicular network as possible explanations. Its pooled population was premenopausal women aged 18 to 45, so its quantitative findings do not directly apply to 68-year-old men. The age-related biological caveat is relevant; the effect estimates are not transferable.

Evidence in women also exists, but it does not validate the same weekly threshold. A randomized trial by Pereira et al. tested recreational team handball in inactive postmenopausal women and reported bone-health effects alongside balance and body-composition outcomes. It did not repeat the one-versus-two-versus-three-session comparison used in the older men's trial. Postmenopausal women should not assume that two sessions weekly has been proven as their minimum effective handball dose.

Who should act on this?

For an inactive man around 60 or older who enjoys handball and can participate safely, two to three supervised recreational sessions per week is a reasonable evidence-based range to discuss with a clinician or qualified coach. In the trial, that schedule created a measurable osteogenic signal; one weekly session did not. The games also delivered substantial cardiovascular intensity, so bone is not the only consideration.

People with established osteoporosis, recent fracture, serious joint disease, unstable cardiovascular symptoms, or a high fall risk need individualized advice before joining a high-intensity contact sport. The study did not establish safety for every medical profile, and fun is not a screening test.

The verdict: Recreational handball twice weekly is the minimum frequency that changed a bone-formation marker in one small, 16-week trial of inactive older men. Three sessions produced additional marker changes. For confirmed BMD improvement, bone structure, and fracture prevention, structured impact-plus-resistance programs still have the broader evidence base—and the optimal long-term dose remains unresolved.