Table 12.
Smoking and bone health in children and adolescents
| Reference | Study description | Population description | Number of subjects | End points | Results | |||
|---|---|---|---|---|---|---|---|---|
| Prospective studies | ||||||||
| Lappe et al. 2001 [42] | This prospective study of female army recruits examined risk factors for developing stress fractures during 8 weeks of basic training. | Sex: female Age: 21.1 ± 3.7 years Race: 31 % black, 53 % white, 16 % other Location: USA Years: 1995–1996 |
3758 | Stress fracture | Odds ratios (95 % CI) | |||
| History of smoking 1.34 (1.05–1.71) Years smoked 1.05 (1.02–1.08) | ||||||||
| •Adjusted for age, race and bone speed of sound | ||||||||
| Elgan et al. 2003 [240] | This prospective study is a 2-year follow-up of the cohort reported in 2002 and investigated the joint effects of OC use and smoking. Smoking reported as monthly average. Twenty-eight subjects smoked as follows: n = 1, >25 cig/day; n = 6, 15–24 cig/day; n = 16, 5–14 cig/day;, n = 4, ≤4 cig/day. | Sex: female Age: 18–26 Race: not reported Location: Lund, Sweden Year(s): 2001 |
118 | Peripheral DXA | Difference from reference group, nonsmoker/no OC use (n = 35) | |||
| Heel aBMD | Smoker/no OC use (n = 9) | Nonsmoker/OC use (n = 57) | Smoker/OC use (n = 17) | |||||
| T2 aBMD (g/cm2) | ||||||||
| Unadjusted | 0.0013 | 0.014 | −0.012 | |||||
| Adjusteda | −0.032* | −0.014 | −0.017* | |||||
| 2-year change in aBMD (g/cm2) | ||||||||
| Unadjusted | −0.035* | −0.013 | −0.016 | |||||
| Adjusteda | −0.06 | −0.010 | −0.021 | |||||
|
aAdjusted for age, physical activity baseline BMD, body weight *P < 0.05 | ||||||||
| Lappe et al. 2008 [241] | Secondary analyses of this randomized calcium and vitamin D supplementation trial examined other risk factors for developing stress fractures during basic training. | Sex: female Age: 19 years (17–35 years) Race: not specified Location: USA Years 2001–2006 |
5201 | Stress fracture | Risk of fracture was 41 % higher in women with history of smoking when adjusted for treatment group, P = 0.0075 Odds ratio for fracture was 1.32 (0.99–1.75) in women with history of smoking when adjusted for treatment, amenorrhea, high exercise, Depo Provera use, age, and running speed. |
|||
| Korkor et al. 2009 [231] | This 4-year prospective study assessed the relationship between smoking and alcohol intake and bone mass. Students in 9th grade were recruited. Number of cig smoked per day reported in the 9th, 10th, 11th, and 12th grades. Subjects classified as any smoking over 4 years vs not | Sex: 37 males, 72 females Age: 14–19 years at enrollment Race: white 83 %, Asian 1 %, Hispanic 4 %, unknown 12 % Location: Wisconsin, USA Year(s): 2000–2003 |
109 | Peripheral DXA | Simple model, any smokinga | Adjusted model, any alcohol or smokingb | ||
| Heel aBMD in 12th grade (g/cm2) | −0.0249, P = 0.29 | −0.0281, P = 0.05 | ||||||
| −4.3 % | −4.8 % | |||||||
|
aAny smoking over the 4-year study (yes, n = 8; no, n = 101) bAdjusted for sex baseline aBMD dairy intake Large overlap in smoking and alcohol use. Any smoking or alcohol combined (N = 14) was associated with a −0.028-g/cm2 lower aBMD. | ||||||||
| Lucas et al. 2012 [235] | This prospective study quantified the association between early initiation of smoking and alcohol intake and forearm BMD in early and late adolescence. Smoking reported at age 13 and 17 years and classified as: tried but not currently smoking, smokes but not daily or smokes daily | Sex: female Age: 13 years and 17 years Race: not specified Location: Portugal Year(s): 2003–2007 |
713 | Peripheral DXA | Smoking at age 13 years vs never smoked (n = 523) | |||
| Distal radius aBMD | Tried, not currently smoking (n = 165) | Smokes (n = 24) | P value | |||||
| At age 13 years | 1.1 % | −0.6 % | 0.410 | |||||
| At age 17 years | −2.3 % | −0.7 % | 0.068 | |||||
| Smoking at age 17 years vs never (n = 390) | ||||||||
| Tried, not currently (n = 224) | Smokes, not daily (n = 33) | Smokes, daily (n = 72) | P value | |||||
| At age 17 years | −2.3 % | −3.2 % | −0.7 % | 0.438 | ||||
| Ever smoked in adolescence vs never smoked (n = 353) | ||||||||
| At age 17 years | Tried 13 years but before 17 years (n = 167) | Tried before 13 years (n = 192) | P value | |||||
| 0.4 % | −1.8 % | 0.103 | ||||||
| Adjusted for menarche age, alcohol intake, sports, and BMI | ||||||||
| Dorn et al. 2013 [233] | This 3-year prospective study of girls examined bone accrual according to smoking, alcohol intake, depression, and anxiety. Smoking measured every 3 months and coded as cigarettes smoked in last 30 days: 0, 1–2 days, 3–5 days, 6–9, 10–19 days, 20–29 days, or 30 day | Sex: female Age: 11–17 years Race: 62 % white, 33 % black, 5 % other Location: Ohio, USA Year(s): 2003–2010 |
262 | DXA Total body BMC (g) Spine aBMD (g/cm2) Hip aBMD (g/cm2) |
No significant main effect or age × smoking interaction on total body Significant age × smoking interaction for spine and hip aBMD. As smoking increased, the rate of bone accrual decreased as girls got older. Regression models adjusted for race, puberty stage, weight, height, and contraceptive use |
|||
| Cross-sectional studies | ||||||||
| Elgan et al. 2002 [237] | This cross-sectional study conducted in nursing students measured lifestyle and physiologic factors. Smoking was measured by questionnaire and dichotomized as yes/no. 23 % smoked on a daily basis, 22 % were party smokers, and 55 % were nonsmokers. | Sex: female Age: 16–24 years Race: not reported Location: Lund, Sweden Year(s): 1999 |
218 | Peripheral DXA Heel aBMD (g/cm2) |
Smoking status not significantly associated with aBMD in bivariate analyses. Data not shown Smoking not significant (P = 0.48) in regression models adjusting for age, weight, physical activity, and hormonal age |
|||
| Afghani et al. 2003 [336] | A cross-sectional analysis of baseline data from a smoking cessation trial. Analyses investigated the role of body composition, physical activity, menarche, smoking, and second-hand smoke on regional bone mass among Asian adolescents living in Wuhan, China. Smoking questions assessed if subjects “smoked at least 100 cigarettes in their life?” and “have smoked in last 30 day.” Smoking was reclassified for analyses as smoking yes/no. | Sex: male and female Age: 12–16 years, 14.5 ± 0.5 Race: Asian Location: Wuhan, China Year(s): not reported |
466 | Peripheral DXA | Correlation with smoking (yes/no)a | P value | Multivariable modelsb | |
| Boys (n = 300) | ||||||||
| Forearm BMC | 0.05 | NS | NS | |||||
| Forearm aBMD | 0.04 | NS | – | |||||
| Heel BMC | 0.05 | NS | NS | |||||
| Heel aBMD | 0.00 | NS | – | |||||
| Girls (n = 166) | ||||||||
| Forearm BMC | 0.15 | <0.05 | NS | |||||
| Forearm aBMD | 0.04 | NS | – | |||||
| Heel BMC | 0.08 | NS | NS | |||||
| Heel aBMD | 0.09 | NS | – | |||||
|
aUnclear how correlation derived given that smoking was a dichotomous variable bAdjusting for age, height, lean mass, fat mass, team sports participation, menarche, and passive smoking Low level of smoking and inconsistent report of smoking in the sample limits conclusions. 38 % of girls and 14 % of boys inconsistently answered smoking questions. Among smokers, 58 % of girls and 51 % of boys smoked ≤1 cigarette in last 30 days. | ||||||||
| Kyriazopoulos et al. 2006 [236] | This cross-sectional study evaluated the influence of current dietary factors (calcium, proteins, alcohol, coffee, and tea intake), exercise, smoking, and sunlight on forearm bone mass in young Greek men. Smoking coded as daily smoking (58.6 %) or no | Sex: male Age: 18–30 years, mean 22 years Race: not reported Location: Greece Year(s): not reported |
300 | Peripheral DXA Distal Radius BMC Distal radius aBMD Ultradistal radius aBMD |
There was no association between daily smoking (58.6 % of sample) and bone measures with and without adjustment for height, weight, calcium intake, sunlight exposure, exercise, and work. Data not shown | |||
| Lorentzon et al. 2007 [239] | The GOOD study is a cross-sectional study involving a random selection of males in Gothenburg. Bone mass, density, and geometry measured by DXA and pQCT. Smoking quantified as cigarettes/day and duration. For analyses, subjects were classified as smokers ≥1 cig/day vs not. | Sex: male Age: 18–20 years Race: not reported Location: Gothenburg, Sweden Year(s): not reported |
1063 | Difference between smoked ≥1 cigarette/daya vs not | ||||
| Unadjusted | P value | Adjustedb | P value | |||||
| DXA | ||||||||
| Total body aBMD | −2.1 % | <0.01 | −1.8 % | 0.01 | ||||
| Spine aBMD | −4.3 % | <0.001 | −3.3 % | <0.01 | ||||
| Femoral neck aBMD | −5.3 % | <0.001 | −3.9 % | <0.01 | ||||
| Trochanter aBMD | −6.6 % | <0.001 | −5.0 % | <0.01 | ||||
| pQCT | ||||||||
| Tibia | ||||||||
| Cortical vBMD | 0.0 % | 0.91 | – | NS | ||||
| Cortical thickness | −4.5 % | <0.001 | −4.0 % | <0.001 | ||||
| Periosteal circumference | 0.0 % | 0.97 | – | NS | ||||
| Endosteal circumference | 2.5 % | <0.05 | 2.7 % | 0.01 | ||||
| Trabecular vBMD | −4.2 % | <0.01 | −3.8 % | <0.01 | ||||
| Radius | ||||||||
| Cortical vBMD | −0.1 % | 0.55 | – | NS | ||||
| Cortical thickness | −2.8 % | <0.01 | −2.9 % | <0.01 | ||||
| Periosteal circumference | 0.7 % | 0.32 | – | NS | ||||
| Endosteal circumference | 3.3 % | 0.02 | 4.5 % | <0.01 | ||||
| Trabecular vBMD | −2.7 % | 0.16 | – | NS | ||||
|
aSmokers consumed 9.3 ± 6.3 cig/day. Mean duration of smoking 4.1 ± 2.1 years. Smokers less active than nonsmokers bAdjusted for calcium intake, physical activity, age, height, and weight. Smokers had reduced BMD due to smaller cortical thickness greater endosteal circumference | ||||||||
| Dorn et al. 2011 [232] | This cross-sectional analysis of data examined how bone mass and density varied according to smoking, alcohol intake, depression, and anxiety. Smoking ever in their life coded as never (n = 104), 1 puff to 2 cig (n = 54), and 3–99 cig (n = 51), >100 cig/day (n = 52) | Sex: female Age: 11–17 years Race: 62 % white 33 % black, 5 % other Location: Ohio, USA Year(s): 2003–2007 |
261 | DXA | ||||
| Total body BMC (g) | No significant differences among smoking groups | |||||||
| Spine aBMD (g/cm2) | No significant differences among smoking groups | |||||||
| Total hip aBMD (g/cm2) | 1 puff to 2 cig group 6.5 % greater than >100 cig group, P < 0.05 | |||||||
| Femoral neck aBMD (g/cm2) | 1 puff to 2 cig group 6.0 % greater than >100 cig group, P < 0.05 | |||||||
| Analyses adjusted for age, weight, height, race, and maturational stage. Significant interactions between alcohol and tobacco use, and depression symptoms on bone outcomes. Stronger negative association between depressive symptoms and total body BMC among individuals who smoked and used alcohol. | ||||||||
| Eleftheriou et al. 2013 [238] | A cross-sectional study evaluating the association of smoking, alcohol consumption, and prior exercise with lower limb bone volume, composition, and structure by MRI and DXA in a large cohort of healthy Caucasian males. Smoking status coded as nonsmoker, ex-smoker (>6 months), recent ex-smoker (≤6 months), or current smoker | Sex: male Age: mean 19.9 years Race: Caucasian Location: UK Year(s): not reported |
651 | Differencea between smoking status vs nonsmoker (n = 329) | ||||
| Ex-smoker >6 months (n = 41) | Recent ex-smoker ≤ 6 months (n = 35) | Current smoker (n = 244) | P value | |||||
| DXA | ||||||||
| Total hip aBMD | −5.0 % | −6.0 % | −4.7 % | 0.0001 | ||||
| Femoral neck aBMD | −5.9 % | −5.3 % | −4.0 % | 0.001 | ||||
| MRI | ||||||||
| Periosteal volume | −5.0 % | −4.3 % | −0.4 % | 0.004 | ||||
| Endosteal volume | NS, data not shown | |||||||
| Cortical volume | NS, data not shown | |||||||
| aAdjusted for height weight, alcohol intake, and weight-bearing activity | ||||||||
| Winther et al. 2014 [234] | This cross-sectional population-based study compared BMD levels of Norwegian adolescents with lifestyle factors. Smoking was classified as daily smoking, sometimes smokes, or never smokes. | Sex: male and female Age: 15–17 years Race: Location: Norway Year(s): 2010–2011 |
835 | Difference between smokinga vs not | ||||
| Age adjusted | Multivariable adjusteda | |||||||
| DXA | Beta | P value | Beta | P value | ||||
| Males (n = 492) | ||||||||
| Total hip aBMD (g/cm2) | −0.039 | 0.012 | 0.029 | 0.037 | ||||
| Femoral neck aBMD (g/cm2) | −0.025 | 0.116 | – | – | ||||
| Females: (n = 469) | ||||||||
| Total hip aBMD (g/cm2) | −0.026 | 0.074 | – | – | ||||
| Femoral neck aBMD (g/cm2) | −0.031 | 0.031 | – | – | ||||
|
aIncludes daily smoking (5.5 % of girls and 3.8 % of boys) and sometimes smokes (15.9 % girls, 20.2 % of boys) vs never smokes (76.6 % of girls and 74.2 % of boys) bAdjusted for age, BMI, height, sexual maturation, physical activity, alcohol intake, diseases, and medications known to affect bone and hormonal contraceptives | ||||||||
aBMD areal bone mineral density, BMC bone mineral content, BMD bone mineral density, cig cigarette, DXA dual-energy x-ray absorptiometry, GOOD Gothenburg osteoporosis and obesity determinants, OC oral contraceptive, pQCT peripheral quantitative computed tomography